The History of Melanesian Society, Volume 1

CHAPTER I

INTRODUCTION

There are two chief kinds of ethnographical work ; one, intensive, in which the whole of the culture of a people, their physical characters and environment are examined as minutely as possible ; the other, survey-work in which a number of peoples are studied sufficiently to obtain a general idea of their affinities in physique and culture both with each other and with peoples elsewhere. There is one feature of low forms of culture which makes these two kinds of work essentially different. In civilised culture we are accustomed to distinguish certain definite departments of social life which can to a large extent be kept apart, but among those people we usually speak of as primitive, these departments are inextricably interwoven and interdependent so that it is hopeless to expect to obtain a complete account of any one department without covering the whole field. In consequence, however deeply one may attempt to go in survey- work, the information gained must inevitably be incomplete and can never possess the accuracy which an intensive study would have given. Another feature of survey-work which has the same effect is that the proper valuation of the evidence of witnesses is impossible.

The work of the Percy Sladen Trust Expedition to the Solomon Islands falls into two distinct parts; intensive work done by Mr A. M. Hocart and myself in the Western Solomon Islands and by Mr G. C. Wheeler in the islands of Bougainville Straits, to be recorded in other volumes ; and survey- work, done by myself during the journey to and from the Solomon Islands, which is the subject of the present book. I have begun by mentioning the difference between intensive and surveywork because I wish at the outset to indicate most emphatically

R. M. S, I

that the work to be recorded in this book belongs to the second kind. It labours not only under the general disadvantages of survey-work, but much of the material was collected during hasty visits to islands, sometimes of only a few hours' duration, and much of it can only be taken as a rough preliminary account of social conditions which will, I hope, be more thoroughly studied before it is too late. At the same time I need hardly say that there has only been included such material as I believe to be correct in essentials, though doubtless often incomplete or even inaccurate in detail. In several cases in which I had grave doubts concerning the value of my material, I have been greatly encouraged by its having received general confirmation through accounts sent to me by various friends who have made further inquiries during the three years which have passed since my visit.

In the arrangement of the material I have departed from the order in which the islands were visited and I therefore give here my itinerary. A month was spent in the Hawaian Islands, a month in Fiji and a day or two only in some of the Samoan and Tongan Islands. Then I travelled in the Southern Cross, the vessel of the Melanesian Mission, through the New Hebrides, the Banks, Torres and Santa Cruz Islands to the Solomons, the same groups being revisited six months later in the reverse order on the way back. Brief visits were also paid to Tikopia and Bellona. In the Eastern Solomon Islands only survey-work was done which will be recorded in this volume and I have also added a short preliminary account of certain results of the intensive work of Mr Hocart and myself in the Western British Solomons which bear most closely on the chief topics of the book.

The information obtained in the various islands visited by the Southern Cross was supplemented by work with natives who were travelling from one island to another and such information forms a large proportion of the evidence to be recorded in this volume. It follows that most of my information was obtained from natives professing Christianity, much of it indeed from native teachers of the Melanesian Mission. I am perfectly confident, however, that their new religion has had little or no influence on most of the facts I have recorded. I deal chiefly with the details of social organisation which suffer little if any change as the result of missionary influence and least of all in the field of work of the Melanesian Mission.

Sources of Error 3

whose policy it has been since the time of Bishop Patteson to preserve as far as possible native institutions. I f this book had dealt with the magico-religious institutions, the case would no doubt have been different, but even here the accounts given of the secret societies and magic of the Banks Islands well show how many native institutions or their traditions have persisted.

At the present moment there exists in Melanesia an influence far more likely to produce disintegration of native institutions than the work of missionaries. I refer to the "repatriation" of Melanesian labourers from Queensland which has been the result of the movement for a "white Australia." Large numbers of men, women and children have recently returned to nearly every Melanesian island. Some had been many years in Queensland and have quite forgotten all they ever knew of their native institutions ; some even have that contempt for these institutions which so often accompanies a smattering of " civilisation." Their influence on native institutions in the future must almost certainly be great, but this influence is so comparatively recent that I do not believe it has had any appreciable eflect on the social conditions which I record in this volume.

I undertook the survey with the definite object of studying the subjects of kinship and marriage. My previous work in Torres Straits and India had shown me the fundamental importance to the science of Sociology of the method of counting relationships which among most of the races of mankind differs so greatly from that practised by ourselves. I had reached the belief that in systems of relationship we have, like fossils, the hidden indications of ancient social institutions and that their study is essential for advance in our knowledge of prehistoric sociology. I hope in this book to demonstrate the correctness of this belief; to show, indeed, that systems of relationship are far more vitally important and their investigation far more fruitful than my utmost hopes had led me to anticipate. I mention this here because there is an undoubted danger in thus going into the field with a dominant belief. To those who fear this possibility I should like to point out two things. One is that systems of relationship are bodies of dry fact the accuracy of which, especially when collected by the genealogical method \ is about as

• See Sociological Review., 1910, iii, i.

incapable of being influenced by bias, conscious or unconscious, as any subject that can be imagined. The other fact is that nearly all the conclusions which have been reached have been the result of a comparative study undertaken only after my return to England, and of many of my most important conclusions, conclusions upon which the whole of my theoretical construction rests, I had not the faintest idea till more than a year after my return, so that any bias from this source in the collection of the data can be absolutely excluded.

I have endeavoured throughout my record to make as clear as possible the degree of its trustworthiness, for owing to the conditions under which the material was collected its value as evidence varies considerably in the different regions of my survey. A most important element in the valuation of evidence is the nature of the linguistic medium through which it is obtained and I therefore give here a brief sketch of this feature of method. Nearly all the evidence from the Banks and Torres Islands was obtained in the Mota language, this having been possible only through the kind help of members of the Melanesian Mission. The information given by the natives was translated into English for my benefit and my questions or requests for further information were translated into Mota. A large proportion of the information from the Banks Islands was obtained from only one man, John Pantutun, but his acute intelligence and his readiness to acknowledge when he was ignorant gave us great confidence in his trustworthiness. In Pentecost Island the information was obtained both in the Mota language and in that of the northern part of the island, Miss E. Wilson kindly acting as interpreter in the latter case. The only native witness who was available for much of the information fell far short of the Banks and Torres informants in intelligence.

The Eastern Solomon Islands form perhaps the least satisfactory region of my survey from the point of view of trustworthiness. For much of the material I was dependent on pidgin English, though I was occasionally indebted for help to those who knew the native languages. The few facts I give from the more western islands form part of an intensive study and can be trusted. The value of the evidence from Tikopia is considered very fully in the chapter on that island. The material from the other Polynesian islands was obtained altogether from natives who knew English or by their help

Methods 5

from others who knew only their own language. In Fiji I had to trust wholly to the services of an interpreter.

In my record I have followed the plan of a previous book^ in keeping distinct from one another the facts collected in the field and the inferences drawn from their study, and I have tried to do this even more completely than before by dividing this book into two distinct portions, in the first of which I give little more than a bare narrative of my field-work, while the second portion consists of a comparative study of these results and of those of previous workers together with their theoretical discussion. Occasionally in the first portion I have gone beyond the bare facts and have discussed their meaning and drawn inferences, but only when such inferences have been immediately suggested by the conditions within a limited region and are not the result of comparative study. Even when facts are discussed in the first portion of this book I hope that there will never be the slightest doubt as to what is observed fact and what is inference.

Nearly all the systems of relationship and certain other data, such as those concerning property in the Banks Islands, were obtained by the genealogical method, though I have only recorded the pedigrees upon which my account is based in a few cases, chiefly in order to illustrate especially complex relationships. Without this method, it would have been quite impossible to have collected the material, not merely on account of the short time at my disposal but also because of the many difficulties, linguistic and otherwise, which are inevitable in such survey- work as I was carrying out. It is certain that without the genealogical method this book could never have been written.

As I have already said, my chief object in making the survey was to obtain systems of relationship together with such other facts concerning marriage, descent and other social institutions as would assist the interpretation of the systems. While on my journey, however, I did not neglect any opportunities of acquiring knowledge on other topics that came in my way and two such subjects seemed to me so important that I devoted much time to their study, even in some cases at the expense of my primary object.

One of these opportunities was provided by the presence on the Southern Cross of John Maresere, a native of Wallis

• The Todas^ London, 1906.

Island, who had lived for twenty years on Tikopia. A few inquiries showed that his knowledge of the customs of this little Polynesian settlement was so great that every effort was exerted to make the account obtained from him as complete as possible. The other opportunity was the readiness of John Pantutun to talk about the ritual of the Sukwe and Tamate organisations of the Banks Islands and especially of Mota, his own home.

These accounts are straightforward and simple and require no special introduction. With my chief topic, systems of relationship, the case is different. All the systems recorded in this volume are of the kind called classificatory, which differs greatly from our own and is to most people so far from easy to understand that a short preliminary account of its main features may not be out of place. The best way to make a complex condition intelligible is to show what it means and I propose therefore to give here a brief theoretical account of the classificatory system. This account will not be accepted by all, but even to those who do not accept it, the scheme may be useful as a means of coordinating the many complexities of the systems to be recorded in this volume.

By many writers the classificatory system of relationship has been held to be of little social significance. McLennan supposed it to be merely a system of titles of address and Westermarck and others have held the same opinion and have attached little or no weight to the value of the system as evidence of social conditions. That the classificatory system is merely what it was supposed to be by McLennan must now be given up by all, for it has been found that the designations of the system carry with them all kinds of duties, privileges and restrictions and are evidently of the utmost social importance and significance. Nevertheless, the former depreciation of the importance of these systems is still having its consequences, shown by the fact that anthropologists of repute still give the most elaborate accounts of social institutions without recording or even mentioning the system of relationship and among those of less repute this neglect is almost universal.

It will, I believe, materially assist the study of the many systems recorded in this book if it be recognised that the essential distinction between the classificatory system and

The Classificatory System 7

that used by ourselves is that the former is founded on the clan^ or other similar social group while our own is founded on the family. A term like that of brother which, except in the metaphorical sense, is limited among ourselves to the male children of our own parents, is in the classificatory system extended to all the male members of the clan of the same generation and also, probably by a later development, to others, of different clans but of the same generation, with whom there exist certain relationships. In the terms of the classificatory system no distinction is made between own brothers and sisters and those whom we should call first, second, third or even more remote cousins, or members of the same social group with whom no genealogical connection can be traced. This character applies to all the terms of relationship ; the term father is used for all those whom the father would call brother and therefore for all the members of the clan of the preceding generation together with other men of that generation. Similarly, all the men of his clan whom the real grandfather would call brother receive the same designation as the grandfather himself.

The same holds good of relatives on the mother's side and of the relatives of the wife or husband. Thus, most classificatory systems have a special term for the mother's brother, but this is used for all those whom the mother would call brother, not only for all the members of the clan of the mother who are of her generation but also for other men to whom for any reason the mother would apply the term which she gives to her brothers. Similarly, the terms for the father and mother of the wife apply not only to these two persons but each is used for the large group of people to whom the wife would apply the terms which she uses for her real father and mother.

Though there cannot in my opinion be the slightest doubt that the classificatory system is founded on the clan or other similar social group, it is used by many people among whom the clan-organisation does not exist, but in all such cases it is certainly a survival of such an organisation, a method of counting relationships which has persisted after the social

' By "clan," I mean an exogamous group within a tribe or other community, all the members of which are held to be related to one another, and bound together by a common tie. In general, this tie is either a behef in common descent from some ancestor, real or mythical, or the common possession of a totem.

organisation which has produced it has passed away. In such cases, however, the terms are usually limited to those with whom some genealogical connection can be traced or with whom the tradition of such a connection persists.

The first and essential character of the classificatory system is, then, this grouping together under one denomination of a large body of relatives whom we distinguish ; but this is far from exhausting its characteristic features. The classificatory system usually distinguishes certain relatives whom we group together ; thus, there is in general a definite distinction between the brother of the father and the brother of the mother to both of whom we apply the same term and usually, though less frequently, there is a corresponding distinction between the sister of the father and the sister of the mother, both of whom we call "aunt."

The distinctions between the father's brother and the mother's brother, between the father's sister and the mother's sister and between the children of these relatives are themselves the natural consequences of the origin of the system in a clan-organisation, for owing to the practice of exogamy, the brothers of the father must always belong to a clan different from that of the brothers of the mother and so with the other examples mentioned. Whenever one finds that these distinctions do not exist, it is also found that the clan-organisation is absent or is in course of profound modification.

Another feature which is very general in the classificatory system is that of reciprocity. It is especially pronounced in the Melanesian systems with which this book will deal and also occurs in Polynesian systems. Relatives, perhaps of different generations, it may be of widely different age, will use between them only one term when addressing, or speaking of, one another. Thus, we shall see that in Melanesia it frequently happens that a man will address or speak of his grandson by the same term which the latter uses when he addresses or speaks of his grandfather. We may say that there is only one, and that a reciprocal term for both grandfather and grandson instead of the two terms which we employ.

This very frequent feature of classificatory systems may be put in another way. It may be taken to indicate that the terms of these systems denote relationships rather than

The Classificatory System g

relatives. Thus, in the example just given, we may say that there is one term for the grandparent-grandchild relationship in the place of our two terms, grandparent and grandchild.

Another widespread feature of the classificatory system, which is probably to be brought into relation with the principle of reciprocity, is a very characteristic and, to us, most peculiar manner of using the terms for brother and sister. In many systems, including most of those to be recorded in this volume, two brothers have a term or terms which they use in addressing or speaking of one another and the same term or terms may be used by two sisters, or these may have terms special to themselves. A brother and sister, on the other hand, use a different term or terms. Most frequently they use only one term which may be regarded as a term for the brother-sister relationship, being used by each in addressing or speaking of the other. There is thus a term or terms for the brother-brother relationship ; the same or a different term or terms for the sister-sister relationship, and a term or terms for the brother-sister relationship. It may help the reader to grasp the nature of this distinction if I point out that it is probably merely another consequence of the clan-organisation and of certain social conditions often associated with it. The early separation of brothers and sisters is a frequent custom among savage peoples and is often met with in Melanesia. While two brothers are thus constant companions and also two sisters, a brother and sister will at an early age come to belong to different social groups within the clan and this separation will be accentuated when the girl marries and, perhaps at an early age, moves to and becomes for all practical purposes a member of another clan. A feature of the classificatory system which seems to us strange and bizarre becomes perfectly natural if the system has been the result of social conditions often associated with the clan-organisation.

Another frequent feature of the classificatory system is the use of terms to denote difference of age, a feature which is especially pronounced in the relationship of brothers to one another and of sisters to one another. Thus, in Melanesia and many other parts of the world an elder and a younger brother apply different terms to one another and use these also in speaking of one another, and the same or

corresponding terms are often used by two sisters. Such distinction accordinof to asfe is, on the other hand, rare, if not unknown, in the case of the brother-sister relationship. A similar distinction usually applies also to distant relatives who under the classificatory principle receive the same name as the brother or sister. A man distinguishes all men of his clan and of his own generation and others whom he calls by the same term as his brothers, as elder or younger, but this does not always depend on the actual relative age. It may depend on this or it may depend on the relative ages of two persons from whom the men are descended. Thus, in the case of the grandsons of two brothers, the grandson of the elder brother may be addressed as elder in some systems even if he is actually younger than the speaker.

This distinction according to age is also often applied to the brothers of the father, usually by means of words meaning "great" and "little" respectively, and this may be carried so far that the actual father, if a cadet of his family, will be only a "little father" while his elder brother, the uncle of the speaker, will be "great father," and this superiority of designation may even carry with it a higher degree of respect and honour ; the " great father " or uncle will be honoured more than the real father.

Owing to these and many other differences between the classificatory system and such a system as our own, English terms of relationship have no real equivalents in the languages of those who use the classificatory system, while the terms of Polynesia and Melanesia are without real equivalents in our language. Consequently there would seem to be endless scope for misunderstanding if English terms of relationship are used in the description of classificatory systems. It would seem as if the only really satisfactory plan would be to employ symbols for the different relationships and it is probable that the time will come when this will be done and many parts of the description of the social systems of savage tribes will resemble a work on mathematics in which the results will be expressed by symbols, in some cases even in the form of equations. The adoption of such a practice at the present time does not seem advisable. The technicality of the subject is already so great that such use of symbols would probably deter most from reading the book at all and thus defeat its chief end, viz., the demonstration of the vast importance of

Terminology 1 1

these systems and the urgent need for their systematic collection throughout the world. We must wait till the subject has become far more familiar than it is at present, even to professed anthropologists, before such symbolic representation will become advisable.

It is clear that, in a book such as this which deals exclusively with classificatory systems of various kinds, our familiar English terms will have to be used in an unfamiliar way and it therefore becomes necessary to define as closely as possible the sense in which they will be used. It will hardly be possible to say in every case whether terms are being used in the classificatory or the ordinary English sense and since all the systems described and discussed in this book are of the classificatory kind, it may be laid down as a general rule that in future the English terms of relationship must be taken in the classificatory sense except when the contrary is stated or when the context makes it clear that they are being used in the English sense. Thus, in general, if I say that a man marries his father's sister, I shall mean that he marries a woman who has the same status and is denoted by the same name as the own sister of his own father. The only general exception to this rule will be in the case of the terms " husband " and " wife," which I shall use in the customary English sense, including under this sense the status involved in simple polygyny or polyandry.

English terms of relationship, however, differ very greatly in definiteness of meaning and there are certain terms which it will often be convenient to use the connotation of which must be more carefully considered. There are certain English terms of relationship such as father, mother, son, daughter, brother and sister which, except when used in a metaphorical sense, are definitely limited to single persons, or to small groups of persons, all related in the same way. There are others, such as grandfather, grandmother, grandson and granddaughter, uncle, aunt, nephew and niece, which include relatives of different kinds, often, and indeed usually, distinguished in classificatory systems ; and there is a third group, such as cousin, brother-in-law and sister-in-law which are used very vaguely, including a wide circle of relatives. If, as has been settled, English terms of relationship are to be used in this book, it might seem necessary for purposes of accuracy to use only the first group of terms which have clear and

unambiguous meanings, and in referring to grandparents or cousins to speak always of the father's father or mother's father; of the child of the father's brother or the child of the mother's brother, etc. Such a practice if carried out systematically would, I am afraid, give the descriptions and discussions of this book an almost unendurable severity and it will often be convenient to use the less definite terms. I give therefore a brief account of the connotation of several of the more ambiguous terms and state how I propose to use them.

Uncle and aunt. These terms include two distinct classes of relative, the brother and sister of the father and the brother and sister of the mother. As a rule I shall only use these terms with the qualification "paternal" or "maternal," except when the context makes the exact meaning quite obvious. Similarly, the terms nephew and niece will only be used when it is clear whether they apply to the child of a brother or sister.

Grandparent and grandchild. As used among ourselves there are two kinds of grandfather, the father's father and the mother's father, and two corresponding kinds of grandmother, and similarly two kinds of grandson or granddaughter, the son's child and the daughter's child, and since there are often different terms for these different kinds in the systems to be recorded they must be distinguished. I shall only use the terms "grandparent" or "grandchild" either with the necessary qualification or when the context makes the exact meaning perfectly clear.

Cousin. This is liable to be a most misleading and dangerous term when dealing with the classificatory system, and yet for the purpose of brevity it will often be convenient to use it. We often apply the term to those of a generation different from our own, as when we speak of a cousin once or twice removed, but whenever I use the term "cousin " it may be taken as certain that this sense is absolutely excluded. In the more limited sense there are four kinds of cousin, the children respectively of the father's brother, the father's sister, the mother's brother and the mother's sister, and when I use the term " cousin " without qualification it may be taken as a general term for all these relatives, the terms brother and sister in the relationships just mentioned being used in the classificatory sense, so as to include those whom we should call first, second, and third cousins.

Terminology 1 3

Often in the classificatory system these four kinds of cousin are grouped in two categories ; the children of the father's brother and of the mother's sister fall into one category and those of the mother's brother and of the father's sister into another. Another way of putting this distinction is that the children of two brothers or of two sisters fall into one category and the children of a brother and sister into the other. It may help the reader to grasp the necessity for, and meaning of, these two categories if I point out that with exogamy the children of two brothers or of two sisters will or may belong to the same social group, while the children of brother and sister must necessarily belong to different social groups. It will often be convenient to have a term by means of which cousins, the children of brother and sister, may be distinguished from cousins, the children of two brothers or of two sisters, and I propose to call the former "cross-cousins," adopting this from the expression "crosscousin marriage " introduced by Professor Tylor. This re- lationship is so important and often so fundamental and will have to be so often considered in Melanesia that it will be convenient to have some simple term to denote it.

Parents-in-law. There are four possible kinds of parentin-law ; the wife's father and mother and the husband's father and mother, and in some forms of the classificatory system each is denoted by a separate term. Whenever I use the term " parent-in-law " all four kinds will be understood to be included and when the father- or mother-in-law is mentioned, it will be understood that both wife's and husband's father in the one case and wife's and husband's mother in the other case are included, except when the context makes it clear that only one of these relatives is intended.

Brother- and sister-in-law. I do not know whether it is generally recognised that these terms, even when used in their more limited sense, denote no less than eight different relatives who are not only often distinguished in the classificatory system but also differ in very important social functions. In addition to the more strict usage, the terms in question are also often used in English for other relationships such as that set up between men who have married sisters or women who have married brothers and these again are often denoted by quite distinct terms in classificatory systems.

It may assist in the understanding of the subject if I point

out that the eight relatives called brother- or sister-in-law in the more strict English sense may be grouped together in four relationships ; one existing between men, a second between women, and the other two between men and women. The first relationship is that between a man and his wife's brother, or reciprocally between a man and his sister's husband. The second relationship is that between a woman and her husband's sister, or reciprocally between a woman and her brother's wife. The third relationship is between a man and his wife's sister, which is reciprocally that between a woman and her sister's husband. The fourth and last relationship is that between a man and his brother's wife, which is reciprocally that between a woman and her husband's brother. It may be mentioned here that in many classificatory systems these relationships are denoted in a manner closely analogous to that used for brothers and sisters, i.e. there is a term to denote the relationship between men ; sometimes the same, sometimes another term, to denote the relationship between women ; and other terms to denote the relationships between men and women. The English terms include so many different relatives that I shall use them very rarely and only when the context makes it quite clear to which category or categories of relationship reference is being made.

I have already referred to the fact that in many forms of the classificatory system a brother or sister is denoted by one term when a man is speaking and by another term when a woman is speaking. The same is true of relatives by marriage, as when the relationship between a man and his brother's wife is denoted by one name and that between a woman and her brother's wife by another. The same also holds good, though more rarely, of other relationships ; thus, in some systems a man and his wife use different terms for their children or for their son's children or their daughter's children. In consequence it is often necessary to indicate whether a man or woman is speaking, and this will be done throughout this book by inserting after the relationship the letters (m. s.) or (w. s.), (man speaking) and (woman speaking) respectively.

In Melanesia terms of relationship are generally used together with a possessive pronoun, often in the form of an inseparable suffix. Further, in some cases the form of the possessive is different in different relationships and it is probable that these differences are of considerable significance.

Definitions 1 5

It will therefore be convenient in general to give the terms of relationship together with the possessive pronoun and in most cases the pronoun chosen will be that of the first person singular. When, as in the comparative tables, the pronoun is omitted, it must be remembered that this is only for the sake of brevity and that the words are never heard in this form.

Definition of Terms.

In the descriptive portion of this book I propose to use as few technical English terms as possible, and certain words such as religion and totemism will not be found till the theoretical part of the book is reached. I hoped to have been able to avoid the use of the word " magic " in the descriptive portion and if I had full knowledge of the native terms, this would perhaps have been possible ; but, as it is, I have been driven to group together a number of facts under this heading the definition of which I do not at present intend to attempt. It must be recognised that my use of the term is purely a matter of practical convenience adopted in order to avoid undue cumbrousness of expression.

In using the terms "ghost" and "spirit" I shall follow the lead of Dr Codrington by whom "ghost" is limited to a personality which has once been a man, while " spirit " is used of a being which has never, according to native belief, been incorporated in a human body.

Of terms applying to social structure I have already defined clan (see p. 7).

By the word "tribe" which occurs in this definition* I mean a group of a simple kind, always in Melanesia settled in a definite locality, which speaks a common language and is capable of uniting for common action as in warfare.

I limit the term "descent" to membership of a social group so that when I speak of patrilineal and matrilineal descent, I shall mean a condition in which a person belongs to the social group of his father and mother respectively. I limit "succession" to the transmission of rank or other similar social distinction so that when I speak of patrilineal succession,

^ This and the following definitions are in general those which have been recommended for adoption in the forthcoming new edition of Notes and Queries by a committee consisting of Miss Burne, Miss Freire-Marreco, Mr Marett and myself.

I shall mean that the rank of chief or other special position in society is transmitted to a man from his father. "Inheritance" will apply solely to property. In Melanesia matrilineal descent may be associated with patrilineal succession and inheritance, or succession may be patrilineal while both descent and inheritance are matrilineal. In this part of the world the special form taken by matrilineal inheritance and succession is that property and rank are transmitted to a man from his mother's brother.

In general I shall speak of "relationship" rather than of " kinship." When I use the latter term I shall mean relationship, real or conventional, other than relationship by marriage, which can be demonstrated genealogically.

In recording systems of relationship, and still more in considering them theoretically, I shall have frequently to refer to certain features which I propose to call " correspondences." I use this term when two or more relationships are included together under one term. A typical and frequent example of such a correspondence is that a person applies the same term to his mother's brother, to the husband of his father's sister and to his wife's father. Another example is that the father's father and the elder brother are denoted by the same term. When in future I use the term "correspondence" I shall mean such classing together of different relationships in nomenclature.

Of geographical terms I shall in general use "region" for a group of islands, an island, or perhaps only part of an island, characterised by the possession of a common culture, while the term "district" will denote a smaller geographical unit, usually part of an island, inhabited by a group of people corresponding closely with the definition of a tribe. In the nomenclature customary in Melanesia there is much scope for confusion between a village and a district. The term village, as often used, denotes a district in my sense about which houses or small groups of houses are scattered. When I use this term, it will denote houses and other buildings which form a discrete group definitely separated from other similar groups of buildings in the same district.

Several of the communities to be described in this book possess a special form of social grouping, the dual organisation, and this opportunity may be taken to point out certain peculiar features possessed by this form of society.

The Dual Organisation 17

In the dual organisation the whole population consists of two exogamous groups which I call moieties, a man of one moiety having to marry a woman of the other. Further, in every case where this form of social organisation is known to exist, descent is in the female line so that a man belonofs to the moiety of his mother. This has certain results illustrated by the accompanying diagram in which the members of one moiety are denoted by the letters A ox a and those of the

A=b

I 1

I

-| I 1

a B b

Other moiety by B or b, the capital letters denoting males and those of ordinary type females. This diagram makes it clear that a man and his own children must necessarily belong to different moieties and a man and his sister's children to the same, while of the two kinds of grandchildren one kind, the son's children of a man, belongs to the same moiety as himself while the other kind, his daughter's children, belongs to the opposite moiety. Further, as a result of the classificatory principle which is always found coexisting with the dual organisation, the whole community of the generation immediately succeeding that of a man will fall into two categories, (i) those classed with his own children, including his brother's children ; and (ii) those classed with his sister's children. If, as usual, a man applies one term to a son and another term to a sister's son, every male of the community of the generation immediately succeeding his own will be denoted by one or other of these terms ; such a male must be either his son or his sister's son, using these terms in the classificatory sense. Similarly, in the next generation, all persons will again fall into two categories, (i) those classed with his son's children, and (ii) those classed with his daughter's children, for the children of his sister will necessarily have married persons classed with his own children so that the children of his sister's son will necessarily be also the children of his daughter, using the terms son and daughter in the classificatory sense, while the children of his sister's daughter must be also the children of his son.

This character of the dual organisation of society will have certain consequences concerning marriage. So long as

1 8 Melanesian Society \

marriage is confined to members of the same community and to those of the same generation, a man must marry a woman who is both the daughter of his mother's brother and the daughter of his father's sister, using these terms in the classificatory sense. If a man marries a person not of his own generation, his choice is similarly limited ; if he marries a woman of the generation next below his own, he must marry one who has the status of his daughter although she may have married his sister's son and thus have acquired the status of the sister's son's wife ; similarly if he marries a woman two generations below his own, he must marry his daughter's daughter in the classificatory sense for his son's daughter would be of the same moiety as himselP.

Phonetic System.

I may consider here briefly the signs which I shall use to denote the sounds of the Melanesian languages. The missionaries to whom we owe most of the written records of these languages have used certain simplified signs to denote sounds which would ordinarily be expressed in English by a combination of consonants, and in one case, that of Fiji, these signs have been adopted officially. Unfortunately these signs have not been used uniformly in different parts of Melanesia and even within the region of one mission a given sign may be used for a large number of different sounds, the mission having had in view the purely practical purpose of providing the natives with the simplest signs possible. The method which has generally been adopted is to use certain redundant English letters, such as q and c. Thus the letter q has been used in Fiji for the ng of the English word "finger," while by the Melanesian Mission it has been used for a large variety of sounds which according to Dr Codrington include pw, mbw, kw, kpw, kbw, kmbw, knipw, and nggmbw. For scientific purposes it is of the utmost importance that the actual sound shall be expressed, for such differences as those between kiv, pw and mbw may be of definite ethnological significance. I have therefore rejected the use of the letter q and have endeavoured to give the sounds as nearly as possible as I heard them, though it does not of course follow that my rendering is correct or even universal in the place

• In this account I have assumed conditions simpler than those of the dual organisation as it now exists in Melanesia.

I

Phonetic System 19

with which I am dealing. Thus in Mota where according to Dr Codrington the sound in question is properly kpw I usually heard it as the simpler kw^ and I have used this throughout.

Another feature of the various phonetic systems used by missionaries and others in Melanesia is that they do not indicate whether the b and d sounds are strengthened by m and n respectively, so that the letter b is used both for the b and mb sounds and the letter d for both d and nd. Here again the differences may have definite ethnological significance, and wherever I heard mb or nd these sounds are so expressed.

There are in Melanesia two kinds of m, that for which I have used the sign m sounding to the English ear very much like mw and in words beginning with a capital letter I have used Mw in place of iii. I use ng for the sound of the English word "singer" and ngg for that of "finger." Dh stands for the th of " though," the harder sound of " through " not being represented in Melanesia so far as I know. Zh represents the sound in the word "fusion" and ck and / have their English values^

The simple letter g (apart from ng and ngg) when used in native words will generally signify a special guttural consonant which is known as the Melanesian g. It resembles the Arabic Ghen and though it varies much in character in different islands, it might often be more correctly transliterated ghr and to the untrained ear the resemblance to r is so close that in place-names in Melanesia the letter r has sometimes been used by mistake for the Melanesian g.

In the case of the vowel sounds I have followed as nearly as practicable the spelling of Codrington^ and other missionaries who have reduced Melanesian languages to writing. The exact value of the letters differs in different islands and in the short time at my disposal it was not possible to attend closely to such variations. The only sign which need be especially mentioned is e which stands for the vowel of the French le.

• This simplification of the proper sound is probably one of the many results of the use of Mota as a lingua franca.

2 For the exact values of other Melanesian consonants see Codrington, Melanesian Languages, p. 198. This book will be quoted hereafter as M.L.

3 M.L., 197; see also Codrington and Palmer, Dictionary of the Language of Mota, i8g6.

2 — 2

CHAPTER II

BANKS ISLANDS

Social Organisation, Relationship, Marriage and

Property.

The group of islands named after Sir Joseph Banks lies immediately to the north of the New Hebrides and consists of a number of islands ranging in size from Vanua Lava which is 1 5 miles in length to Rowa and Merig which are only tiny islets.

The island with which my account will chiefly deal, Mota* (PI. I), is little more than two miles in diameter, the natives inhabiting a raised coral base surrounding the central volcanic cone.

Dr Codrington's account^ has made anthropologists familiar with the Banks Islands as an example of the dual organisation of society and as the seat of a highly developed secret organisation. The following account is intended to supplement that of Dr Codrington by describing in greater detail certain customs and institutions of the islands.

The Social Organisation.

All the islands, with the exception of Merlav, possess the dual organisation of society, there being two chief social groups or moieties, called veve or vev, a term also meaning "mother," and the members of the same moiety are sogoi to one another. In general these groups have no special

• The initial letter of this and other islands of the group, viz. Motlav, Merlav and Merig, should be M {Miv) but the spelling with the ordinary M is so well established that I do not venture to alter it.

2 The Melanesians: Studies in their Anthropology and Folk-Lore^ Oxford, 1 89 1, quoted hereafter as M.

17 O"

Torres I*

Q Toga J Urepaj-aparal

^ ^t^-^^Saddlel BANKS VamiA Layri^-fY^^rtgaJ. ISLANDS

Santa Maria U-) .Men^/

OMerlavI.

IS*

Espiritu Santo I

Lepers L \ ) Auroral. '^oj \ \ Pentecost I.

Malikolo

!V^

cT^S Aanbrym I.

^

^ 'O^Api I.

Mai I. (P

O Mdkuyal. o Mataaol,

Sandv^ich l.C^ ?

(EfaU) * —

^

Eromanjyal.

oArtuvOyJ.

Ok

2c

20°

I.OYALTV ,^^

Halgaxi I ""^Av (UeoQ "^

ATUJuteturv I. '^

JL

17SL

Map oj Banks and Torres Islands and Neiv Hebrides.

names\ but people are spoken of as belonging to the same side or to the other side of the house. The only exceptions of which I could learn were that in the Lakon district of Santa Maria one of the two vez^e, believed to have been created by the sun, is called Atealoa while in Mota one of the two moieties certainly has a name, being called the Takwong- {ta, belonging to, and kwong, night). According to one informant the other moiety of Mota is called Tatalai, talai being the name of the giant clam-shell. It is doubtful whether this name is in general use, the Takwong people usually referring to the members of the other division as tavala ima, the other side of the house, but it will be convenient to use it in the following description. In Mota the people of the two moieties are said to have different characters ; the Takwong are reputed to be ignorant and unimportant^ always quarrelling and unable to manage their affairs properly, while the members of the Tatalai are well versed in social lore and live peaceably with one another, capable of governing both themselves and others. In the old time the members of the two moieties hated one another and even now there is a feeling of enmity between the two. It is said that in ancient times there was a very \or\g gamal or club-house (see Chap. Ill), one end of which was inhabited by the Takwong and the other end by the members of the other moiety. Sometimes a man changed his division by crossing from one end of the gamal to the other and in doing so changed his disposition, a man of the Takwong thus becoming a peaceable member of society. It was rarely that a man of the Tatalai crossed over to the side of the Takwong, for he ran the risk of being killed by those he was attempting to join. The site of this ancient building can still be seen and so far as I could judge from the spots said to have been the sites of its two ends, the building must have been at least a hundred yards in length*. There are still stones on the

' Mr Durrad was especially struck by the contrast with the Torres Islands in this respect. Even when names for the moieties were obtained in the Banks Islands they were not given with any confidence and only after long hesitation, whereas in the Torres group every child is at once able to give the names of the social groups to which he and his friends belong.

2 This should properly be Takpwong.

3 It may be noted that this information was obtained from a member of the other moiety.

^ The whole region is now covered with bush and the distance could not be measured.

Plate I

'iSgSi&:^Miiati£^i

"^'3^i4

The island of Mota.

Social Organisation 23

island which are said to have formed part of this old gamal, these stones being cut quite unlike any others on the island. It is tempting to suppose that when the two veve are spoken of as of the same or the other side of the house, we have a survival of this ancient habitation of a building in common by the two moieties, or perhaps rather of a time when the separation between the members of the two moieties was maintained within the house.

Each of the moieties has a number of divisions and these subdivisions^ have names. Those of the Takwong are called Talosara, Talogapmarus, Mareanusa, Nusawo, Tasalaima, Nuarara, and Punui ; those of the other moiety are Gatava, Pengusiu, Maligo, Talvatvat, Lolirwora and Lisaveve. These subdivisions are often called veve, the same word as is used for the moieties. Many of them take their names from rocks and in one or two cases certainly, and probably in all, these rocks are regarded as sacred and as inhabited by spirits (vui). Two divisions, the Tasalaima and the Maligo possess rocks called O pagoa (the shark) and at the rock of the Maligo a sacred owl inhabits a banyan tree. The Talogapmarus have as a sign or mark the talai or giant clam-shell, and the Lisaveve have a creeper sacred to their division which they must not cut or break. The members of some of these lesser divisions are said to have special characters, thus the Lolirwora quarrel and make friends quickly, while the Talvatvat of the same moiety are especially combative and are nicknamed the Pulgalegale (deceitful friend), otherwise the name of a spirit or vui at the village or district of Luwai (see p. 24). Both the Talvatvat and the Gatava are known as veve kwakwae or eccentric divisions. Their behaviour is different from that of the members of other groups, the following being instances of their eccentricity. It is usual to begin a meal with yam or breadfruit which may be followed by eating pig or the leaves of a kind of hibiscus called toape, the meal being finished with the pudding called lot, but the veve kwakwae eat in the reverse order. Further, when they have difficulty in drawing up a canoe, they will hack off the outrigger so that when they start again they have to make a new outrigger from unseasoned timber. The members of one division, the Tasalaima, are said to be unable to sing songs. The Talosara division of the Takwong

^ I am indebted for the following account of these subdivisions to Mr Durrad.

perform definite rites at their sacred rock or rocks (see

p. 157). . .

In all the islands with the dual organisation, the two moieties are definitely exogamous, a man never marrying a woman of his own veue. The restriction is not limited to the inhabitants of one island but extends to the other islands of the group, thus, a man of Saddle Island or Rowa going to Mota will at once learn which is the moiety corresponding to his own and who are those whom he should call sogoi, and if he marries he will take a woman who is not one of his sogoi. The same holds good for natives of the New Hebrides or at any rate of those islands with which the Banks islanders have regular dealings. It might have been expected that the carrying out of these regulations would have been much easier if the moieties had names, but the usual absence of names did not appear to raise any difficulty.

The divisions of the moieties in Mota do not influence the regulation of marriage, the men of any subdivision of one moiety being allowed to marry women of any subdivision of the other moiety.

In addition to the moieties and their subdivisions, the islands are divided into districts, often inhabited by people speaking obviously different dialects. Even in so small an island as Mota there are such districts each of which seems to be in many ways independent. I was unable to make any inquiry into these local divisions in Mota, the names of which have been given by Dr Codrington' on a plan of the island. It may be noted that several of the districts which he records bear the same names as certain of the subdivisions of the moieties, but it must be left for further investigation to determine the exact relation between the two modes of grouping.

In Merlav it was found that the social organisation was of a kind widely different from that of the other islands of the Banks group. I obtained the names and something of the histories of ten different groups called tagataga and my informant thought that there was at least one other. The names of these ten tagataga are Mweo, Sarana, Dungdung, Ronalung, Luwe, Bule, Ondoa, Kwerivia, Voslav, and Maketuk. These groups are not connected with any sacred objects such as animals or plants and have no restrictions

' M., 15.

Social Organisation 25

on food peculiar to themselves but they are exogamous and the origin of some of them is ascribed to certain women who came from other places. The ancestors of three of the tagataga, the Mweo, Sarana and Dungdung are reputed to have come from Maewo in the New Hebrides, for which the Merlav name is Mweo. Nothing could be told about the origin of the group which takes its name directly from this island, but the Sarana have a tradition that a woman of Sarana in Maewo was driven away from that place and floated to Merlav on a dish used for making large puddings. She landed at a place called Ngerungara and made a garden at a place called Visangut. While working there one day she heard a child crying whom she could not see, so she went quietly and found a girl sitting by some long grass. The woman looked after the child who was called Tangerungana when she grew up, and it is from this child that the Sarana people are descended. The Mweo and Sarana people believe in their close relation to one another and call each other tasi or brothers.

The Dungdung group is descended from a party of Maewo people who came to Merlav in a canoe. They lived for a time on the beach and one day some of them wandered about the island and found an old woman sitting alone in a house. Those who had found her went back to tell their companions and it was agreed that they should return and ask her for fire. The old woman gave it but did not speak. When the people returned to the beach with the fire a woman of the party asked what had happened, and when told she said that they were to put out the fire and go back to the old woman to ask for it again. The people did so and were given fire a second time in silence and were again told on their return to put it out. They went a third time to fetch the fire and on that occasion the old woman asked, " My children, why have you killed the fire?" They told their companions what she had said when they returned and it was agreed that the old woman must be their sogoi and they went to live at Dungdung which was the name of the place where the old woman had been found.

The Ronalung group is connected with Merig through a woman of that island who saw a bird give birth to a girl child in a nut tree (ngai). The woman looked after the girl and also had children of her own. One of these children

came to Merlav and went to the village of Vanmisi where she married. One of her sows went one day to a place called Ronalung and there gave birth to a girl who is the ancestress of the Ronalung division. Though the people are thus connected with Merig it is not from the woman of that island but from the pig that they are believed to be descended. The Luwe and Bule are descended from women of the island of Santa Maria but no stories connected with them were known.

The four other groups are called Ondoa, Kwerivia, Voslav and Maketuk. Their mode of origin is doubtful but it is believed that they are descended from certain people who for some unknown reason were appointed to start new groups.

The traditions connected with the origin of the different tagataga suggest that the island of Merlav has been populated by means of a number of small immigrations. It is the most outlying island of the Banks group and lies rather nearer to Maewo of the New Hebrides than to Santa Maria which is, except for the tiny island of Merig, the nearest of the other islands of the Banks group. The Merlav traditions suggest that the island was populated by small parties of people, each of which became the starting point of a social group.

Systems of Relationship.

The different islands of the Banks group differ from one another very considerably in their mode of naming relationships, these differences being not merely in the forms of the words but in the way in which they are employed. The leading features of the system have been already recorded by Dr Codrington\ the island of Mota having been taken as an example. My most complete information was also obtained from that island and agrees in all essentials with that given by Dr Codrington, though I give a few amplifications and modifications in detail. The accompanying pedigree will illustrate several points in this and succeeding sections. It may be noted that it gives the genealogy of the same family as was used by Dr Codrington i^M., 38). In neither case is the pedigree complete, only such persons being included as are needed for the purpose of illustration.

' M., 3S-

r^D —

^ 0+

I

<

s

-H

<

h]

rt

g

^

-0

as

z

^^

Q

W

(=< §

o

-H

o

o

s

T3

C

i

E

:) l-< iq

K o ^

a -o

o .c

Pi II

< -z

M

1°

II- Z

o

-4

!M

''5

^ 2

c4 >-■ O

(A S

i2

^•^

Mota.

The following are the Mota terms' :

Tamai. This is the term for the father, the father's brother and the husband of the mother's sister, these uses being those customary in the classificatory system. In addition the term is applied to the son of the father's sister, to a first cousin according to our method of nomenclature, i.e. John will give this name to Dudley.

Veve. Mother, mother's sister and the wife of the father's brother. The father's sister is also called veve but is distinguished as raveve, ra being an honorific prefix which is used for anyone who is highly thought of and is thus often applied to all who are closely related, in distinction from those more distantly connected. The father's sister whose full designation is veve vus rawe would always receive this prefix. The father's sister may be called vevegae to distinguish her from the mother and mother's sister who would be veve sogoi ; she would not be of the same veve as her nephew and therefore not one of his sogoi. Corresponding to the exceptional use of tamai, the daughter of the father's sister is called veve.

Natui. This is in general the reciprocal of the two preceding terms so that it is used, not only in the customary ways for a child and the child of the brother of a man or the sister of a woman, but also by both men and women for the children of the mother's brother, thus John and Agnes will be the natui of Dudley.

Tugui. The term for an elder brother of a man and probably also for the elder sister of a woman but it is seldom used, the term which properly denotes younger brother taking its place.

Tasiu. The term for the younger brother of a man and the younger sister of a woman and it is often used for brother or sister generally, not only between those of the same sex but sometimes even between those of the opposite sex.

Tutiiai. A reciprocal term used between brother and sister.

The three terms tugui, tasiu and tutuai are used not only for own brothers and sisters but for the children of the father's

' Most of the terms of this system are given without a possessive pronoun but with the usual termination of independent substantives.

I

brother and the mother's sister and in general for all those of the same moiety or veve who are of the corresponding generation. It is in the case of these more distant relationships that the term tasiu is especially used in its wider sense, and in these cases the terms which connote difference of age or sex seem to be rarely used.

Veve vus rawe. This is the term for the father's sister but it is also given to a number of other women who take part in a certain ceremony, of whom the father's sister is always the most important. The father's sister would be classified as a veve with the mother but is distinguished in the ways already mentioned.

Usur. This is a term used only for the husband of the father's sister who is also sometimes called maraui.

Maraui. The mother's brother, also applied sometimes to the husband of the father's sister.

Vanangoi. The term reciprocal to maraui, being given by a man to his sister's son.

Mateima. A term used by a man for the wife of his mother's brother, his wife's sister and the wife of his brother. It is also applied by a man to the wife of his sister's son.

Welag. A term used as the reciprocal of mateima, and thus applied by a woman to the son of her husband's sister, her sister's husband and her husband's brother. The terms mateima and welag are thus used both for relatives belonging to generations once removed and for those of the same generation.

Tupui (in address pupua). The terms used for the four grandparents and reciprocally for the four kinds of grandchild, their use being extended as customary in the classificatory system.

Rasoai. A reciprocal term for husband and wife. Another term also used is amin or amai.

Kwaliga. In general a reciprocal term for the parents of husband or wife and the consort of son or daughter, but in the case of some of these relationships, other terms are also used. The wife's mother is distinguished as rakwaliga, the same honorific prefix being used as in the case of the father's sister and other near relatives called veve.

Itata (in address tata). I only heard of this term as applied to the husband's father but I am not sure that it

Melanesian Society %

is not also used for the wife's father and possibly for other relatives\

Tawarig. This is probably a term denoting status rather than relationship, used especially for the son's wife by both man and woman. It seems to be used for all women who should help on certain occasions. It is used for the wife of a son's son by a woman, her brothers and their wives ; for the son's wife, for the brother's son's wife and for the wife of the mother's brother's son. Thus Jessie, the wife of John, was the tawarig of Leveveg and Rivlava and their wives, of Tapermaro, of Pantutun and his wife, of Maututun and Dudley (see pedigree).

Wulus. A reciprocal term for the wife's brother and for the sister's husband (m. s.). It is only used between men.

Walui. A reciprocal term applied to one another by two women, the husband's sister and the brother's wife (w. s.).

Gasala. This is a term applied to one another by the parents of a married couple, i.e., it is the name for the parents of a son's wife or of a daughter's husband, thus the father and mother of Jessie would be }^^ gasala of Pantutun and Emily.

The system so far recorded is that of Mota and there were very important variations in some of the other islands.

Saddle Island or Motlav'^.

Here the differences are chiefly in the terms, not in their application, thus, veve became vev ; veve vus rawe became vev wuhe rau ; kwaliga became kwiliga though in address the word with the first person possessive is kwalgak. I did not hear of any modes of application of these terms differing from those of Mota.

Vanua Lava.

There are definite variations in different districts of the island, certainly in the forms of the words and probably

' In the story of Ganviviris recorded by Dr Codrington (yl/., 384) this term is used in addressing an uncle (mother's brother).

^ Motlav is properly only the name of part of Saddle Island, another part being called Ra, but I do not know of any differences in custom between Ra and Motlav proper and in general in this book I shall use Motlav to denote the whole island. There is also a district of Saddle Island called Vlow (which in Mota becomes Valuvva or Valuga). The people of this district speak a special dialect and it is therefore possible that their system of relationship may differ from that of other parts of the island.

I

in their application. The only system obtained from this island which is at all complete is that of the district of Pek. Here the word for father is imam (emek^ in address), and this was also used for the father's sister's son and other relatives as in Mota.

The word for mother is imu (in address emu), also used for the same relatives, including the father's sister, as in Mota.

The elder brother of a man or sister of a woman is ogak (in address) and the younger isik, an interesting point here being the disappearance of the initial letter. The word for the brother-sister relationship is ehwek which departs more widely from the Mota form. The child is nenik. The four grandparents are popo, this word resembling the vocative form of Mota and Motlav.

The parents of husband and wife and reciprocally the consorts of son or daughter are kwelgek, and wulus and wuluk are used in the same sense as the wulus and zvalui of Mota. It is in the other terms for relatives by marriage that the most peculiar feature of this system is to be found. The wife is rengoma and this term is also applied to the wife's sister and the brother's wife. The husband is amanma and this term is also given to the husband's brother and the sister's husband. I failed to ascertain the term used for the wife of the mother's brother, for if the system is to fall into line with that of Mota she should also have been called rengoma.

Rowa^.

In this little island of the Banks group which is the chief place for the manufacture of the shell-money ^ the system of relationship has the same general features but there are distinct phonetic differences and one or two very instructive expressions to denote certain relatives. The father is imam, the mother ivev, and a child ching^neruk or werwer, and these terms are used in the same sense as in Mota except that it was said that the father's brother would be always addressed by name and would not be called imam. Brothers call one another chisik and the same word is used by sisters, but the

' With first person possessive, -^, suffixed. In this and the accounts which follow, most of the terms are given in this form.

2 It is probable that the language and system of this island are closely related to those of Ureparapara.

3 See Chap. VI.

32 Melanesian Society S

term used between brother and sister is chiochiok. The father's sister is called ivev and vev wuswus rawe. The mother's brother is maruk and the sister's son van-nguk. The wife of the mother's brother is called wunu munidal which means "the wife of all of us." The reciprocal term for grandparents and grandchildren is popo.

According to one informant the parents of husband or wife and reciprocally the consorts of son and daughter are kwilia, but according to another the wife's mother and the husband's father with their reciprocals are distinguished as kwalgak. The wife's brother is wulus and the husband's sister wuluk and these terms, as elsewhere, are used reciprocally. The wife's sister and the husband's brother, on the other hand, are addressed by name, this being reciprocated. The husband is called tamanrok and the wife ligenrok.

Merlav.

In this island, the social organisation of which departs widely from the general type of the Banks Islands, there are also decided differences in the nomenclature of relationship. The father is imama or in address mam and the mother correspondingly ivev or vev. The child is natuk, the elder brother tugak, the younger tasik and the brother-sister relation tatak. The father's sister is ivev and her husband moruk. The mother's brother is also moruk and his wife is ivev.

So far as I could learn there was a departure from the usual practice of the Banks Islands in that the name for the child of the father's sister is natuk, the child of the mother's brother also having this name, so that here two cross-cousins call one another by the term also used for a child. It was a point about which my informant, a very intelligent man, was quite positive and he gave a reason for it, saying that the child of the father's sister is called natuk because he is the child of a morui, i.e., it would appear that this method of nomenclature has come about through a process of generalisation. The mother's brother's son is called natuk and he is the son of a morui and it would seem that the same name has come to be applied to the son of the father's sister whose husband is also a 7norui, thus obscuring the nature of the reciprocal relationship between the children of brother and sister found elsewhere in the group.

I

The terms for the grandparent-grandchild relationship also differ from those of other islands. The four grandparents are all addressed as tumbuk and this term is used both by men and women for the children of a daughter. A woman also calls her son's son tumbuk but a man calls his son's son imbua. Similarly, a man calls his son's daughter tumbuk while his wife might call her either tumbuk or imbua. hnbua may also be used for the daughter's children but only in familiar speech, i.e., when there could be no chance of their taking offence. This means that except for this latter familiar use the term imbua is only used between two persons of the same sex who are grandparent and grandchild through a son.

The terms for the relationship with the parents of the husband or wife have also several exceptional features. The wife's father is called kwaleg and the wife's mother rombu, the reciprocal term to both being welag. The father and mother of the husband are similarly kwaleg and rombu but here the reciprocal term is taivarig. This is the only example in the Banks Islands so far as our information goes in which a distinction is made according to the sex of a parent-in-law and further we have the Mota term welag used in a new sense for the husband of a daughter.

The terms walus and waluk or nawal are used in the same sense as the similar terms of Mota but there is a most important variation in the terms for wife's sister and husband's brother. These are either addressed by name or as tatak, the term used for the brother-sister relation. Here relatives who in Vanua Lava call one another by the same names as husband and wife, address one another as if they were brother and sister.

Attention may here be called to a few of the special features of these systems though their full discussion must be left for the theoretical portion of the book. The characteristic feature of the systems of most of the islands is the use of terms ordinarily denoting the relationship of parent and child by persons of the same generation, viz., by crosscousins, the children of brother and sister. This peculiar feature was recorded by Dr Codrington who pointed out that it forms the starting-point of other anomalous features ; thus, the child of the daughter of the mother's brother may be regarded as a tupuiy i.e., a person one generation younger

than the speaker may be addressed in the same way as a grandchild. The reason for this is that the daughter of the mother's brother is a natui (child) and therefore her child will stand to the speaker in the same relation as a grandchild and be called tupui. The only island of the group where, so far as I know, this designation of cross-cousins as parent and child does not exist is Merlav but there can be little doubt that the nomenclature there is only a modification of the characteristic Banks feature which at one time existed in that island. It will be seen later that this special feature of Banks relationship is to be connected with a special marriage regulation of these islands, viz., one according to which a man marries the wife or widow of his mother's brother.

Another striking feature of the systems of this group is to be found in the names given in different islands to certain relatives by marriage. In Mota the wife's sister, the brother's wife (m. s.), and the wife of the mother's brother are all grouped together as mateima, the reciprocal to which is welag. We shall see later (p. 48) that each of the three relatives called mateima is a person whom a man may or should marry, so that mateima appears to be a term for a potential wife, while welag would be a corresponding term for a potential husband. In another island, Vanua Lava, the wife's sister and the brother's wife receive the same name as the actual wife while a woman calls her husband's brother and her sister's husband by the same word as her actual husband. In this case I did not obtain the name used for the wife of the mother's brother so that I do not know whether she also is classed with the wife as might be expected from analogy with the Mota practice. In another island, ' Merlav, the wife's sister and the brother's wife, if not addressed by name, are denoted by the same term as the actual sister, while in Rowa these relatives are always addressed by name. We thus find four different ways of designating relatives by marriage of the same generation but of different sex ; by name, as wives, as potential wives and as sisters. It may be noted here that the island of Vanua Lava, where sisters-inlaw are classed with the wife, probably stands lowest in the scale of culture of those islands from which material is available. It must be sufficient here to point out these special features ; their discussion must be left till later.

Functions of Relatives 35

A term for a relative by marriage about which I obtained much information, but which remains somewhat obscure is tawarig. This seemed to be applied primarily to the son's wife and to be used for the wives of many other descendants, direct or collateral. A woman is called tawarig by her husband's father's mother and her brothers but not by the husband's father's father or his brothers. A difference between these relatives is that the former belong to the same veve as the woman while the latter belong to the veve of her husband. Similarly, while a woman is called tawarig by her husband's father's mother's brother, she must not be so called by the sister's husband of the father's mother and here again the latter belongs to a veve different from that of the woman. The fact of belonging to the same or the opposite veve will not, however, account altogether for the distinction, for the husband's mother and the wives of the grandmother's brothers, though of the opposite veve, use the term but it is probable that they do this by right of their husband's relationship and that with this exception the use of the term is limited to those of the same veve as the woman.

The terms for the grandparent-grandchild relationship are interesting, partly because the systems of some islands make distinctions according to the sex of the speaker which are absent in others, partly because some islands use terms for all purposes which in other islands are only used in address. Thus, the term popo used in Vanua Lava and Rowa corresponds closely with the vocative (orm,/>upua, of Mota.

The Mota system is remarkable for the possession of a term, gasa/a, applied to each other by the parents of a married couple. So far as I could discover there is no special term for this relationship in other islands of the group.

T/ie Duties and Privileges of Relatives.

The Banks Islands are exceptionally rich in regulations concerning the behaviour of relatives to one another, and breaches of these regulations are looked on as matters of great importance. It will be most convenient first to consider the regulations connected with each bond of relationship and to consider later a general custom called poroporo, to which the islanders attach very great importance.

Parent and child. At the present time the behaviour

3—2

towards one another of parents and children has at least one feature which seems to us very unnatural ; a man and his father do not eat together, it being thought that by so doing the child might acquire his father's character. The unnatural character of this, however, becomes less pronounced when it is remembered that men always eat in the gamal or club-house and that it must be very unusual for father and son to be in the same division of this institution, so that their eating together would not be a regular proceeding even in the absence of this prohibition. Further, the fear that the son might acquire his father's character becomes much less strange in connection with the belief that the members of the two veve or moieties of the community have different characters, and that father and son belong to different veve^. It is evident that at the present time there is the usual natural affection between parents and children, and the ordinary observer would probably notice nothing in their relations with one another to differentiate them from the members of a European family, but it must be remembered that the people have now for a long time been under Christian influence which would have tended towards strengthening the bond between parent and child.

Brother and sister. There is little to be said about this relationship, the avoidance between these relatives found in other parts of Melanesia not being present in the Banks Islands. There is, however, one indication that the relation between brother and sister is not of that familiar kind customary in civilised society, and that is the fact that they do not poroporo or chaff one another. This practice is especially frequent between members of the same veve and, as we shall see later, the custom has much significance in the eyes of the people. It is therefore a fact of some importance that the prohibition should exist between brother and sister. It suggests a survival of the much greater restrictions on intercourse which are still to be found even so near the Banks as Lepers' Island in the northern New Hebrides.

Grandparents. The chief custom concerning the relation of grandparents and grandchildren of which I learnt came from the island of Rowa but it is possibly present throughout the group. If a man with children dies while his parents are still

' For examples which illustrate further the relation between father and child, 5ee the section on the mother's brother.

J

The Mother's Brother 37

alive, the children do not succeed to his land till they have bought off the rights of the grandparents with money and pigs.

The mother s brother. The relationship between a man and his mother's brother is very close throughout the Banks Islands. In the old days there is no doubt that the sister's son would have been the heir of his uncle and would have taken all his property, including any objects of magical value, but at the present time this has been much modified. There are even now complicated regulations enjoining certain payments from the children of a dead man to his sister's children, but at the present time when these payments have once been made, the sister's children have no further right to the property of their uncle. In addition to this potential right to the property of an uncle after his death, any of his goods may be taken by his sister's son during his lifetime, a right which is the source of many quarrels. It was said that even the most valuable possessions, such as pigs or canoes, might be taken in this way, but there is little doubt that this is a prescriptive right which is not put into practice. The difference in the status of the uncle and father is brought out very clearly in one way. If a man imposed on his sister's son a difficult task which resulted in his death or injury, no compensation could be demanded, but if a man put his own son to such a task compensation could be demanded by the mother's brother of the boy. If a man who is fighting is told to stop by his mother's brother he will do so at once, and it was said that if he had refused in the past he would have been killed by his uncle, but a man told by his father to stop would only do so if he felt inclined. A man treats his mother's brother with far greater respect than his father and may not put his hand above his head. If the uncle were sitting down in the house and something his sister's son wanted were hanging above his head, the nephew would not think of taking it but would wait till his uncle had gone. A man is usually initiated into the Sukwe^ by his mother's brother, and directly after the birth of a first child a ceremony takes place (see p. 146) in which the maternal uncles of the child take part. Before a child has been initiated into the Sukwe he may go into the gamal but only into that of his uncle, not into that of his father.

1 See Chap. III.

38 Melanesiaii Society

The sister's son always goes in his uncle's canoe ; it would not be necessary for a man to ask his nephew but the latter would take his place naturally, and if he refused to go he would be a despised man and if young he might be thrashed by his uncle.

Various objects which possess power connected with different kinds of magic are usually inherited even now by the sister's son and the latter may be called on to help in any ritual which his uncle is carrying out.

At the death of a man his sister's son can naro or taboo any food he chooses. The result of this is that the widow may neither eat this food nor cook it. Other relatives can thus na7'o food but it seems to be especially the privilege of the sister's son.

The foregoing account applies to Mota where I was able to make the fullest inquiries. It is probably true, however, for the whole group though the connection is perhaps less close than elsewhere in Merlav, where I was told that the relation between a man and his walus (wife's brother) is closer than that with the morui (mother's brother). In this island a man and his sister's son help one another and do not poroporo. A man will not take anything belonging to his mother's brother without permission, and if given anything it was said that something would be given in return. A man, however, must obey his mother's brother.

In Rowa it was said that it was the duty of a man to give food to the children of his mother's brother.

The father s sister. The relation between a woman and her brother's child is one of the utmost importance in the Banks Islands. She is called veve vus rawe in Mota and corresponding terms in other islands, this name being due to a ceremony in which she takes a leading part. When a man is initiated into the division of Avtagataga in the Sukwe^ a number of women assist, who by this ceremony come to stand in the relation of veve vus rawe to the initiate, the name being connected with a ceremony in which the initiate strikes {vus) the head of a tusked pig of the kind called rawe. In this ceremony the leading part is taken by the father's sister.

The father's sister may also be called 7naranaga, a name used for one of high rank and now adopted as the word for king or queen. She receives the greatest respect from her

1 See p. 79.

The Father's Sister 39

nephew who will not say her name, though he will take food from her and will eat with her from the same dish. It is a sign of the times that children now sometimes call their father's sister by name in order to annoy her, and I was told of a case where a woman was made to cry by her nephews and nieces treating her in this unceremonious fashion. If there is any rumour about a man or if it is known that anyone is desirous of injuring him, it is his father's sister who tells him of it, and she will always keep her ears open for any news which would concern her brother's child.

It is in connection with marriage that the role of the father's sister becomes of the most importance. She arranges the marriage of her nephew in the ordinary course, and if the latter chooses for himself, she may forbid the match. A man would never marry against the will of his father's sister. It seemed also that if an unmarried woman wished to have sexual relations with a man she would first approach his father's sister, it being usual in Melanesia for such a proposal to come from the woman.

A woman plays a great part in ceremonial following the birth of her brother's children (see Chap. VI).

There is a certain amount of community of goods between a man and his father's sister. The latter could take her nephew's things but only those which he had received from his father. If a man wants any of the possessions of his father's sister, it is customary first to ask her son but they may be taken without his permission.

This account applies to Mota. In Motlav a definite contrast was pointed out in the behaviour towards mother and father's sister. The mother may be spoken to strongly {maremare, i.e., emphatically and with assurance), but this is not permissible in addressing the father's sister. In Rowa it was said that the father's sister would not choose a wife for her nephew but would forbid a match of which she did not approve and in such a case she would always be obeyed. Here it was said that a woman would not speak "boldly" to her nephew, i.e., in the way which in Mota is called wuwang. On the other hand, however, a man will never speak to his father's sister till she has first spoken and this was definitely said to be a sign of respect. In Merlav the father's sister arranges her nephew's marriage or forbids one arranged by himself. It was especially said here that a man would not poroporo his

father's sister and that if he does so, she has to give a feast, all the expenses of which have to be paid by her nephew. This custom probably holds good of the other islands. In Vanua Lava the father's sister was said to find part of the money paid by her nephew for his wife.

A woman calls her brother's child natui vus rawe {nat wuhe rau in Motlav) and will never say his or her name.

The husband of the father s sister. The special feature about this relative, who is called usur, is that he is continually the subject of chaff from his wife's nephew. People will poroporo this relative continually, and will use special words in speaking to him which are called usur-gae. As examples I will take the special case (see pedigree) in which Virsal is the usur of my informant John. If John or his sister were to see a pig wallowing they would say, "There is Virsal." If they heard a flying fox in the night and met Virsal next morning, they would say, "We heard you last night." If they heard a kingfisher, they would say, "Your food is the body of Virsal," and anyone hearing this would know how they were related to Virsal. If they see Virsal going to the beach, they will say that he is going to eat manvat (a worm-like creature) or the sea-slug. If people are talking about Virsal and inquire where he is, John will say that he is in Panoi (Hades) or in some sacred place which he will name\ If a dance is about to take place, and John knows that Virsal is to be there, he will go too, rush up to his uncle and threaten him with a club ; he will seize him and will only relax his hold when he has been given money, which Virsal will have brought with him, knowing quite well what is about to happen. According to our informant the idea of all this is to emphasize the importance of the father's sister. Before Virsal married, John would have heaped all sorts of opprobrious epithets on him because he would not think him good enough to marry his father's sister, and John thought that the chaff and usur-gae are a continuation of this mode of conduct after the marriage, intended to magnify the importance of the father's sister. The 2isur is sometimes also called marazik, being thus classed with the mother's brother.

Husband and wife. A man may address his wife by her personal name though the wife should not take this liberty,

' These examples suggest all sorts of interesting ideas concerning flying foxes, kingfishers, etc., about which I could not learn anything definite.

Relatives by Marriage 41

but should speak of her husband as / gene if he has no children, or as the father of his child " Taman X'' if there have been children of the marriage. For a wife to address her husband by name shows a great want of respect, and it would seem as if at the present time this custom is used bywomen to flout their husbands. I was told that there are no less than three women in the district of Veverau alone who address their husbands by name, thus showing that they do not respect them. There is little doubt that this is the result of external influence, and that in former times the usual signs of respect would have been enforced.

Though a man may speak of his wife by name, he usually calls her either Ro gene ; the mother of his child, ''veve X'' \ or irananatuk, his children being iranatuk. Irananatuk was translated as meaning "my children," so that one of the ways of speaking of a wife is as if she were equivalent to her children. The explanation given by my informant was that the usage showed the superiority of the husband, the wife standing in a relation to her husband similar to that of his children, she being subject to his command while he is responsible for her behaviour. It was clear that a distinction is made between irananatuk and iranatuk^ and that it would not be right to apply the latter term to the wife. People often speak of a woman as amen followed by the name of the husband ''amen XT There are thus five different ways of referring to a wife : (i) by name, (ii) as Ro gene, (iii) as the mother of her children, (iv) as the children themselves, or (v) as the wife of her husband.

Parents- and children-in-law. In Mota the suffix k used in address is only applied to those nearly related to the husband or wife, as the actual father of the wife, one whom the wife would call father through a more distant relationship being addressed simply as kwaliga. The wife's mother may have the prefix ro and be addressed as ro kwaligak. In Motlav the wife's mother is addressed as ro kwaliga and spoken of as kwaligak. Neither a kwaliga nor an itata (husband's father) may be addressed by name, nor may their sons- or daughters-in-law use any word which forms part of their names, and in consequence of this regulation there are a number of substitution-words in each island for use when the proper name of an object is forbidden. These words are called un words in Mota and siem words in Rowa and, as

42 Melanesimi Society

Dr Codrington has pointed out', they are either old words or words from another island and therefore of great interest, both philologically and ethnologically. One way of avoiding the utterance of the forbidden names is to address or speak of a relative by marriage as the father or mother of his or her child.

In the case of the wife's father there are, in addition to this name-avoidance, other signs of respect. Thus, a man will not take anything from above the head of his father-in-law nor will he pass him when he is sitting; if he wishes to do either of these things he must wait till the father-in-law rises. He will vi€\t\&r poroporo this relative nor address him familiarly. With the wife's mother, on the other hand, the avoidance is far more thorough. A man must not enter a house if his wife's mother is within and near the door; he will wait till she goes or moves away from the door, and in the latter case he may enter but must sit as far away from her as possible. If he meets her in the bush he will turn off the path and make a wide detour through the bush in order to pass her. If on the other hand he has climbed up a tree, his mother-in-law must turn out of her way to pass him. She may not drink the water from any bamboo that he has carried and if she wants him to help with any work she will speak first to her daughter who in her turn will speak to her husband. Any infringements of these rules can only be condoned by payments of money.

One regulation of interest in connection with the prohibition on the name of a kwaliga is that if the name is uttered, the money which has to be paid by the offender is not kept by the injured recipient but is divided between his brothers and sisters.

The same rules apply to the relation between a woman and her husband's father, called itata in Mota and Motlav and kwaliga in most of the other islands. It is the part of the daughter-in-law who is called tawarig to help her itata and especially to get them firewood and to work in their gardens. She may not take anything from above the heads of her parents-in-law when they are sitting nor may she eat a bird which has Hown over the head of the itata and further the latter has only to say of an object ''Iloke ?ia kwatuk,'" "That is my head," and the object cannot be eaten by the woman.

' M., 44, note 2.

Relatives by Marriage 43

The itata in his turn may not say the name of his daughterin-law but must call her ratawarig, and here again if this rule is infringed by either party money has to be paid which is divided between the brothers and sisters of the recipient.

The chief idea which seems to attach to the word tawarig is that the woman so called must help those who give her this name. After the birth of a first-born child the father's sister holds her newborn nephew and makes a speech in which she mentions the tawarig of the child, thus referring to the future wife of the child as her tawarig (see p. 146).

In Merlav, where the terms are kwaleg and rombu, the same rules apply, and in addition it was said that these relatives must be spoken to softly and that food must not be taken from their heads or shoulders though food so carried by them might be eaten. It is probable that these rules also apply to other islands.

The wulus and walui or wuluk relationships. These are terms for relatives by marriage belonging to the same generation, the former being used between two men and the latter between two women. They do not say each other's names nor poroporo one another, and people object to hearing others poroporo their kwaliga or wulus. These relatives have regard to one another's heads but the avoidance of the use of their names when these correspond with those of natural objects is not important, i.e., un words are not necessary. (This is certainly so in Rowa and probably throughout the group.)

In Merlav the customs concerning a man and his wife's brother differ from those of Mota. These relatives may say each other's names and to a large extent they have their possessions in common. One will always help the other and if a man is in danger it is the duty of his walus to warn him and if necessary to die with him. If a man wants his walus to go in his canoe with him the latter would always go. When a man dies, his wife's brother gives money and pigs to the relatives of the dead man and helps to support his widow. He fasts from certain foods and a man who has lost his walus is carefully watched and if he is seen to eat everything he would fall greatly in the estimation of the community.

The reference to the head plays the same part in connection with the wului or walui relation as in the case of the relationships of kwaliga and itata and I am indebted to Mr Durrad for a very instructive example from Modav

which illustrates the great influence which this belief and custom may have in the lives of the people.

Georgina^ was making a mat and, wanting two stones to put on the corner to hold it down, she asked her husband, Katlavi, to get them but he refused. In order to make him obey, she said, "They are the head of your wului,'" meaning by the wului Charles who was sitting in the house. Then Katlavi went to fetch the stones but brought them in a wet and dirty state so that they were not fit to put on the mat. Either this or the fact that Georgina had made use of him to compel her husband to obey made Charles angry and he said, "If you go on making that mat, it is the head of Mabel," and she could then only finish the mat, after she had paid compensation to Mabel. It will be noticed here that the reference to his wului made Katlavi obey, though Charles was only a distant relative, a second cousin of his wife according to our nomenclature, while the relationship between Georgina and Mabel was equally distant. The example shows clearly that the various duties connected with these ties of relationship are not limited to the husband of a man's own sister or to the wife of a woman's own brother.

Pedigree II.

6 = 9 I

RUKWONG= ? WORAU = Petinge

r -J 1

M'^KENZiE = Petimdo BENJAMIN = Ellen

I Katlavi = Georgina

John Sem = Maria Walter Woser Mostyn = Eliza

I I

Charles = Mabel Nicholas = Eleanor

Another example of the reference to the head of a wului is that if a man beats his wife, she may say, "If you beat me again, I am the head of your wului,'' and this will make him stop. If he persists, the woman tells her brothers (her husband's wului) who will gather together and demand twenty fathoms of money and, if the husband refuses, they will cut down some of his trees. Again, if two men are quarrelling and the people wish to stop it, they will say, "If you go on, it is the head of your kwaliga or wului,'' and if the men disobey

• See the accompanying pedigree.

I

Relatives by Marriage 45

they will have to pay money or have their trees destroyed, this custom of referring to the head of relatives by marriage thus doing much to lead people to keep at peace with one another.

The wife^s sister and the brother s wife (m. s.). The very interesting gradation in the names applied to these relatives in the different islands of the Banks group has already been considered. The important feature of their relation to a man is that they are his possible wives and in Mota we find them sharing a name mateima in common with another potential wife, the wife of the maternal uncle, while the women have a reciprocal term, welag, for their potential husbands. In some of the other islands these persons address each other by name only and do not appear to have any definite terms of relationship to use when speaking of or to one another. In Vanua Lava, on the other hand, they call each other by the same names as a real husband and wife. There is a definite difference in the various islands of the group in the attitude towards these relatives as shown by \^ poroporo custom. In Rowa a man would never poroporo his wife's sister or his brother's wife but in Mota he might do so, though in the case of the former relative I was told that he would only do so to a small extent. He must poroporo in a respectful manner, this probably meaning that definitely sexual references must be avoided. It was said that if a woman carried the poroporo custom too far, she would have to be taken as wife by her sister's husband. In Vanua Lava I did not inquire into this point and this lapse makes it impossible to decide how far there is a connection between the use of special terms for these relatives and the poroporo custom.

I owe to Mr Durrad an example of the poroporo custom in which the relationship of brother's wife is involved. The relationship of the persons concerned is shown in Pedigree II. In this case Charles sees the wife of Nicholas going by and says, "Ah! Here is the wife of Nicholas and me coming hither." This makes all those present laugh, the wife ot Nicholas included. Walter Woser is sitting by and says to Charles, "You are worse than she is," and the people laugh at this retort. Charles says, "No. N icholas and I are all right ; it is she who is bad." Walter answers, "Wait a minute. You two sit here a little while," and he runs to get an old stinking leaf-mat which had been used as a covering for an oven. He

comes back with it and Charles's arms are seized by his maraui while Walter (also one of his maraui) jams the mat on his head and all the people laugh uproariously. The wife of Nicholas then mocks Charles in her turn and says, "I want some fish ; go and get me some and plant my garden for me." Charles's sister then joins in and says, "There is no food for you two" (Charles and Nicholas); but they answer, "O yes. We have plenty of food." This again makes the people laugh and Charles's sister then stops the joking by saying, ''Atogo !'' a word equivalent to our "Shut up ! " There is no doubt that this joking and horseplay which seem very pointless to us have a very definite meaning to those who indulge in it and it seemed to Mr Durrad that the gist of the whole joke lay in the original reference to the wife of Nicholas as also the wife of Charles. The point of the whole business lies in the recognition by all present of the possible relationship between Charles and the wife of one he would call brother.

The Poroporo Custom.

The custom of chaffing or mocking certain relatives seems to be so well established a custom of the Banks islanders as to be a regular social institution. There is a recognised difference in the way of regarding the custom according as those of the same or different moieties are concerned, the custom in the latter case being called poroporo tavale ima.

1 was told that a person should not poroporo one of the opposite veve and the relative in whose case the custom reaches its highest pitch, the husband of the father's sister, is of the same moiety as his tormentor. Even among those of the same veve there are, however, restrictions, for it is not right for brother and sister to poroporo one another. Although it was thus said that a man should on\y poroporo his own sogoi, it is clear that there is a definite variety of the custom, the poroporo tavale ima, in which those of opposite moieties are especially concerned. The case oi poroporo which was recorded in the last section is one in which people of both veve were concerned and I was given a definite instance in which a man had mocked the children of his mother's brother and of one who was his maraui in a more distant sense, and another case in which a woman had mocked her husband's maternal uncle. It is clear that there is an established custom of mocking certain relatives by

Marriage 4-7

marriage and the example which has been recorded suggests that the whole business has a deep-seated meaning to those who indulge in it. What this meaning may be, however, must be left for a later portion of the book.

Marriage.

The dual grouping which exists in most of the Banks Islands is the most obvious means of regulating marriage, but the simple rule that a man of one moiety must marry a woman of the other moiety is very far from exhausting the marriage regulations. There are other prohibitions or ordinances based on relationship; there are certain women who, though belonging to "the other side of the house," are yet ineligible as wives, while there are others whom it is more or less the duty of a man to marry. A man and his father's sister belong to different moieties and therefore the dual organisation alone would allow these relatives to marry, but at the present time it is clearly contrary to custom for this form of marriage to take place in most parts of the group.

It is probable, however, that the objection to this form of marriage is of relatively recent growth. In Merlav it was said that a man often married his father's sister in the old days, not merely if she were a widow, but as her first husband, and it seemed probable that this form of marriage still occurs occasionally. In Mota it was said that sometimes a man marries one whom he classes with the father's sister by the classificatory system though he would never marry the own sister of his own father.

Similarly, the dual mechanism alone would not prevent the marriage of persons belonging to generations twice removed from one another, such as the marriage of a man with the daughter's daughter of his brother or other women whom he would class with this relative according to the classificatory system, and it must be remembered that where this system is followed such relatives may be of approximately the same age and except for their relationship suitable mates. Such marriages did not seem to be definitely prohibited and were said not to be unknown, but they are regarded as abnormal and as a sign of mental weakness on the part of those who contract them.

The cross-cousin marriage appears to be unknown in most

of the islands and this is to be expected, since according to the usual Banks system of relationship cross-cousins stand to one another in a relation corresponding to that of parent and child. I obtained one case of this form of marriage in Merlav where, it may be noted, the usual Banks relationship has been modified (see p. 32) but in this case there seemed to have been especial conditions through which the marriage was brought about. Even in this case, however, the married couple were not children of own brother and sister though it was said that, even if they had been so, the marriage would still have taken place.

Not only are certain marriages between those of different moieties prohibited or regarded as unorthodox, but there are others which are especially enjoined, and one of these is so prominent that it may be regarded as the orthodox marriage of the Banks Islands. This is the marriage with the widow of the mother's brother. At the present time it is the established custom of the islands that a man shall marry the widow of his mother's brother if he is not already married, while in the old times when polygyny was still practised it is clear that he would take the wife or wives of his dead uncle even if he were already married. Further, there is the tradition that a man might hand over one or more of his wives to his sister's son during his lifetime, and there is an example of this in the story of Ganviviris recorded by Dr Codrington^ in which Ganviviris was given the third wife of his uncle.

When a man thus marries the wife of his mother's brother, he would come to stand in the position of father to his uncle's children and there can be no doubt that there is a close relation between this form of marriage and the peculiar feature of the Banks system of relationship whereby the children of the mother's brother are regarded as children and conversely the children of the father's sister as parents.

This marriage with the wife of the mother's brother may be regarded as an exceptional form of the Levirate, but this institution also exists in its more usual form. It is right for a man to marry the widow of his deceased brother and it seemed that the widow might be taken either by the elder or younger brother of the late husband. A third orthodox form of marriage is that with the wife's sister. At the present time

' M., 244 and 384.

Marriage 49

or till recently a man naturally married the sister of his deceased wife while still farther back it often happened that in case of polygyny the wives were sisters.

These three forms of marriage, with the wife of the mother's brother, with the brother's wife and with the wife's sister are further brought into relation with one another by a feature of the system of relationship found in Mota. In this island all these three potential wives are classed together as mateima and the use of this common term for the three persons seems to be clearly connected with their position as potential wives.

As already mentioned, marriages take place between the inhabitants of different islands of the Banks group and these are subject to the same restrictions as those between natives of the same island, a man being limited in his choice to the women of the moiety corresponding to that from which he would have taken a wife in his own island. It seemed that there were certain customs, if not positive regulations, concerning these marriages between different islands. Thus, it was said that Mota and Motlav intermarried in general, but while men of Vanua Lava and Ureparapara took wives from Motlav, the men of Motlav did not marry women of Vanua Lava and Ureparapara. I have, however, no exact genealogical data on this point.

I did not go fully into the mode of arrangement of marriage nor obtain an account of any ceremonial connected with its but this appears to be almost completely absent. A marriage is usually negotiated by a third party who arranges the amount which shall be paid by the bridegroom to the relatives of the bride. In one case of which I was told in Rowa the amount so paid was sixty fathoms of shell-money and two pigs. It appeared that such payment was not made when a man married the widow of his mother's brother, but on the other hand a man would not receive anything if his uncle's widow married elsewhere.

The most noteworthy feature of the arrangement of marriage in the Banks Islands is the power of choice or veto which rests with the man's father's sister, a power which was said to be absolute in some islands (see p. 39). It was said that occasionally a girl is married before she is old enough to live with her husband.

» See Codrington, M.^ 239.

When in the old days a man married a woman of his own veve the offending couple were killed in public by their relatives, usually by shooting with bows and arrows, less frequendy by clubbing to death. At the present time no definite punishment is inflicted but the offenders become the object of the scorn and derision of the whole island.

Adoption.

The practice of adopting the children of others is very frequent in the Banks Islands and is accompanied by many interesting features. Of these features the most important is that a man who fulfils certain conditions may take the child of another in spite of the unwillingness of the parents to part with their offspring. The true parents may be unable to keep their own child if others want it and it is interesting in this connection that the word for adoption, raiho, seems to have primarily the meaning of "snatch."

There are certain differences in the customs of different islands of the Banks group and I will begin with the island of Mota about which I obtained the fullest information. In this island a newly born infant becomes the child of the man who pays the chief helper or midwife at the birth. The sister of the father settles who shall be the midwife, so that the father usually has priority of information on this point and as he will usually be on the spot, he has thus two advantages in the contest for being the first to claim his child, but if he has not the necessary money or if he is away, it may happen, and frequently does happen, that another may step in before him and become the " father " of the child. Two definite instances may be given. My informant, John Pantutun, before his wife had given him a family, wished very much for offspring. He heard that the wife of a man named Matthew had just had a child and he knew that Matthew had no money, so he ran to his village, heard who had assisted at the birth, paid this woman the necessary money and became j the father of the child. Plenty of other people were said to
have wanted the child and John ascribed his success to his swift running which enabled him to reach the village earlier than anyone else. The other instance is one in which the same man lost his own child. John had left his island at a time when his wife was expecting her delivery. He knew

Adoption 51

that another man, Luke, was very anxious to adopt the child, so he gave the necessary money to his sister Agnes with instructions that she was to pay it to the midwife directly the child was born. When the time arrived Agnes did this, but soon after Luke arrived, paid her the sum, claiming that this was equivalent to having paid the woman himself, and on John's return to the island he was not able to upset the claim of his rival to the fatherhood of the child.

The relations between the real and adoptive parents do not cease with the payment which has been mentioned. When the real father takes his next step in the Sukwe he receives a further payment which is known as "the money of the hand of the child." Further, the transaction is not absolutely final till a further payment has been made which is called vat ngorag sasai, or "name concealment." The fact of the change of parentage is very carefully kept from the child and this concealment becomes imperative on the part of the father after this last payment has been made. The money is given in the presence of the child who is, however, led to think that it is only being paid in connection with some Sukwe function.

It seemed clear that the real father could recover his child from the adopting parents by making certain payments, but the conditions were such that this is rarely possible. The father would have to repay all money which had been expended by the adopting father on behalf of the child with cent per cent interest. If the real parents wish to recover their child when only the payment to the midwife has been made and food supplied to mother and child, this may be possible, but after any of the later payments have been made by the adopting father, and especially after the boy has been initiated into the Sukwe, it becomes as a rule quite impossible for the father to return double the sum which has been expended. Even if the real parents wish to recover their child when only the initial payments have been made, it is probable that they will not have the money by them and if the adopting parents hear of the intention, they have only to make one of the further payments or to have the child initiated into the Sukwe to defeat the purpose. I was told that the adopting father usually makes the payments large in order to remove all possibility of their repayment by the real father. I n the case of initiation into the Sukwe, it may be mentioned that the

4—2

father would not only have to repay the sum expended by the adopting father but also the contributions made by relatives of the child in its new relationship and any payments made on behalf of the child during the initiation of others. Even supposing that all the money could be repaid with cent per cent interest, further difficulties would be raised. If land had been planted for the child, this would have to be redeemed or compensation of some kind given and a return would have to be made for all the food consumed by the child or at feasts connected with the child.

It was said that the adopting father rarely if ever relaxes his hold on the child if he can possibly help it. It may happen sometimes that a man sees his own child growing up in the possession of many good qualities which make him regret that he is known as the son of another and the man may ingratiate himself with the boy and with the adoptive father in the hope of obtaining his child back again. Such an attempt will only make the adoptive father angry and may lead to quarrels or even bloodshed between the two. In some cases it appears that the adoptive father will apparently consent, probably in cases where he fears the magical power of the real father, but the yielding will only be feigned, while he sets in action a process to take the life of his adopted son
He will kill his supposed child rather than allow him to return to the real father.

As I have already indicated, the fact of his real parentage is carefully kept from the adopted child and any approach of the real father to his child is resented on account of the fear that the true parentage may be revealed. The ceremony of vat ngorag sasai should properly take place when the child has grown up but it is sometimes performed when it is quite young in order to remove all possibility of such revelation, for the true father regards his lips as absolutely sealed when this payment has been made. If, on the other hand, the child learns its true parentage while young, the payment called vat ngorag sasai is not made.

Sometimes the real parentage of a child is revealed by some third person, usually in the course of a quarrel, and I was told that no Mota man is ever wholly free from doubt as to his real parentage.

Two kinds of adoption are recognised in Mota. In one,

• For the method of doing this see p. 158.

Adoption 53

which is regarded as the more normal form, a man adopts a child of a moiety different from his own, i.e., of the same moiety as his wife. A man's own children belong to the opposite moiety, so that in this method of adoption the adopted and real children of the man belong to the same social division and the child has not to change his division on adoption. In the other form a man adopts a child of his own division and consequently the child has to change its veve. An adopted man is thus compelled to marry a woman of the same division as that in which he had been born and the possible awkwardness of this makes this second form of adoption very unpopular. In the first form there can be no danger of a consanguineous marriage. In the second there may be such danger and if a marriage of this kind is projected the relatives of the girl intervene and forbid the match and this arouses the suspicions of the adopted child and leads him to make inquiries. The unpopularity of this second form of adoption seems thus to be partly due to the greater difficulty of concealment. If a man discovers that he has been adopted, he may sometimes marry a woman of his own veue, i.e. a woman of his veve by adoption, and it was said that in such a case he would not be subjected to the ridicule incurred in ordinary cases of this kind (see p. 50), for people will know that he had been born into a veve different from that of his wife.

Although the man who makes the necessary payment to the midwife is regarded as the father of the child, he does not at once take possession of his new acquisition. The true mother suckles the child and only gives it up when it is weaned. During the whole time of suckling, however, the food for the child and mother is provided by the adopting father and if there is ever any question of resumption of the child by the real parents, the value of this food has to be repaid. Further, the adopting father has to supply the feast which is customary when the weaning takes place.

The time at which the child is weaned is decided by the adopting father who generally tries to hurry on the event. We were given graphic accounts of the grief often experienced by the parents on losing their child, and we heard of one case in which the parents were so miserable that they could not sleep and got up in the middle of the night to recover their child. The father went to the house to which his child had been taken, found the new mother sound asleep and carried

off the child. The loss was first discovered by the adoptive father who woke up and, finding the child had gone, went to recover it from the parents who got into great trouble with the community for their breach of Mota custom. It may be noted that it was the adoptive father, and not the mother, who was awakened by the theft.

In the island of Motlav the practice of adoption differs in several respects. Here a man who wishes to adopt a newly born child or one about to be born sends his wife to the mother to give food. Some food is chewed and given to the child by the woman who wishes to adopt it and this giving of food seals the transaction. Another important difference in Motlav custom is that the child is taken by the adopting parents at once. If the new mother has had a child recently she will suckle the adopted child herself; if not, she will give it to another woman. In no case is the real mother allowed to suckle her child.

The only other Banks island from which I obtained an account was Merlav. Here a man who wishes to adopt a child about to be born waits till he hears of the birth and then runs quickly to plant the leaf of a cycas tree in front of the door. The man who first plants a leaf is the father of the child. The leaf remains in front of the house for ten days and then the adopting father pays the necessary money to the woman who has acted as midwife as in Mota. In Merlav the child is not taken by the adopting parents till it begins to cut its teeth when money is paid to the father of the child.

In addition to the adoption at birth a child may be taken by another later in life, and one occasion for this is the death of a woman with a young child when the child may be adopted by another. In a recent case the adopted boy has turned out a fine youth who works well and the real father is anxious to recover him for his own. The father has remarried and he and his wife are always trying to entice the boy to come back to them. Recently the sister of the father has been accused of telling the boy his real parentage and the people concerned are continually at loggerheads about the matter. In this case the following difficulties stand in the way of recovering the child. The adopting father has been responsible for the initiation of the boy into the Stikive and when some of the boy's sogoi were initiated the adopting father gave money to the boy to contribute {tnaiegae) for them. Further,

I

Property and Inheritance 55

the boy has had trees planted for him and there is the supply of food to the boy which has now been going on for many years.

Several motives were given for the practice of adoption. Childless persons have no honour and may be the object of ridicule because they have no one to call or command. A child is spoken of as tano loglog, which means a "place of calling " and the way the matter was put was that a child is adopted because a tano loglog is desired. Childless people are generally extremely kind and generous because they desire the affection of the children of the village. If they are kind and generous the children of others will obey them, obtain fish and other food for them and help them in various ways. It was said also that adoption is designed to prevent property passing out of the family; the more children there are in the family the less is the risk of this. It is a question whether these motives furnish a sufficient and satisfactory explanation and the matter will be more fully considered later.

Property and Inheritance.

The mode of inheritance of land in the Banks Islands has been briefly described by Dr Codrington^ who brings out clearly the distinction between ancient hereditary cultivated ground which goes to the sister's son and land recently reclaimed from the bush which is inherited by the children of the man who has cleared it. He also points out that the property in trees, which may be distinct from that in the land on which they stand, also goes to the children of the planter.

There are almost certainly in this as in other features of the Banks culture different regulations in the different islands and I began an inquiry into the question of ownership in Mota by taking concrete examples of what had happened in the cases of certain plots of land, but I was unfortunately only able to work out one such example and that not completely. Since, however, that single instance brought out several features of interest, I propose to record it here in spite of its imperfections, especially because it seems to me to furnish an excellent example of the genealogical method. The land in question was cleared by Takwale (see pedigree I on p. 27) and his

1 M., 63.

Melanesian Society

brother, who reclaimed it from the bush and divided it into two equal portions, one for each brother. I was only able to follow out the fate of the portion which was taken by Takwale, and my account will deal only with this. As the children of Takwale were born the land was portioned out, a plot being planted for each. This in accordance with the usual custom was done by the father who, as he planted the plot he proposed to assign to the child, carried the latter on his left shoulder. If it is wished to plant land for a son who is grown up, the father will plant with his son walking behind him and putting his right hand on his father's left shoulder. There is also a custom of planting land for an unborn child, in which case the father plants with a dried coconut under his left arm or fixed on his left shoulder to represent the child, a coconut being the customary representative of an absent person. These customs are still followed and my informant had recently planted a piece of land with his sister's children on his shoulder.

The plot of Leveveg, the eldest son (see pedigree) was

planted when he was young so that he was carried on the shoulder of Takwale and his portion was in one corner as shown in Fig. i. When the next child was born the adjoining plot was planted for him ; his portion was large, larger than that of any of the rest and therefore he was called Rivlava {I'iv, plant and lava, large). The rest of this strip was planted for the next child, a girl named Tapermaro. The next strip was planted in three plots for three children whose names were not obtained. The first part of a third strip was planted for the youngest son, Matthias, who is still alive and the rest was left common to the whole family. So long as Takwale was alive, this unallotted portion was regarded as especially his, while the allotted portions were regarded as property common to father and child. Thus, if anyone pointed out the portion of Leveveg and asked to whom it belonged,

Leveveo

Rivlava

Taper- -maro

Matthias

Common Land

Portion of brother of Takwale

Fig. I.

Plan of land cleared by Takwale and his brother.

I

Inheritance of Land 57

the answer would be, "It belongs to Leveveg with his father," while if the question was about the unallotted portion they would say, "It is the property of our father," and when the further question, "In common with whom?" received the answer, "In common with no one," the status of the land would be clear to the questioner. When Takwale died, this unallotted portion was common to the whole group of children and any of the children might plant there. The children of Takwale's brother might also plant there and if they planted coconuts, these would be always recognised as theirs though the land would always remain the property of the descendants of Takwale. When any of the latter died, the owner of a tree on the land would pay a sum of money to the senior representative of the survivors of Takwale's family. It might perhaps have been expected that the existence of this common land with the rights of others than its owners to plant and take produce would have been a continual source of disputes, but it was said that quarrels in connection with such common land are unknown although disputes about land which has been assigned to individuals by their fathers are very frequent.

The plot of land given to Leveveg was portioned out by him among his children who were carried on his shoulder while their plots were planted as in the original partition. All the descendants of Leveveg died before Leveveg himself and at his death his land was taken by Rivlava. He adopted Matthew, the son of Tanenkwat, who at the death of Rivlava took his adoptive father's land, i.e. the plots of both Leveveg and Rivlava with the exception of a piece of the garden taken by Matthias because Matthew did not pay him a sum of money which should have been given to him. If Rivlava at his death had left several brothers Matthew should have paid money to each but in this case Matthias was the only surviving brother. If Matthew had been the real son of Rivlava there would have been the same claim for money, i.e. the payment was not the result of Matthew being an adopted son. If all the brothers had died before Rivlava, Matthew should have paid money to Robert Pantutun, the son of Rivlava's sister. John and Agnes and the children of Agnes may all take coconuts or breadfruit from the land which Matthew has thus inherited from his adoptive father and if they wanted part of it for a yam-garden, they could have it. John's son could also take

coconuts from the garden but his wife would not be allowed to do so.

The third plot planted for Tapermaro was again portioned out among her four children of whom all but Robert died before her and he took the whole patch at her death. This, however, was only because Dudley, the son of the daughter of Tapermaro, was away from the island at the time and when he returns to Mota, he will take three-quarters of the plot, Robert keeping the remainder. The fourth plot now belongs to Mary, the granddaughter of Takwale, but as she is very old and her daughter Bessie far away, the plot is used by Robert. Most of the fifth plot now belongs to Patteson, the greatgrandson of Takwale, but part of it is owned by Robert, this land having been taken as a return for certain money given to Patteson to enable him to pay money to the surviving relatives of his mother when the latter died, the claim of the relatives being of the same kind as that already considered in the case of the succession of Matthew. The last patch is still in the possession of Matthias and when he dies his daughter Alice will pay money to Robert if the latter survives his uncle. If Alice were unable to pay the money it might be paid for her by John. This would be because John is regarded as the brother of Alice, being really her father's sister's grandson, i.e. of a different generation. This is a good example of the characteristic feature of the Banks system of relationship. Robert is regarded as the father of his cousin Alice and in consequence Robert's son becomes the "brother" of Alice, and as a consequence of this relationship we see that a man will pay to his own father money which is due from one we should regard as a distant relative on account of a relationship of a peculiar kind set up between his father and the relative in question. We have here a striking example showing that even these anomalous terms of relationship are not merely titles but carry with them very definite duties and responsibilities.

The common patch left after the children of Takwale had received their portions is still common and is used by all the living descendants of Takwale.

It has been already mentioned that the ownership of trees is distinct from that of the land on which they stand. This separate ownership may also arise if a man plants a tree on the land of a man to whom he is unrelated. In such a case

Inheritance 59

the planter uses the produce of the tree but he can only transmit his rights in the tree to his son if the latter pays money to the owner of the land when his father dies. If he does not do so, the tree becomes the property of the owner of the land on which it stands. Sometimes a man pays this money during his lifetime to ensure the right of his son, but in such a case it may happen that the owner of the land dies first and the land goes to his brother who might then demand a second payment when the owner of the tree died and his right would be recognised. An example may be given of a dispute now in progress about a ngai or nut tree on the land cleared by the brother of Takwale. The tree was planted before this land was cleared by an ancestor of Takwale who ate of it. Tapermaro and Robert have also eaten of this tree and paid money to descendants of the brother of Takwale when they succeeded to their rights in the tree. Rorolea, the grandson of Takwale's brother, now claims possession of the tree and demands money if Robert and John are to continue to be regarded as its owners. If Robert died the right of Rorolea to receive payment would be recognised but it is regarded as unnatural that he should claim payment now. The dispute is still in progress and John said that he expected that his son and the son of Rorolea would still be quarrelling about the tree in the future. When there is no relationship a man will only allow the planting of one tree on his land, so that in this case the planting evidently requires the permission of the owner of the land.

A piece of land may be let by its owner to another person and money given by the latter in return. This happens especially when a man has a piece of land at a distance from his home so that it is inconvenient to cultivate it himself.

In the case of property other than land, it seems to be the custom at the present time for the children to inherit, but there are the clearest indications of an earlier succession of the sister's son. The latter may still take anything he chooses and if this were not conceded, it was said that the sister's son would have the right to take everything.

There is a ceremony after death called rave epa ("draw mat ") which is evidently connected with rules of inheritance. The dead body is covered with strings of money which are drawn away one by one by a child of the deceased, who says with one, "This is for X's garden," X being the name of the

6o Melanesian Society

dead man; with another he says, "This is for X's pigs," and so on for all his possessions. The money is given to the relatives of the dead man, but I did not obtain its exact destination nor did I ascertain whether this ceremony would be performed by the sister's son in case he inherited.

The gamal or club-house of a village usually or often has an individual owner and this is certainly so when the man by whom a village has been founded is known, the gamal then belonging to one of the descendants. In this case the inheritance is to brother or sister's son rather than to child and in the only case of which I have a record the gamal was inherited by a brother. A gamal belonging to Rivlava was inherited at his death by Matthias, his only surviving brother, and not by his adopted son, Matthew, and at the death of Matthias it will go to Robert, the son of the sister of Rivlava and Matthias.

CHAPTER III

BANKS ISLANDS

The Sukwe.

The Banks Islands are the seat of a very complicated organisation which is in general known as the Sukwe, though this term is usually employed in a more limited sense. The organisation has two main divisions; one, closely connected with the village, the Sukwe in the more limited sense or the *' Sukwe of the village " ; the other, the Taniate or ghost societies which meet in the bush and are sometimes known as "the Sukwe of the bush."

Dr Codrington has given an account^ of these two branches of the organisation considering them as two separate institutions, the club and the secret society. The description which follows must be taken as an expansion of his account in which certain features are treated in greater detail. The most important modification is that a definite connection will be shown to exist between the two organisations which were supposed by Dr Codrington to be separate institutions.

Before considering these organisations in detail, a short general account of them may be given.

In every village of the Banks Islands there is a building called the gamal, which is used (except on very special occasions) by men only, and it is in this building only that men sleep and take their food. In the last chapter I have spoken of it as a club-house, the Sukwe being regarded as a man's club. Within, each gamal is divided into a number of compartments, often only separated by logs on the ground, and each of these compartments is connected with a division or rank of the organisation of the Sukwe. These divisions vary in number, chiefly owing to the fact that in many villages there

1 M.I Chaps. V and vi.

62 Afelanesian Society

is no one who has attained the higher ranks of the organisation. Each compartment of the gamal possesses an oven and it is at the fires of these ovens that the food of the men is prepared on important occasions.

If you leave the village and take a path into the bush you will sooner or later come to a place where two ways diverge and on the path which bears evidence of being less used there will be some mark or decoration, — a cycas or hibiscus tree or a decorated stake, — which will indicate that it is prohibited to all except initiated persons. If you are allowed to take this path it will lead you presently either to a building or to a clearing in the bush which is called a salagoro, the lodge of one or more of the Tamaie societies.

Lastly, there will be found decorated stone-platforms called wona on which perhaps there may be the decaying remains of masks or other objects. The decorations of such a platform may be the relics of a ceremony called a kolekole of which the most obvious feature is a dance, and these kolekole ceremonies are closely connected in many ways with the organisations of the Sukwe and Tamate societies, though they are largely public affairs in which both initiated and uninitiated take part.

I propose to describe in order the organisation: first, of the Sukwe ; secondly, of the Tamate societies ; and then to consider the part which this organisation takes in the general life of the people, including a description of the kolekole.

The Sukwe of the Village.

The divisions or ranks of the Sukwe vary in number in different islands. In Mota the following are the names of those at present in existence: — Avrig, Kwatagiav, Avtagataga, Luwaiav, Tamatsiria, Tavatsukwe, Tavatsukwelava, Kerepue, Mwele, Tetug, Lano, Kworokworolava and Wometeloa. The names of these divisions are said to be derived from — they certainly correspond with the names of — certain images which may either be carried in the hand or worn as hats or masks, these images being called in general tamate.

Dr Codrington gives Rurzvon as the lowest rank, but this is said to have long disappeared everywhere. He also gives four ranks above Wometeloa, viz. Welgan, Wesukut, IVetauro-meligo and Tikivangivono, but these seem also to have disappeared, at any rate in Mota.

Plate II

Plate III

The Sukwe 63

The size of a ganial in a village depends largely on the number of ranks of which there are members in that village and sometimes there may be seen a gamal which has been allowed to fall into ruin at one end owing to its higher ranks having no members. In a village in which all or most of the ranks are represented the gamal will be a long, usually low building with entrances at intervals along the sides, as well as at each end. Such a building is shown in PI. II, representing a ^<a!;;^?«/ from Ra in Saddle Island. PI. Ill, Fig. i, shows the front of an ancient gamal in Santa Maria in which the figures are of solid stonework. Mr Palmer, to whom I am indebted for this photograph, believes that this building was only used by those of high rank in the Sukwe.

The gamal is the eating and sleeping house of its members as well as their general meeting place in the village. Each man sleeps and eats either in the compartment of his own rank or in one belonging to a rank lower than his own.

At the present time the strict limitation of the use of each compartment of the gamal to the members of its proper rank is becoming lax and in Mota uninitiated boys have even been known to run right through the gamal from one end to the other, but formerly a man who entered the compartment of a rank even one degree higher than his own would have been trampled to death.

The first initiation may take place at any age and it is not uncommonly carried out during infancy, most of the essential ceremonies in this case being performed by the man who acts as introducer. Otherwise initiation takes place on entrance upon manhood or even later if for any reason, such as poverty, a man has been unable to enter before. If a man cannot enter the Sukwe he has to feed with women and this may sometimes so excite the pity of a friend that he may undertake to act as introducer, knowing that he will thereby have to spend a large sum of money. Even in cases where the initiate and his family are rich, it was said to be usual for the introducer to be the loser by his good offices. It is not necessary to pass through all the ranks of the Sukwe, beginning at the lowest, but, provided certain conditions are observed, initiation may take place into any rank directly. These conditions are, hovi^ever, so difficult of fulfilment and the expense would be so great that it is customary for the initiate to enter low down, though the lowest or the two lowest ranks are probably so often passed

over that in many villages these ranks are not represented in the gantal.

There is one most important limitation on entrance into the higher ranks. No one can pass beyond the rank of Ta7natsiria without having first been initiated into the Tamate society called Tamate Izwoa, this being one of the most definite indications of the connection between the Sukwe of the gamal and these societies.

The Ceremony of Initiation.

In passing on to describe the process of initiation, I propose to take a real case as the basis of my account, viz., that of my informant John Pantutun introducing his sister's son, Mark\ to the rank of Kwatagiav. One day John would say to Mark, "I will now make you stikwe," the last word being used as if it were a verb. This would make Mark sad, for he would know that he had not enough money for the purpose and he might even refuse to accept his uncle's proposal.

If he consented, John would announce that Mark would '^wusulie about Kwatagiav^ Then Mark would bring a pig and tie it to a stake near the door of the gamal. Some man would blow a conch-shell five times, three long continuous blasts and two interrupted blasts, upon which Mark would smack {wusulie) the pig on the back, a man standing by saying, "Let Mark now smack the pig for Kwatagiav.'' The pig then becomes the property of John. If Mark has plenty of money the initiation may proceed at once, but more usually there is now a long interval to allow the initiate to collect the necessary sum.

For initiation into Kwatagiav it was said that the sum which the candidate himself should produce is 60 fathoms, but the actual amount expended much exceeds this. In order to obtain the necessary amount Mark would put into action a special procedure. If he already possessed ten fathoms he would give this to ten of his friends, a fathom to each. After some months he would go to them and ask for his money back and each man would give two fathoms, thus returning what had been given with cent per cent interest. Mark would then distribute the twenty fathoms thus received

' It will be noticed that Mark is not included in pedigree I, and I do not know his exact relationship to John. It is possible that he was only a "sister's son" in the classificatory sense.

Initiation 65

and at the end of a few months he would demand these back and again receive interest at the rate of cent per cent, raising his wealth to forty fathoms. If this amount were sufficient the repayment of the money would form part of the ceremonial of initiation. In the meantime he will have been working, perhaps helping others to build a house and may thus have earned the remaining twenty fathoms. If Mark is able to tell John that he has enough money to proceed, a day will be fixed for the initiation. If Saturday be appointed, the preceding Wednesday would be devoted to collecting firewood and this would be a sign to all that an initiation is about to take place. On the next day those who had assisted in the collection would be paid and the real proceedings would begin on the Friday. On this day all would gather nuts, and food would be collected with leaves {gae) in which to wrap the puddings and in the afternoon the nuts would be cracked, fires would be lighted in the ovens, those of the women in the houses and those of the men in the gamal, and all would be busy preparing and cooking food. While the people are thus employed the men of the Kwatagiav division will be making a pudding and when it is ready all the women will be sent away from the village. One of the Kwatagiav will then fetch a cycas {iiieie) tree, drag it into the gamal and place it just inside the door, no women being near while this is done. Later in the night when the fires are no longer blazing, the men will make a noise with the instrument called iiteretang, consisting of a leaf acting as a reed between the two parts of a bent piece of wood. A piece of wood so used is illustrated in Fig. 2, the two ends being tied together with a piece of grass. All those initiated in the ikeretang^ go to a place by themselves in one part of the gamal in order to make the noise and at the same time others will blow conch-shells and make a whistling sound with dracaena leaves. After a time this noise ceases and there is perfect silence; no one will move about except those belonging to Kwatagiav who will sit down by the cycas tree at the door of , ^.•' the gamal. Three or four men sit on one side of the tang. tree and on the other sits Mark, who will place on the tree as many half fathoms as there are men. He then

• See p. 112, R. M. s. S

1'

takes up the end of one piece of money and John, his introducer, says to the first man, "X, Mark eats Kwatagiav to you," and gives to him his half-fathom, and this is repeated with the other men, each man receiving his piece of money. The special pudding is now brought which so far Mark has not touched; pieces are cut off corresponding in number to the people sitting by the tree and each man holds his piece before his mouth while John says words of which no full translation could be given. They run something like **Neg^, neg, neg, neg,...vawo, wo waso, ti, of of of of...'' These words are chanted, the chant rising in pitch as far as the word ti and then descending and at the last of each man puts the piece of pudding in his mouth and eats. The tree is then planted outside the gamal door and strangers seeing it will ask what it means and will be told "It is the sign of Kwatagiav' and all will know that this division has received a new member. As soon as the tree is planted the women may return to the village. Only those belonging to the Kwatagiav division eat at the mele tree and it is this eating which makes the initiate a member of that division of the Sukwe.

On the following day the food which has been prepared is distributed to the men in the gamal and to the women in the houses, but this is not done till after a distribution of money has taken place at the gamal. Mark goes to the house to fetch his own money while those to whom he has lent money bring what they owe him with its increase. Mark's relatives (sogoi, i.e. those of the same veve) also bring him money while friends may also contribute. It was said that one man might give as much as forty fathoms. Mark sits down outside the gamal with a basket before him into which he puts all the money he receives, this money being called mategae. If Mark were married, his wife would be the last person to contribute; if he is not married, his mother would take her place. The relatives of the wife would give her at least forty fathoms, the first four fathoms would be stretched on the ground and the rest wound round it and then Mark would take up the money and bring it into the gamal. Then would begin the distribution to the members. Mark would put two sticks in the ground six fathoms apart and would wind ten coils round these, thus making 1 20 fathoms. While the money is lying on the ground

Initiation 67

a man important in the Sukwe stands within the gamal and blows a conch-shell as before, three continuous and two interrupted blasts, and at this signal Mark takes up the end of the money and another important member of the Sukwe says to John, "John, let Mark now give you your money." Mark puts the end down and John takes the money which is now his. Each person present in the gamal then gives John half a fathom called som veni. To each half fathom John adds a full fathom and each man takes back {sar) the half fathom he had given together with the added fathom, the result being that John pays as many fathoms as there are members present. As soon as this distribution of money has taken place, the ovens are opened and a general distribution of food follows. In every village there is a big oven and the food for the general population is cooked in this but the food for the men who are members of the Sukwe is cooked in the ovens of the gamal. In the case of an initiation into one division the ovens of the divisions below this are not used. It was found very difficult to understand the exact nature of the financial transactions. It was said that the introducer usually lost money, but from the foregoing account it would seem that he has only to pay as many fathoms as there are people present. On the other hand he was said to have received in this special case 1 20 fathoms, which would imply that there must be more than this number of persons present, if he is to lose\ In Mota the 120 fathoms are called 60 rova, although in other islands of the group the amount would be called 120 rova. The money need not be paid in the gamal but this may be done in the house. This means that it is not regarded as strict ceremonial and women may be present. It is the rite of the previous night which forms the real ceremonial occasion.

The ceremony of initiation is of the same kind for all ranks in the Sukwe ; in all cases there is the eating by the ntele tree and the distribution of money. The difference is in the amount of money and the details of its distribution. For the next rank above Kwatagiav, viz. Avtagataga, it was said that one and a half rova are added for every half rova given. For Luwaiav every man present puts down one fathom and receives back this together with an equal amount.

• It is probable that there are many other expenses and that the above account is far from exhausting the financial transactions connected with initiation.

5—2

For Tamatsiria each man puts down two fathoms and has another two fathoms added to it.

When Tavatsukwe is reached there is a difference. So far there have only been concerned the men belonging to the village or district of the initiate, but for Tavatsukwe and above the whole island is concerned. I n the old days fighting was put aside and all belonging to Tavatsukwe or superior divisions would come and each man would put down two fathoms to which the introducer would add another two. In addition to this general distribution to the members from the whole island there would also be an additional distribution to the members of the initiate's own village to whom the payments would be larger. Each man puts down five fathoms and the introducer adds another five ; then each puts down one and is given back this with one added and this is repeated, the net result being that each man receives seven fathoms from the introducer. The distribution to the whole island is called sar\ the first giving of five fathoms to the people of the initiate's own village is called ul lokwamal\ the second distribution of one fathom is maniligoligo and the third distribution is rkaniganga7i.

The special feature of all these distributions is that something is given to be received back with addition. The giving is called veni and it is the receiving back which is sar.

The introducer is usually of the same veve as the initiate and is very often his mother's brother, this being so in several cases about which 1 made inquiry.

The foregoing account, given to me by John Pantutun, applies to the island of Mota. It is not based on actual observation.

I am indebted to Mr Durrad for the account of the ceremonial attending the taking of a step to one of the higher ranks of the Sukwe, viz. the passing of a "chief" of Vanua Lava, Charles Wunau (see PI. Ill, Fig. 2), from the rank oi Kworokworolava to Wometeloa, which took place at the village of Vatrata. At the time Charles was the person of highest rank in the village and as there had been no one there belonging to any rank above Kworokworolava, the upper end of the gamal had fallen into disrepair and had been renovated shortly before the ceremony about to be described. The lowest rank existing at present in this gamal is Avtagataga and the other ranks present are shown in the plan of Fig. 3. Vatrata is a small

IVo/v/i

M

X

V

1 ^. - ^

D

s

X

X

i

firewood

firewood

W0ME.TELOA

o

oven

KWOROKWOROLA VA

Lano

Mel

Kerepue

Ta VA TSUKWELA VA

Tavatsukwe.

Tamatsiria

LUAV

AVTAGATAGA 1 I

Fig- 3-

Plan of Vatrata. The wona are stone-platforms. The dotted line round the drum indicates the position occupied by the dancers during most of the dance. The numbered crosses show the position of the posts or tree to which pigs were tied. The arrow indicates the direction taken by the dancing party as they came through the wood into the village. The gamal is on a larger scale than the rest of the figure and shows the ranks now in existence at Vatrata.

village which only possesses one house belonging to Charles Wunau in addition to the gamal. The whole village lies in a hollow of the hills, sloping down to the shore. The gamal is at the higher part of the sloping ground and the lowest rank is at the upper end, so that the Wometeloa end was built on a high platform of stones. Below this, with the chiefs house on one side and two stone platforms (wona) on the other, was a small flat open space used as the dancing ground. On one side of this, as shown in the figure, was a large drum laid on the ground.

The man who was acting as introducer or was "making the Sukwe " was Abraham, an old man of the rank of Wometeloa, who came from another village. When Mr Durrad arrived the pigs had already been handed over and the payments of money were being made. Many people were assembled and were watching the money being given to Charles Wunau, who was standing in the centre of the open space m front of his house. As people gave the strings of money to Charles, he measured them with his outstretched arms and laid them in a flat basket lying on the ground before him. A string of money often had leaves at the end to signify that the payment was being made as a return for food which Charles had given on some previous occasion, the leaves being either of the yam or coconut according to the nature of the gift. The last person to present money, a man named Jimmie, whom Charles called mak (father), stood for some time explaining the nature of his payment. He was giving five fathoms on his own account and another five on behalf of his daughter who was dead. When the payments were finished the basket containing the money was taken into the house. About sunset this money was measured carefully. Charles Wunau sat with some other men in his house, in other parts of which women were grating nuts for puddings. Two pegs a fathom apart were stuck in the ground forming the floor of the house, the distance being carefully measured by a measuring stick that lay handy. The basket of money was brought and the different lengths of money were strung together. Starting from one peg the whole string was wound round and round the two pegs till twenty fathoms had been measured off. The string was then cut and the end fastened with a knot to prevent the money from slipping off, and the twenty fathom lengths were laid in a basket. It was found that 144 fathoms had been

Contributions 71

presented during the afternoon. Charles Wunau then produced 46 fathoms of his own money which was added to what remained of the contributed money and measured off. Ten more fathoms were still wanted, for Abraham, the old man who was " making the Sukwe" required two hundred. When the measuring was finished and it was seen that more money was needed, all the men sat quietly for a time and looked at one another. Presently one went out and returned shortly with a piece of money which he threw on the ground and it was joined on and measured. Then followed another pause till a second man went to fetch money and it was a long time before the amount was completed, the money coming in amounts of only one or two fathoms. At the end it was found that 202 fathoms had been given. The extra two fathoms were kept by Charles and the remainder was divided into two equal parts. One hundred fathoms were put in one basket and the other hundred in another and the two baskets were hung up in the house.

Mr Durrad recorded the names of all those who gave money together with the amounts given. Thirty-seven people contributed altogether, their amounts varying from one to twelve fathoms, three also giving pigs. Both men and women contributed and it may be noted that certain persons gave their money to the wife of Charles who then handed it to her husband. Mr Durrad also recorded the pedigrees of the father, mother and wife of Charles and it may be noted that of the thirty-seven people who contributed there were only twelve whose names appear in these pedigrees, while several near relatives who were living failed to contribute at all. Thus no contributions were made by the man's father's sister or her husband, the latter being also his mother's brother, though one of their two sons gave a single fathom. Again no contributions were made by two brothers and one sister of Charles though the husbands of this sister and of another who had died contributed, one seven and the other six fathoms. Among the relatives who contributed were the father (four fathoms), the father's brother (two fathoms), two daughters of a father's brother (one, two fathoms ; the other, seven fathoms and a pig), a mother's brother's daughter's son (10 fathoms), a mother's brother's daughter's daughter (one fathom), the wife's mother's sister (one fathom) and the husband of the daughter (12 fathoms). It is thus seen that most of the near

relatives who contributed did not belong to the veve of Charles but to that of his father. So far as can be judged from this example the contributions come chiefly from distant relatives and from those of the opposite veve and it is probable that every contribution is a return for money or food, etc., previously given by Charles to the donors or their families. Thus the duty of supplying money does not seem to fall to the sogoi of an initiate but to those who have previously been indebted to him, and this example seems to indicate that these persons belong chiefly to the moiety of the community to which the initiate does not belong. It may be noted that of the nine I

people who contributed the last ten fathoms required to make ]

up the amount only three were near relatives, the father, the sister's husband and the daughter's husband, so that even in these payments which were apparently not in return for previous benefits, the supply came from persons of the other veve. None of the men who assisted in other parts of the ceremony |

were nearly related to Charles. i

After the measuring of the money, which took place about 8 p.m., a number of men sat in the middle of the open space in front of the ga7nal with small bamboo drums which they beat while women danced. As they played the drummers sang a song. Before the women came out of the house to dance three little boys ran round the group of drummers several times. The drumming went on till midnight with frequent pauses while the women danced round the drummers in pairs holding each other's hands. The drummers had to be paid by Charles. While the dance was going on some of the men were drinking kava just outside the door of the lower end of the gamal, and within the gamal a large fire had been lighted in the Tavatsukwe division and was heating the oven of stones. At midnight when the dance was finished, the fire was raked out and the oven filled with a large pile of food, and all went to sleep, Mr Durrad occupying the new division of Wometeloa while Charles Wunau slept in the adjoining division of Kworokworolava separated from Wometeloa by a row of stones.

On the following morning at six o'clock the oven was opened and the food pounded on a large flat pudding dish, grated and cooked nuts were added, and when ready the pudding was covered with leaves and left in the gamal. While the pounding was going on a number of young men

Dance 73

went up the hill into the woods and played the meretang for about half an hour, no woman leaving the house during this time. Meanwhile other men were preparing decorations. Two long saplings forked at the top were peeled of their bark and smeared with red paint and two cycas trees, cut off at the base just above the roots, were brought in. Other men were stripping coconut leaves of their fronds which they tied in knots and adorned with feathers. To the man who fetched the cycas trees Charles gave a pig and the two extra fathoms which had remained over after the 200 had been measured, and this man in his turn paid three fathoms to the men who had assisted in making the decorations.

About 9 a.m. the dancing ground was swept clean by some women and an hour later the arrival of a dancing party from a neighbouring village, twenty or thirty in number, was announced. The people were said to be " bringing the drum," which meant that they were coming to dance to the accompaniment of the large drum already mentioned, not that they were actually bringing a drum with them. Amid perfect silence on the part of the Vatrata people the party approached in single file from the wood below the village, three men who were "the drum" preceding the rest. These three men took their seats at the drum and the rest of the party ranged themselves in front of and around the drum, squatting in a crouching posture (see PI. IV). Each man carried in his right hand a bow and arrow and in his left a handful of broken leaves. Some of the men had their hair dressed flat on either side and all were decorated with hibiscus blossoms, with sprigs of scented leaves in their armlets while some had fillets round their heads. At a signal the drum began and every man rose at the instant, threw up the broken leaves which rained down upon them and brandishing aloft their bows and arrows, the men gave a shout and began to sing, stamping rapidly with the feet in rhythm with the song. Beyond the stamping which somewhat resembled "marking time" there was little movement, but now and then they would open out and then close up again. Two singers sang together one part of the song in an alto, then the rest sang in a rather lower key, the spreading out taking place at the end of each break in the singing. At intervals they paused to rest and after several periods of dancing, seven men with conch-shells assembled in the Wometeloa division of the ga?nal and blew together, two long and two short blasts.

As they blew Charles Wunau, who had painted his legs red, ran leaping out of his house to the dancing ground. In his right hand he carried two rods for measuring money tied together near the top with a band to which was attached a small packet tied in a leaf. This packet, kept by its owner among his strings of money, contained a small stone, tied up in various leaves, which is supposed to possess virtue for increasing wealth. Mr Durrad writes, "it would be difficult to exaggerate the dramatic effect of the scene. The sun now high up shone upon the gaily decorated dancers grouped in close formation round the drum whose booming roll accompanied the shouted chorus and the trumpet blasts, while in the open space was the tall athletic figure of Charles Wunau dancing by himself."

After dancing for about five minutes Charles put the measuring rods against the thatch of the house and went to bring three stout pointed posts (nos. i, 2, 3 in the plan of the village. Fig. 3), which he stuck in the ground in a row on one side of the dancing ground, the posts being about four feet apart. Then from a shed he brought a very fine pig which he tied to post no. I and then taking up the measuring rods danced again. After a few turns he again put down the rods and fetched a second pig which he tied to the second post. He danced again without the rods and then brought a pig to tie to the third post. Two more posts (4 and 5 in the plan) were then erected and two pigs were tied to these. Then Charles himself brought a sixth post and tied another pig to it and again danced, but now with a pig's jawbone in his left hand and the measuring rods in his right, and after a few turns he went into the house.

Abraham, the old man who was "making the Sukwe^' then set up two stakes (7 and 8) close by the gamal and Charles came out of the house with seven fathoms of money in his right hand and taking up the pig's jawbone in his left hand, he danced again a few turns. A woman belonging to the village of the dancing party then came and stood in the dancing ground near the house and Charles laid the seven fathoms of money over her left shoulder. Three of the fathoms were to repay money she had given to the dancers, a tenth of a fathom to each, while the other four fathoms were given in addition. When he had given the money Charles again took the measuring rods and danced. A man

Plate IV

Dance 75

then brought out three more pigs and tied two to the stakes set up by Abraham (7 and 8) and the third to a cycas tree (9) near the ga7?tal and then Charles went on dancing. When he ceased Abraham put a leaf of the tree-fern in the thatch under the eaves of the house just over the doorway, and then went to the dancers and received from one of them a bow and arrow which he carried into the house. Charles then danced again, holding four fathoms of money in his right hand and the pig's jawbone in the left and when he stopped, a small boy, whom Charles called mak, came from the dancing party and Charles laid over his head the four fathoms so that one end of the string of money hung over his face and the other down his back. The boy had helped to hire the dancers and had paid them two fathoms and he received the other two fathoms as interest on his outlay. Then Charles took up the measuring rods and danced again.

While this was going on the dancers had also been performing and they now called for water which was brought to them and two women appeared at the door of the house carrying a large pudding on a dish. Some of the party danced up and down and took the dish from them and returned with it to their companions. The pudding was cut up and each man took a slice and ate as he went on dancing. Abraham, his legs smeared with red pigment, then joined Charles and both danced, the former holding the tree-fern leaf in his right hand and his bow and arrow in his left hand. At the same time five men stood outside the gamal and blew three blasts upon conch-shells. After they had danced for a while, Charles stood still in the centre of the ground and the old man walked to and fro before him and addressed him in a short speech \

At the conclusion of his speech the old man threw down the fern leaf at the feet of Charles who took it up with his left hand. It was said to be a symbol of peace, signifying that there would be no destruction of Charles's property. The dancers were then told to continue and Abraham and Charles again danced, the former now holding the arrow in his right and the bow in his left hand. Both men then stood by pigs nos. 7 and 8 and Abraham slapped them in succession on the back, speaking to Charles as he did so, and then both

1 For an account of a speech made on a similar occasion, see Codrington's M., 109.

danced again. Then Abraham fitted his arrow to the string of his bow and drew his bow to shoot the pig no. 9, but Charles laid his hand on his arm to stop him. The pig was to be killed for a feast, but as the next day was Sunday the killing was to be deferred to the following week'. Charles then went on dancing alone with the jawbone and measuring rods and then sat down to rest in the shade near his house. When he again rose he untied the little packet from the measuring rods, took it into his house and came out again bringing a bundle of kava roots and danced with the bundle, first, in his right and then in both hands. At a period in the song of the dancers the kava roots were laid on the ground by the side of pig no. 2, and a little later Charles brought out the two baskets of money that had been measured on the previous night and laid them by the kava roots and then danced again with the measuring rods in his right hand.

Stacked up against the end of the gamal was a large heap of firewood which had been gathered by the whole community and at this point in the ceremony pieces of it were handed into the Wometeloa division of the gamal and two men prepared a fire. Six men then stood outside the gamal with conch-shells and blew fourteen or fifteen times, blowing " for the pigs and the money." As they blew Charles and his father Alfred stood by the pigs nos. 5, 4, i, 2, 3 and 6 and both smacked the backs of the pigs in succession one after the other. Then there was a pause in the dance and when it began once more the conch-shells were blown again, and Charles and Alfred touched the baskets of money and Charles laid down the measuring rods upon the baskets. At this stage the conchshells were handed into the gamal, but probably by mistake, for they were at once passed out again and blown while Charles brought out another pig and handed it by the thong to Abraham and then danced again with the jawbone in his hand. Abraham then carried the money and measuring rods into the house and coming out again loosed pigs nos. 3 and 4 which were taken away. The pig which should have been shot was also taken away to be kept for the feast and as it was led away the dancers danced behind it and thus made their exit from the dancing-ground. Then a fire was lit for the first time in the Wometeloa division and that evening

1 This is, of course, a modification of the ancient ritual due to Christian influence.

Songs 77

Charles ate food cooked on this fire, Abraham took away the remaining pigs which became his property. The men of the village and those of the dancing party sat about the village near the lower end of the gamal and ate the food which had been cooked in the Tavatsukwe division.

On the evening of this day the trees which had been prepared and painted red in the morning would be set up. When this is done all the women leave the village and the men play the meretang and six conch-shells are blown, with two short and two long blasts. The cycas trees are planted near the door of the gamal where they will take root and grow, being the sign that a man has ascended in the Sukwe. The saplings painted red which are now called balak are set up wherever the man who has taken the step wishes, usually either in his garden or at the place on the shore where he bathes. They tapu the place and no one but himself may go there. Anyone who wishes to visit the tabooed place to fish or fetch food has to pay a large sum of money to the man whose signs of rank they are. This payment removes the tapu, upon which occasion the owner has to kill a pig and make a feast for all the members of the Sukwe who belong to the same rank as himself.

The following are the songs sung by the dancing party on this occasion. The first is an old war-song composed by a man of Santa Maria whose name was Ristan. The second is a rain-song.

(i) Pagaloi^.

Rera rera raloke^ t

Vetur riaris mi h ma. Turn again hither to me,

Na we vet nalngak at minik. I speak my speech to you.

Lige be vus e Agreement for a fight

Ve siu ma. Na ve ti naasik. Has come down here. I have made

my song. Ve siu ma. Na ve tia naasike. Has come down here. I have made

my song. Dero ve sarsar doro. We two are equals*.

Ina m^ ven we tamat me lele mermer. I shot a dead man in my youth. Verong na melevnat, verong narak Sacred is my fame, sacred is my fame

mega, that has spread abroad,

Nararov suk me ga sur vono. The fame of my bow has spread in

every land.

' Mr Ray suggests that this is a variant oi vagalo, a fight. 2 This line could not be translated. ' This refers to the man and his son.

Melanesian Society

(2) Ligo Naslang. Ligo naslang: na vet nalingake.

Wen tur kwonge

Ve asas ameke Ser Sere ven dung tal

we e vofio. Na me mbul nangusuk, Vear bat na maran. We riimbu Mataka ve rumbus we

vano. Na me mbul nangusuk ; wu iogala ni

ve kal ma ni vus menesvesi.

Wu rerono e tur ni me on. Na te meneg ve tawag amaki Merelav. Na ngamusuk ni 7ne tev le melik.

(3) Ligo Naslang.

Ligo naslang: ina me wot na ive

iuwali. Iroteme ve tu ler na. Nau rembi gar tur; na voluk vasi.

Navoluk Wegamal niik ; nike ve tangtang rit goro na me le ti.

Vetameni nasak

Tevogol ris wule mina,

Tevogol gar lele tani vereket wo vono mina.

Nau melig vanvan ; na vie van lere me sur kemi.

Aea. Namau elngak ve vega sur la.

Ve asas levarang rangrang^ wovere lumgev.

(2) Appoint your messenger.

Appoint your messenger: I have

uttered my speech. A day of rain floods indeed Has come upon them up there at Ser^

and resounded through every land. I have anointed my Hps, Very sacred every morning. The foam at Mataka^ has splashed

as rain over the land. I have anointed my lips; the North

West wind has come up hither

with a rushing sound. A calm is on the harbour. I will charm the clouds that open

out over Merlav. My lips have

pierced the clouds.

(3) Appoint your messenger.

Appoint your messenger: I am the

only one born here. The people have forsaken me. I am like a cockle-shell on the beach;

I have no companion. My companion is my son Wegamal;

you have wept over me. Like my name Has disputed the law to me, Has disputed over the land that has

fallen to my lot. I am a floating cloud; I have wandered hither to you. Howl. My voice has reached the

shore. It has pierced your withered breast,

the breast of youth.

Mr Durrad was told that every man must possess a song. Such a song is recognised as his song though it may be used by others. There are a number of men who are noted as song-makers and when a man wishes for a song he goes to one of these men and gives him half a fathom of money. The song-maker retires to some place by himself and with some rite in which he utters formulas and uses objects with mana, such as sprigs of trees and sea-water, he composes the song. He revises and revises it till it is perfect and then calls together the people of the village of the man who wants

' Places in Santa Maria.

Songs 79

the song, this man himself not being present. He sings the song to the people till they know it and then they go to sing to the man who has ordered it. The new owner gives a fathom of money to the composer and tells the people to fetch food from their gardens and cook it. While the food is cooking the people sing the song to the man, and when the food is ready the ovens are opened and a specially choice pudding is made and the new owner of the song cuts the pudding and as he does so he gives five or six fathoms to the composer, saying, "I pay for my song with this," and the song-maker gives three or four of the fathoms to the people who sang the song for the first time to its owner. It was said that every man must have a song, for the nature of his reception after death by the ghosts of the dead will depend on whether he has a song or not.

While singing at a dance there is a good deal of gesticulation. At every mention of money the singers touch the left forearm with the right forefinger ; when clouds, rain or sun are mentioned the right hand is lifted above the head ; at any mention of shooting the people bend the arms as if drawing a bow while at any reference to speaking they make a gesture similar to that known as "blowing a kiss."

It will be noticed that this account does not include any actual ceremony of initiation unless the eating of the first food in his new division by Charles Wunau is so regarded. In Mota I was told that the ceremony of eating by the side of a cycas tree took place for every step in the Sukwe. It is probable that the ceremony by the side of the cycas tree had already taken place before Mr Durrad arrived, having been performed as in Mota during the night preceding the distribution of money.

In Saddle Island or Motlav I was told of a special feature of the initiation into Avtagataga, and this probably also happens in Mota and other islands of the Banks group. In the following account the terms given are those of Motlav. The cere- mony is performed in the village when the sun is setting. The people gather in the open space of the village where mats are spread. In the old days the initiation and these ceremonies would always take place during childhood and I therefore assume that this is so in the following description. The child is seated on one of the mats and numerous women, it may be as many as twenty, who are to take part, sit round him on

8o Melanesian Society

mats. The head of a pig with tusks is put down on the mat before the child. Each woman gives a short fathom (from the fingers to the opposite shoulder) and the child gives back three tal (double fathoms) and this money is put in a bag by the side of the pig's head. The child then gives a piece of money to his mother's brother who is standing by, but does not get up for this purpose. An old man now blows the conch-shell four times and at the end of each blast the child takes a stone and puts it down on the top of the head of the pig, this being called nir or nir kwat rau. The boy is then carried into the gamal on the back of his father's sister who is always the chief among the women taking part. The mats on which the pair had been sitting are taken up and all the other women take up their mats and follow the pair into the gamal, only going into the lower divisions up to Kwatagiav. The child, now called nat vuhe rau, is left in the gamal for five days, being given mats on which to sleep and at the end of this time he would kole^. The child would be painted on face and body with red and black stripes, the person who paints him receiving a fathom of money. When the child comes out of the gamal the chief men beat drums. Twenty men take the boy to the drums and each receives from him a short piece of money called leme gambak. All the women who have taken part in the ceremony then kneel down so that as the boy goes round the drums he passes by them and steps over each of them and then sits down on a stone. While the drums are beaten songs are sung by the chief men. The men then go to the gamal to make their feast while the women return to feast in their houses.

At the present time there are two modifications; it is no longer customary to be initiated in childhood and in consequence the father's sister leads the initiate into the gamal instead of carrying him on her back and it is now the custom to stay only one day in the gamal, this innovation being due to missionary influence which has been directed to cut down the periods of seclusion during the ceremonial as much as possible.

All the women who have taken part in this ceremony call the child (or man) nat or natui, child, while he calls them vev or vev vuhe rau (Motlav) or veve vus rawe (Mota), this meaning "mother strike (or kill) hermaphrodite pig." The child

• See p. 131.

Masks mtd Knives 8i

or man may not marry any of these women and sexual relations between them would be regarded very severely. It is the duty of the women to help the man in every way and give him food while in return he would respect and obey the women and would help them in their gardens. The father's sister would always be the chief among the women taking part in the ceremony and would receive more money than the rest. This forms the sole occasion on which women go into the gamal and as we have seen it sets up the closest relationship between the child and the women who take part.

Certain objects or acts especially connected with the gamal may now be briefly considered.

Each of the ranks of the Sukwe has connected with it a hat or mask which has the same name as the division to which it belongs. These objects are made in two ways ; either with handles so that they can be carried in the hand, or hollowed in the place of the handle so as to be worn on the head. I could discover litde about the functions of these objects. It seemed that a man might go right through the Sukwe from its lowest to its highest rank without concerning himself about these hats in any way, while, on the other hand, a man may have a hat made for a kolekole ceremony which belongs to any division, even one superior to his own. When used for this purpose they are made in the salagoro^ of a Tamate society, and then taken to the gamal from which they are brought out later to be exposed to the common gaze (see Chap. V). It seems clear that whatever may have been their importance in the past, they have now little connection with the actual business of the gamal, and since they resemble the masks or hats of the Tamate societies, and like them are called tamate, it seems probable that we have in their existence the survival of a relation between the Sukwe and the Tamate societies, which was once closer than that which exists at the present time.

In some islands knives of a special form called igot are used to cut the puddings made from breadfruit or other materials. These knives seem to have been especially developed in the island of Ureparapara and, with two exceptions, all those shown in PI. V, VI and VII come from this island. The exceptions are PI. V, B 9, which is from Santa Maria, and PL VI, C 6, shown both in full face and profile, the exact provenance of which is unknown. These knives, most of which

were presented by Mr J. C. Palmer to the Cambridge Museum, are of two kinds. The upper part or handle of most of them represents the human figure in various degrees of conventionalisation, while two, PI. VII, D i and 2, are derived from the til or sword-fish. Other examples illustrated by Mr Edge-Partington' are shown in PI. VII, E i — 4.

These knives, or at any rate some of them, are used only by certain ranks of the Sukwe, thus A i and A 2 are used by those of the Kworokworolava rank and of higher ranks than this ; B 7, C i and C 2 and also the larger knife of the fish motive, D i, are used by the Lano division and those of higher rank ; and B i to B 5, A 6 and C 3 are used by the Mwele rank and above. The knives of the Mwele division are said to be distinguished by the fine serrations of the handle, and if this is correct, A 4 and A 5 may also be assigned to this rank. It will be noticed, however, that these serrations are also present in C i and in D i, ascribed to the Lano rank, and round the heads of A i, 2 and 3.

The dancing dress called malo-saru^ shown in PI. VIII is also connected with the Sukwe. These have now almost disappeared, but were worn in the old time in the dance connected with initiation. This malo-saru is almost certainly woven and not plaited.

Kava Drinking.

One prominent feature of the life of the Banks islander is closely connected with the Sukwe and may therefore be considered here. Kava which is called gea is only drunk by men, and at one time only those high in the Sukwe, of the rank of Tavatsukive and above, were allowed to drink it and even now those below the rank of Tavatsukwe may not drink kava in the gamal, though they may do so in the salagoro of a Tamate society, where at the present time all ranks participate. When used at the ga7nal, it is made and drunk either within the building or in its immediate neighbourhood.

Women should never drink kava or take any part in the ceremonial connected with its use'. At the present time the

1 See Man, 1909, ix, 182.

2 This dress has been presented to the Cambridge Museum by Mr Durrad.

3 I never saw my informant more heartily amused than when I told him that I had seen kava being made by a woman in Samoa.

Plate V

B I

5 6 7

Pudding- Knives.

Plate VI

C I

4 5

Pudding- Knives.

Plate VII

D I

E I

Puddino^-Knives.

Pl.itc I'll I

Malo-sarii. from Ureparapara.

I

Kava 83

people are beginning to drink kava in the house, on which occasions women are allowed to be present, but it is only recently that they have even seen it being made. It is said that men of high rank in the Sukwe have been known to give kava to their wives, but this only quite secretly.

The act of drinking kava in the Banks Islands is called woana, and its special feature is that it is prepared by one man who chews the root of Piper methysticum for each member of the circle in turn. In the island of Mota four vessels are used, all small cups called lasa (see Fig. 4), made of coconut or wood. Names were given for these four vessels, but it was not clear that they are in general use. As, however, they will make the following description clearer, I propose to use them. The first is called the lasa woroworo, the cup into which the Fjg- 4-

juice is squeezed ; the second is the Kava-cup from Mota.

lasa unun, the drinking cup ; the

third is the lasa samwai, the cup for the dregs, and the fourth containing water is the lasapei. The maker, who should be of Tavatsukwe rank (though this rule is now being relaxed) must sit cross-legged with his feet under his knees. He washes his hands by pouring water over them and rubbing them together, and then claps his hands, this being the sign to those without that kava is being made. He takes a piece of kava root in his mouth and chews for some time, taking also small amounts of water in his mouth with it. When the mass is sufficiently chewed it is taken out of the mouth and rolled with the hands to the form of a ball (it was said that this must be done) and the ball is then squeezed so that the juice drops into the lasa woroworo. Some water is then added to this from the lasa pei and the contents of the cup are poured from a height into the lasa unun and the dregs left in the lasa woroworo are poured into the lasa samwai. The liquid is then poured back into the lasa woroworo from a height, and the dregs in the lasa unun are put into the lasa samwai, and this pouring from one vessel to the other is repeated several times. Finally, the liquid is poured into the lasa unun, water is added and the mixture stirred with the thumb and handed to the oldest man present. If there is a man present of higher rank in the Sukwe than the oldest

6—2

man, the latter will hand him the cup, but it will usually be returned so that the oldest man will drink first. The cup is then handed back to the maker who will take the same ball of kava root into his mouth and proceed to chew it in the same manner as before for the second person. The maker may chew the ball and squeeze out the juice several times for any one person before he gives it to him to drink, thus making a larger amount of kava.

After using the same mass of root for several persons the kava will become weak, and when the maker thinks that it is becoming too weak he will ask the next man to taste. The man will perhaps say, "It is only water," and the maker will then take a fresh piece of root. When all have drunk or when the kava is exhausted, the maker washes out his mouth with water, and puts the chewed mass of root into the thatch of the gamal, where it will remain as long as the gamal stands, becoming squeezed into a flat cake by the pressure of the thatch. Its presence in the roof of the gamal is a memorial of the kava drinking, and the owner will be proud if there are very many cakes of kava root in the thatch of his gamal. The maker of the kava receives as his reward a piece of the root.

The kava is drunk in a special manner. As the maker hands the cup he holds it on his outstretched right hand and the drinker takes it with his right hand and sucks up the liquid without putting his lips to the edge, and when I saw the drinking, the liquid was sucked up by the old men with much noise. It was said that the younger men often now drink from the edge as in the customary way of drinking from a cup. Before the first man drank on the occasion at which I was present, the maker poured some of the liquid on the ground before handing the cup to the first drinker^ I did not observe any spitting or blowing out the breath on this occasion. The vessels used for making the kava are usually kept in a basket in the dwelling-house. There is a rule that they should never be put on the ground, but kept either on a basket-framework or on a ring of coconut shell. It is a sign of the change which is now affecting custom, that the young men ridicule the care of the elders that the kava cups shall not touch the ground, and if an old man wants a plate, they will say in derision, " Why do you not take that.-*" pointing to the frame on which the cup is resting.

• See p. 85.

Kava 85

The dregs which have been poured into a special cup are not thrown away but may be drunk by anyone of Tavatsukwe rank. Before being drunk the Hquid dregs are strained by pouring on coconut fibre and then squeezing this into a cup, but this is the only time that there is anything of the nature of straining.

On Mota kava is drunk especially on days when anything is going on in the Sukwe, and I was not told of its use in connection with ceremonial of other kinds. I am indebted to Mr J. C. Palmer for an extract from a notebook belonging to the late Archdeacon Palmer, giving a prayer used on the occasion of drinking kava which would indicate the existence of a more strictly religious aspect of the practice.

This formula runs : Tataro mamo^ motangarogea nake neira golgol ti

Prayer drop of kava this they trouble continually

amenau neira tamate. me they ghosts These words were said when the kava was ready for drinking, and as they were said a very small quantity was poured out on the ground ^

At Pek in Vanua Lava it seemed that only two vessels are used and sometimes only one. One seen by us was a shallow wooden cup with a handle which was called lasa woana^, and sometimes the chewed root is merely squeezed into this, mixed with water and drunk. When a second vessel is used this is a coconut cup called lasa un. In this case the juice is squeezed into the lasa woana, poured from that into the lasa un, then back again into the lasa woana from which the person drinks. It was said that a wooden cup is only used low down in the Sukwe, and that the higher ranks only use coconut cups. As in Mota the vessels should never be allowed to touch the ground. Here also the liquid should be sucked up without putting the lips to the edge of the cup. After drinking the breath is blown out strongly {pupsag) in such a way that some of the kava is blown out too,

1 Mr Ray to whom I am indebted for the translation of this formula suggests that this word should be fnama, father. Another formula in Archdeacon Palmer's notebook begins ''''Tataro, pupua''^ or "Prayer, grandfather." An alternative reading is mam 0 tangaro gea, "a drop of kava for you to drink." In each case the difficulty is the mo immediately preceding tangaro. Mr Ray cannot translate it and it is possibly an error of transcription.

^ See Codrington, M., 147, where another formula will be found.

3 Woana is the name for the act of kava- drinking.

exactly as is done in the Torres Islands (see Chap. VII). The motive of this was said to be merely to clear the teeth of fragments of kava root which might be sticking between them.

Mr Durrad observed an unusual method of drinking kava, which was practised during the dance when Charles Wunau was taking the step to the rank Wometeloa (see p. 72). In this case the men sat just outside the lower end of the gamal, the maker sitting in the middle of the group with a pudding-dish, hollowed deeply in the centre, before him. He scraped the roots of kava on a stick called vat like that used for grating nuts, viz. a stalk of the tree-fern (kwatia) studded with numerous and very short prickles. When a mass as big as the two fists had been scraped it was moistened with water and the juice squeezed into a very small and shallow coconut cup. A somewhat larger cup, the lip of which was hollowed at one part, was then filled with water and the kava juice poured into it, and it was then passed to a man who put the hollowed part to his lips and drank with a prolonged sucking noise, pouring the last drops on the ground.

The anomalous features of this procedure are that the root was scraped and not chewed and that a large pudding dish was used, and Mr Durrad was told that the kava was being made in the Fijian manner, the knowledge of which had been brought recently by natives who had been to Fiji. The pudding-dish evidently represented the kava-bowl of Fiji, but the most characteristic features of the Fijian ceremonial seem to have been absent.

Dr Codrington states that the introduction of kava into the Banks Islands is recent, and brings forward as evidence of this that the custom had not yet spread to Santa Maria. According to Mr J. C. Palmer the use of kava, instead of not having spread to Santa Maria, has been given up by the people of that island, only the chiefs having been allowed to drink it. Mr Palmer believes that it has never been used in Merlav but as they have had the Sukwe there, and as the use of kava is elsewhere closely connected with the Sukwe, we ought to hesitate before being sure that it has never been used in that island. At the present time the root is grown most plentifully in Ureparapara, and is used more frequently there than in the other islands of the group, where its use is limited by the smallness of the supply.

CHAPTER IV

BANKS ISLANDS

The Tamate Societies.

These societies which take their collective name from the word for ghost or dead man are very numerous. In the following pages I give some account of 77 societies from the island of Mota alone ; many of these have numerous subdivisions and there are probably many more in other islands.

The societies are of several kinds. Some can be entered by boys who have not yet been initiated into the Sukwe, who cannot therefore eat in the gamal, while in one case at least a girl has become a member. They are called Tamate matawonowono, meaning 'Hamate with their eyes closed," and the ceremonial connected with them is extremely simple.

The nature of a second group of societies is doubtful. These societies are only open to those who belong to the Sukwe but are not specially connected with the characteristic lodge of a Tatnate society which is called a salagoro. They are known as talo 7naea or "belonging to the open," and were said to be pardy of the Sukwe, partly of the salagoro. One of them, the Kwat, was not counted as a Tamate society by Dr Codrington, though, as he says, it closely resembles them in all its main characteristics, but John Pantutun included it in this group without any special comment. We have at present too little knowledge to enable us to define the exact nature of this group.

A third group is clearly marked by a definite connection with a salagoro. The chief of these societies is the Tamate liwoa of which a man must be a member before he can be initiated into the others, called collectively the wangarai (or branching) Tamate.

A further distinction is that some societies are regarded as male and others as female, but it is a question whether this distinction does not apply rather to the hats or masks of the societies than to the societies themselves. Often two or more societies are associated together, the members of one enjoying some of the privileges of the other or others, and in some of these cases the societies so connected, or their hats, are regarded as male and female.

There are certain societies called Tamate kiskislag the members of which chase, injure and terrify people. The members often pelt the uninitiated with stones or they may shoot at them. In the old days they even killed people but this was stopped owing to the influence of Bishop John Selwyn.

Each society is entered by a ceremony of initiation which varies greatly in complexity and length for the different societies. The native word for initiation is tiro which is used in two different senses. A man is said to tiro the Tamate of the society which he enters and he is said to tiro mun (to) the man who undertakes his introduction. The money which the candidate collects is paid to this introducer who in his turn has to provide a pig. Before this pig is killed it is laid on its back with its legs stretched out, and each man who is to be initiated stamps once with either right or left foot on the breast of the animal which is then killed and eaten. For the unimportant societies which can be entered by those who have not eaten in the gamal, pigeons may take the place of the pig. If a hat or mask is made the initiation has to be followed by a period of seclusion ; the candidate has to goto.

In some societies a number of men maybe initiated at one time, the necessary money being made up by equal contributions from all, so that the expense is divided among many. A society which can be entered in this way is called a Tamate ninganifiga, the latter word being used for any purchase to which several persons contribute. Other societies, including the most important, can only be entered by one man at a time' who has to contribute the whole of the money necessary, these societies being called Tai?iate kerkerei. Two societies which were formerly kerkerei have recently changed their rules and can now be entered by a number of men at once.

Mt is possible that two or more men may be initiated at the same time, but in such a case each has to pay as much as if he were entering alone.

Plate IX

Drawings of Tamatc Ihooa and I'ioi', etc.. on the door of a building in Alerlav. The two simpler lozenge-shaped figures represent Tamatc viov and the three large and more elaborate figures Taiiiate Ihuoa.

The Satagoro 89

It may be noted here that the word tamate is used in several different senses in connection with these societies. It denotes the societies themselves, the individual members of the societies, the hats or masks worn by the members and the beings, probably in all cases ghosts, they are supposed to represent. Uninitiated persons will speak of meeting or seeing the tamate, meaning the members of the society in their disguises, in which case no distinction is supposed to exist, theoretically at any rate, between the members, the masks they are wearing and the beings they are representing. In addition to these uses, the term tamate is also given to the hats of the Sukwe and at the present time this usage has been extended to include hats of any kind worn as the result of European influence. Lastly, tamate is the usual term for a dead man or a ghost \

The Salagoro.

The term salagoro, the possession of which marks the third group, does not necessarily mean that the society possesses a definite building having this name. The salagoro may be only a clearing in the bush or it may be merely a spot in the neighbourhood of a salagoro building. Thus in the island of Vanua Lava I visited the salagoro nivat which was merely a well-defined oblong clearing which had evidently been used for a long time and bore every sign of careful attention. Again, round the salagoro building of the Veverau district of Mota to be described later there are a number of cycas trees which mark the sites of places regarded as the salagoro of smaller societies connected with the Tamate liwoa. Again, a man may plant a cycas in his garden as a sign that it is saretamate, i.e. the lodge or resting place of a tamate. This place may remain so for an indefinite time and can only be released by selling the land when, if the new owner wishes, he can make it ul or "free" again. Every salagoro building has a door of a special kind, in the shape of a lozenge or diamond, the conventional representation of the face of a tamate. (See PI. IX.)

The salagoro forms a second kind of men's meeting place, in addition to the gamal, being used even when there is no

1 In the following record I distinguish the sense in which Tamate is used for a society by writing it with a capital letter.

especial ceremonial in preparation or actual progress. It is in the salagoro that the special objects connected with the societies are manufactured and kept.

Certain objects and processes are denoted by special terms in the salagoro, i.e. there is a special language connected with the salagoro, at any rate in vocabulary. A few examples of such special terms will be given.

Masks, etc.

Most of the Tamate societies possess objects, either worn as hats or masks or carried in the left hand. These often bear the shape of the animal or other object from which the society takes its name. All these objects are called tamate whether used as hats or masks or merely carried. They are used when the members go about the island and may therefore be seen by the uninitiated, though only at times when they cannot be closely observed. The tamate is carried in the left hand in order that the right hand may be free to strike people who are in the way. The object of the tamate is partly disguise, partly to produce fear among the uninitiated and enhance the mystery of the societies.

At times other than these expeditions the tamate objects may neither be seen by women nor other uninitiated persons, nor should women come into relation with them in any way even when hidden. Thus when the mask shown in PI. X was given to me, the condition was made that it must be put in a box and taken down to the landing-place next morning before any women were about. A special point on which stress was laid was that no woman should be in front of it.

Many of the tamate objects have the form of animals and in these cases it is from the animals that the societies take their names. When a society has a number of subdivisions each of these may possess a tainate object from which it takes its name.

Many of the masks resemble one another closely and it may require an expert to detect the difference. At the present time long intervals have elapsed since certain societies have received a new member, and as in some of these cases the people have not been certain about the exact form of the tamate, they have ceased to make it, so that it now sometimes happens that a man joins a Tamate society but never wears

Masks 91

its mask, this part of the ceremonial of the societies being thus in process of disappearance in some cases.

As already mentioned the masks are of different sexes. Male and female tamate objects differ definitely in shape, male hats being lozenge-shaped, tall, or oblong, while female hats are rounded (see Fig. 5). The Banks islanders regard everything large or long as male, thus a long soma fish is regarded as male and a short one as female, and it is possible that the difference of sex ascribed to the masks merely indicates their shape and is not a point of great importance. It may be noticed that the lozenge forms the orthodox representation of the face of a tamate.

Another object connected with a Tamate society is the saka stick used to frighten women and children. In order to do this a man blackens his face and body with charcoal so

a be

Fig. 5. Shapes of hats; a, female; b and c, male.

that he will not be recognised and goes out carrying the saka in his left hand while he wields a club in the right. He hits the people who do not get out of the way with the club and moves the saka about continually that no one can see it distinctly and he is careful never to be still himself for a moment. The saka has on it the representation of the face of a tamate with the two eyes showing definitely.

A hat or mask should never be allowed to fall to the ground when being worn. This is a most grievous offence which in the old days was punishable by death. Even in the initiation of a baby into one of the minor societies the hat is tied tightly on the head lest it should fall to the ground.

When a man is initiated into a society it is not necessary that its mask or other special object shall be manufactured. If the man wishes to have this done, he has to pay some

9^ Melanesian Society

one with the necessary knowledge, but the making of the hat is not the occasion of any special ceremonial.

The mask shown in PI. X is that called ivis or owl and is one of the tamate of Tamate liwoa. I obtained the names given to different parts. The white hour-glass shaped objects represent teeth, and both these and the curved or straight white bands which run transversely round the hat are made from the pith of the weslaui tree, the pith being called pagasula when plain and iheo when it has black marks on it as in the case of the transverse bands. The surface in which the eyes are situated is a glistening black which is very greatly admired by the people of the Banks Islands and is known as pei ta Vava or the water of the Torres Islands, the method of making this pigment having come from these islands. The surface itself is called lama or sea. The bands separating the upper iheo from one another and those above and below the teeth are covered with Abrus precatorius, which is ordinarily called wowele^, but when used in this way or in any other way connected with the salagoj'O, it is known as wo gaviga. The fowl's feathers at the back of the hat are called weo. The general framework of the hat is made of pandanus roots.

Badges of the Societies.

These are objects, most frequently the leaves of croton plants, which are used to protect the property and especially the gardens of members. They are signs resembling those which protect property elsewhere in Melanesia and when used for this purpose are called soloi.

The special feature of the sign of the Banks Islands is that it only protects property from persons other than the members of the society of which it is the badge. Thus a man who only belongs to one large society would find the badge of this society of very little service, for it would still leave his property at the mercy of all the members of the society. A man always belongs, however, to more than one, often to many societies and by combining their badges he may be able to protect his property against the whole of the rest of the population. At the present time there are some societies which only have three or four members. If a member of such a small society contrived to join another society to which the

^ wo is a contraction of ivoai^ fruit.

Plate X

The ■-i.>is mask of Tamaic liwoa, as seen from the side and front.

Badges 93

other two or three members did not belong, he would be able to protect his property against the whole island by using only two signs. In consequence a man is always anxious to become a member of small societies and if a society becomes numerous its usefulness from this point of view is diminished. If the society is not important in other ways, new members will no longer wish to join it and it will gradually decline in numbers. The large and most important societies, such as the Tamate liwoa or viov, are so large that they are of no value for this protective purpose and are joined for other motives.

A man who takes property protected by the badge of a society to which he does not belong, and is found out, has to give a pig to the members of the Tainate he has disregarded. Sometimes a man will put on his property the badges of many or all of the societies to which he belongs and it would seem that in this case the transgressor would have to pay as many pigs as there were badges of societies of which he was not a member. The fine is so large that a man who has stolen will sometimes confess in the hope that the matter may be kept secret and that he may get off with a reduced fine.

It could not be discovered that there is any machinery for enforcing these punishments nor that there was any need for such a machinery. Here as among other communities of the ruder culture the respect for custom appears to be so great as to remove the necessity for any tribunal to enforce the will of the community. I could not discover that the breaking of a soloi carried with it any consequences apart from those just considered, i.e. there was no evidence that infraction brought about any punishment except that involved in the fine paid to the society or societies whose soloi had been disregarded.

Sometimes a man may disregard a soloi by inadvertence. We shall see that a surprisingly large number of species of croton are used as badges and it is possible that a man may sometimes mistake one for another and thus transgress.

A man must not cut or use in any way a plant which is the badge of a society to which he does not belong, the penalty being the fine of a pig to the members of the society whose badge he has used. Thus, recently a man of the Viov society tabooed some bananas with the croton of the- Tamatpilagi society thinking it to be the croton of the Viov. Both kinds of croton are white but the leaf of the Tamatpilagi has a spot of red at the base which he did not notice. A man

saw the leaf in use as a soloi and asked who had put it there. The man of the Viov society said he had done it, and when it was found that he had not been initiated into Tamatpilagi he was told he would have to pay the usual fine.

One way of avoiding the fine is to be initiated into the society whose badge has been wrongly used and in this case the man joined the society instead of paying the fine. Sometimes men will adopt this device in order to get into a society. They will decorate themselves with a croton to which they are not entitled and when discovered will join the society. It seems as if they take advantage of the pity of the members who will allow a man to join the society rather than submit him to the heavy penalty to which he is otherwise liable. The members of the society have the right to refuse admittance to a man who has transgressed, for it was said that a man could not force himself in this way into a society whose members did not want him.

When a society has several subdivisions each of these may have its own badge or badges, and in another recent case a man belonging to one branch of the Tamate nivat had to pay a fine of a pig to the members of another branch because he had unwittingly used their leaf instead of his own.

A man may not destroy a croton bush or other plant which provides the badge of a society to which he does not belong, and as no woman is initiated into any society, it follows that no woman may destroy these plants at all. If a woman finds a croton when clearing the bush for a garden, she must not touch it. Recently when John Pantutun and his wife were making a clearing for a garden, the latter came upon a croton which she asked her husband to cut down. John saw that it was the badge of one of the less important societies of which he was not a member and refused, but it happened that a little boy who had been initiated into this society was present and John called to him to cut the plant.

There seems to be one exception at least to the rule that an uninitiated person may not use a badge, viz. the case of the Malop society, one of whose badges is the green leaf of the coconut which is stuck upright in the ground as a soloi, but this can be used and carried, not only by uninitiated men, but also by women.

The use of the badges of the societies as protective signs plays a large part in disputes connected with gardens. If two

Tmnate liwoa 95

men have a dispute about a garden, one may keep the other away by putting up the badge of a society to which the other does not belong and the latter may retort in a similar manner and keep his opponent away by the use of the badge of another society. A man thus warned off a garden may ask a friend who belongs to the same society as his opponent to remove the sign but in such case there is nothing to prevent its being reimposed so that the advantage is only temporary. I obtained one piece of evidence, doubtless only a fragment of the whole, which shows that these societies have certain functions in connection with the growth of crops. When the yam crop in Mota is ready for digging a great feast is held by the Tamate liwoa. Its members collect food, seeking out the largest yams and hundreds of coconuts, and eat and drink in the salagoro. The shells and husks of the coconuts are not thrown away but are placed in a special enclosure elaborately decorated by the side of which is put a stake or palako, painted red, with coconuts on it. This enclosure is made near some especially good coconut tree the fruit of which becomes taboo for a time, the palako acting as a sign of this condition. We passed such an enclosure in Mota on our way to the salagoro of Veverau, and the palako and several growing crotons were still to be seen by its side although it was three years old. As an explanation of this ceremony it was said that the Tamate should always have the best of everything.

Tamate liwoa.

The chief of the Tamate societies is Tamate liwoa or Tamate werewere, the first name meaning the great or chief Tamate while the second is derived from the special sound which may only be made by those who have been initiated into this society for several years. It is into this society that initiation is necessary for progress in the Sukwe beyond the rank of Tamatsiria (see p. 64). In the Veverau district of Mota the salagoro of this Ta7?tate is a fine house constructed in the same manner as the ordinary house of the village, except that the door is in this, as in all salagoro, diamondshaped so as to resemble the face of a tamate. In front of the building is an open space with a platform or kwer on which there is room for two men to sit and make the werewere

96 Melaiiesian Society

sound, and there is a large oven where the food is cooked for a great feast which takes place at the end of the period of seclusion which follows initiation. At the sides of this open space and all round the salagoro-\i\iAdi^\x^ cycas trees are growing which mark spots to which the name salagoro is also given. These are the sites devoted to a number of minor Tamate societies which are connected with the Tamate liwoa. Each spot or salagoro may only be approached by those who have been especially initiated into the Tamate of that salagoro, and if a man is not a member of these minor societies localised outside the large building, he may only go a certain distance inside the main building. The three men who accompanied us to the salagoro belonged to different minor societies and one man could on this account penetrate further into the main building than the other two. In some cases the cycas tree marking the site of a salagoro has died, so that there is no obvious sign to indicate the exact spot of its salagoro, though it remains perfectly well known to all.

The Ta77iate liwoa has many badges, viz. the fruit of the gasor tree ; leaves of the sala, 7nai and gamas trees ; five varieties of croton, red, white, turturuga (green), yellow and one very long and thin ; and lastly a piece of the stem of the hibiscus {var), about six feet long and stripped of the bark.

In PL IX there are shown three representations of Tamate liwoa, apparently drawn upside down. The figures evidently represent the human face, the eyes and mouth being definite, but I could not obtain any explanation of the other features.

The Tamate liwoa has certain subsidiary societies or, at any rate, different hats, two of them being wis (see PI. X) and kwatwasawasa. These do not require separate initiation, but a man who becomes a member of Tamate liwoa is ipso facto initiated into Wis and Kwatwasawasa.

The Werewere Sound. The characteristic feature of the Tamate hwoa is that only its members may make the sound called werewere. The method of making this sound is the secret which is revealed to a new member at his initiation. It is produced by rubbing a stick called sur on a stone. In the salagoro of the Tamate liivoa in Veverau there are two large stones on which the sound is produced on special ceremonial occasions, and these are named, one being Lova and the other Kwasematika. There are other smaller

The Werewere Sound 97

stones which may be taken away from the salagoro and are used by those learning how to zverewere. These may also have special names usually derived from the place whence they have been brought. The sticks are made of several kinds of wood, one very old stick shown to us being made of aru (Casuarina) wood said to have been brought from Motlav. The bottom of the stick is bound round with twine to prevent it from breaking, and the whole stick is covered with leaves of the umbrella palm, one of which hangs down from the upper end, and the other from the middle of the stick. The midribs are removed from leaves of the ugava tree and after being heated, these are put lengthwise on the palm-leaves to which they adhere. Over these are put long leaves of the nokaria shrub, which are tied on to form the outer covering of the stick. The sound is produced by rubbing the end of the stick backwards and forwards on the surface of the stone, over which has been rubbed a heated univalve shell, ordinarily called lata, but which in the salagoro has the name of wokwasa. The stick is rubbed on the stone at definite rates and with varying rhythm, the stone lying between the outstretched legs of the performer. On ordinary occasions each performer works independently ; the stick is first moved slowly to and fro and then quicker and quicker, the movements decreasing again in rapidity after a time. The rhythm of the movements varies greatly in complexity.

On special occasions, such as the talasa feast (see p. 104), two men make the werewere sound together, sitting on the platform (kwer) outside the salagoro building. The performance begins with ten movements, each more complex in rhythm than the preceding and then ten more which are the same as the first but in reverse order, so that the rhythm becomes less complex with each movement. Up to this point the two performers work in harmony ; then begins a far more complicated series of movements in which, so far as could be gathered, one performer carries out a comparatively regular series while the other makes sounds which are variations of this. John Pantutun spoke of the latter series^ of sounds as decorating the regular series which it accompanies. It takes very long to learn these complex rhythms and still longer to execute them properly, and at the present time there are only two elderly men in Mota who know how to perform properly in concert, none of the younger men having succeeded

in learning, and there seems a prospect that the more complex performance will soon be lost. A man who knows how to iverewere completely attains thereby a very high position in the Sukwe, and it was said that he would rank above everyone else, even those of superior grades in the Sukwe. A man has to be initiated into Tmnate liwoa for five or six years before he is allowed to begin to learn how to make the sounds, though he is allowed to carry the stones out to the platform during the ceremony of initiation. It is the stick or sur which he is not allowed to touch.

The werewere sound has a very piercing quality and carries very far. We were told that when made in Mota, it would be heard not only over the whole island, but that the sound might reach Motlav and Vanua Lava.

A mistake made during the production of the werewere sound is visited very severely on the performer. He has to pay a tusked pig, and if he cannot provide it himself, his relatives have to do so for him. We were told that in the old days a man who could not expiate the offence with a pig was hanged, the regular method of inflicting the death punishment in the salagoro.

Initiation into the Tamate liwoa.

A man cannot progress beyond the rank of Tamatsiria in the gamal till he has been initiated into the chief of the Tamate societies, Tamate liwoa or werewere. He cannot enter Tavatsukwe till this has been done and the feast connected with entrance into this rank takes place in the salagoro and not in the gamal, although it is primarily a step in the latter institution. A man wishing to join the society gives a pig {I'awe) to some one already initiated, and in doing so would smack the animal on the back just as in initiation into the gamal, a man blowing a conch-shell and saying " Let X now smack the back of the pig." This takes place in the village so that all the people know what is going to happen. The introducer gives back a pig of the ordinary kind which is known as his answer. The candidate has then to give six other pigs to six men already initiated. These men all belong to the village of the candidate, but they represent six different districts of the island. All those initiated into Tamate liivoa then give to the introducer half a fathom of shell-money, and

Initiation 99

to each he returns this together with an added full fathom (sar). The candidate then gives to his introducer a sum of money for each of the districts of the island, 30 or 40 fathoms for each. Each initiated man in the district of the candidate gives three fathoms to the introducer who returns them with three fathoms added. This money is not necessarily paid at the time of entering the salagoro, but it has to be paid before the man is allowed to leave the building (see p. 106).

The actual initiation need not take place immediately after the smacking of the pig, but as in initiation into the gamal, an interval may be allowed for the collection of the necessary money.

Certain proceedings take place before a new member is initiated into the Tamate liwoa. A house has to be pulled down and the choice of the house lies with the father of the candidate. On the appointed day all the members rush out, some in masks and with the object called saka to frighten away women and the uninitiated, and pull down the house. Some time after this the father of the candidate asks the members of the society to beat the candidate, and pays them money to do so. The candidate knows that sooner or later he will be beaten and keeps out of the way of the tamate. These go in search of him and beat not only him, but any members of the Sukwe, up to the rank of Tavatsukwe, whom they meet. They do not touch men not in the Sukwe nor women, and further, if a man who would otherwise be beaten is with women, he will not be touched. My informant was once walking with some women to a feast when he met a party of tamate, and was on the point of going into the bush to avoid them, when one of the women seized him by the hand, and in consequence the party left him alone. Sometimes a man will be able to avoid his beating for a long time, and there is a story of a man belonging to the district of Tasmate who once succeeded in avoiding it altogether. In this case the tamate became so furious at their failure that they determined to catch the man when visiting a waterhole and club him to death. When the man came from the waterhole he saw the people in front of him, and jumped completely over them only to find himself surrounded, but they all crowded round him so closely that they could not club him without risk of injuring some of their own party, and the man managed to wriggle between them and escape.

7—2

loo Melanesian Society

When the day of initiation, which is called vena inalui or vanua gona\ is announced all the uninitiated, including women and children, leave the village. Each of the men who have been given pigs takes three yards of money to the salagoro which he represents, and returns to the village when the ceremony is to begin. In the district of Veverau this takes place in a village where there is an ancient stone taken from the site of the old long gama/ (see p. 22). Those initiated will assemble from all parts of the island and the whole place is thrilling with excitement. About three o'clock in the afternoon, they hear the sounds of the werewere, the special feature of Tamate Ihvoa, when some one leaves the assembled people and goes secretly to the candidate and says to him, "Come! I want you to climb a coconut." He takes him into the bush and tells him to climb and pluck a large number of coconuts for the assembled people, taking them from trees which are the property of the candidate, and then the people come and carry the coconuts into the village. The nuts are eaten and the milk drunk, and those who have eaten may not leave the village till each has received a fathom of money from the recipients of the pigs. Then all those belonging to other salagoro leave the village and go to their own places, but the people of the district stay and cook food provided by the candidate. After eating and roasting yams, the people begin to make the zverewere sound, using for this purpose small stones, and not the large and sacred stones of the salagoro. Only those who have been initiated for some years may werewere. This goes on till the next day and is called ora maran, or playing till morning. When those who have returned to their own villages hear the werewere, they may return if they like, and in this case they bring their own wereivere stones with them, playing as they come. The noise goes on all night, and in the morning all signs of the fires and of the feast are carefully removed and the village swept clean. The new day is called O ut som or " break money " and the whole island remains gona, i.e., the uninitiated have to avoid the district where the ceremonies are taking place. After getting the coconuts on the previous day, the neophyte has returned to the village with the uninitiated so that he has

• Mwaliii means "pierced," and signifies that something secret is this day made known ; gona means "tied up," the proceedings being tied up from all except the initiated.

Initiation loi

not seen how the sound of the werewere has been made. On this morning six stakes of var or hibiscus wood are put up in the centre of the village opposite the door of the gamal. The six stakes are put up in a direction from east to west and are called mategawoso. Some one is then sent to fetch the candidate, and when he leaves his mother and sisters and other female relatives, they cry as if he were leaving them for a long time, and we were told that in the old days it not infrequently happened that they never saw him again (see p. io6). The candidate is taken into the bush and decorated with flowers and leaves ; one hibiscus flower is stuck on the top of his head, another upright at the back, and a third upside down in the hair at the back of the head. A belt known as kworotal is tied round the waist, croton leaves are drawn through it three times and left in it the fourth time, and hibiscus flowers are put on the top of the crotons. Meanwhile the people in the village will have chosen a stone which they think the initiate can only just lift, and this is put in the middle of the open space of the village and covered with a plaited mat called tapangau. The door of the gamal is covered with a plaited mat and a stick used for making the werewere sound is bound round with leaves and is put through a hole in the mat, the part which has passed through the mat being left uncovered. There is a small werewere stone inside the gamal and a man sitting there ready to use it. The candidate is then brought into the open space of the village, and from 20 to 40 fathoms of money are bound round his right arm, more round his neck and a large amount is put into a bag hung over his right shoulder, while two fathoms are hung on his outstretched left hand. The candidate and the assembled members then form a procession in which the senior member goes first, immediately followed by the candidate. The procession passes round the stakes five times, and then the candidate goes to the door of the gamal, and hangs on the end of the stick projecting from the door the two fathoms of money which he has been carrying on his left hand. He does not know the nature of the object on which he is putting the money. Some one then says to the last initiated member, " Let him buy the eye of the Tamate for you," and this man takes the money put on the stick which now belongs to him. The neophyte then goes to the stake nearest the gamal, and stands close to it clasping it with his hand,

and the man inside the gamal draws in the werewere stick which had been in the door. Some one then says to the first man to whom a pig had been given, " X, let him now be initiated to the Tamate ta Ureparaparal' the legend being that the Tamate liwoa came from this island. The man inside the gamal then makes the werewere sound, and the neophyte moves on to the next stake. The same man then says to the second man who had received a pig, " X, let him now be initiated to Kwat wasawasa for you." Again the man inside makes the werewere sound, and the neophyte moves on to the next stake. Then the third man is addressed, " Let him now break the wan" (a tall hat) and at the sound of the werewere the neophyte again moves on. Then to the fourth man, " X, let him be initiated to Kwaskwasoroa " (the name of a bird) and the neophyte moves to the fifth stake. To the fifth man is said, " X, let him be initiated to />ei ta Vava " (i.e., to the water of the Torres Islands)\ and the neophyte again moves on. Lastly the sixth man is addressed, "Let him be initiated to Tamate werewere'' and when he hears the werewere, the neophyte moves on to the sixth stake. The man who has spoken in the ceremony just described then addresses the neophyte as follows: "Look here, brother! I am going to speak to you about the rules of the salagoro. You were sulky about the Tamate ; you were a person of no consequence. That is finished. According to the custom of the salagoro it is your duty to keep clean the path, to fetch water and scrape coconuts. Should you go out and see a fellow-member of any other village, invite him, saying, ' Let us two go into the salagoro to eat.' When you enter to-day into the salagoro, should you see anything that your friends are doing, be careful that you do not touch."

After this speech the people go to the stone in the centre of the space, and some one snatches away the mat which had covered it and says, " Look here ! Formerly you were discontented, you thought it was something good, but look ! it is nothing but a stone." Then the man who had been in the gamal rushes out with the stick and makes the iverewere sound on the stone, which has been chosen so as to be suitable for the purpose, and thus is the great secret of the linge tamate or cry of the tamate revealed to the novice. The neophyte then lifts the stone, or if he cannot do so, his

• The black pigment used in the manufacture of the masks (see p. 92).

Sechision 103

father or mother's brother puts three or four fathoms of money on the stone to become the property of the person who carries it. All then go to the salagoro, the neophyte or his substitute carrying the stone. Before the people disperse there takes place the distribution of money called kere, if it has not been already done, the money used being that in the bag carried on the right shoulder of the neophyte. If the distribution of money has already taken place, the money in the bag is kept by the neophyte. As soon as the people have gone to the salagoro the island becomes free again and people can return to the village where the ceremony has taken place.

The newly initiated member should stay in the salagoro for a hundred days, and during this time he should not wash, and must give food to all who come. He may not leave the salagoro himself, but his friends provide food which is brought into the building by his father or mother's brother. It is the duty of the last member previously initiated to fetch water. A mat is laid in the salagoro on which the neophyte sleeps, and this must not be touched by anyone else during the hundred days or longer time that the probation lasts. During these hundred days and for a time afterwards the neophyte has to submit to all sorts of trial of his patience. He is fined a pig or money if he offends in any way against the rules of the salagoro, a cycas tree thrown down in front of the building being the sign that a fine has been inflicted. People come at any time expecting food and may inflict a fine if it is not ready. Often the people try to irritate the neophyte ; they may throw his best yams into the fire, break his knives or other possessions and if he utters any complaint, he is fined. Sometimes a man will have provided plenty of food, but his visitors will see that some special kind of food is not there and will demand it, and when this is procured and cooked they will only treat it as rubbish. Sometimes the neophyte will be set to scrape out coconuts through a hole so small that he makes his hands bleed in the attempt, and when he has succeeded, his tormentors only throw the food out of the door or rub it on his head. Even when the hundred days have come to an end a newly initiated member may be sent to fetch something, and after going with the utmost speed will return only to see the cycas tree thrown down because he has been too slow, and it was said that in such a case the man might be unable to control his tears, for

I04 Melanesian Society

the expenses of his initiation and previous fines may have completely exhausted his resources. The people who inflict these trials often come from another district, and a man who has been so treated will often revenge himself by imposing similar tasks when there is an initiation in the salagoro of his tormentors.

Sometimes an infant is initiated into the society in which case the father or maternal uncle has to go through the process of initiation. If the whole money is paid before or at the time of the initiation the seclusion is not necessary, but this is probably a modern innovation.

Throughout the hundred days following initiation, whether the neophyte is in the salagoro or not, there is a feast every five days when many assemble and eat in the salagoro. The feast is larger and more important on the fiftieth day, and at the end of the hundred days, or at any later period when the money has been fully paid, there takes place the most important feast called talasa, for which the food is cooked in the large oven in front of the salagoj'o. On this day for the first time the double werewere sound is made. This linge tamate or cry of the tamate is produced in the morning after the food has been put in the oven and covered over ; again when the oven is opened and again in the evening. When they hear the sound on this talasa day, the people say, " Let the tamate die."

The neophyte leaves the salagoro on the same day as the talasa feast or on the day after ; if on the same day, he becomes free after the feast has taken place. He is elaborately decorated ; a white powder is put in his hair, being puffed on his head from the mouth of the man who is decorating him. Young unopened leaves of the coconut are obtained, lozengeshaped holes are cut in them, and they are placed on a special kind of banana leaf which is glistening black, so that the black shows through the holes, and the leaf with its covering is bound on the head of the neophyte as a fillet. His head is decorated with three hibiscus flowers as in the entering ceremony. (If he has already been initiated into Tamate viov four flowers are used.) A belt is put round the waist and young coconut leaves are tied so that their ends hang down over either hip. Dots of charcoal are put on the cheeks and a long black streak above the eyebrows. Bleached pandanus is bound round the ankles, arms and neck. Four

End of Seclusion 105

single croton leaves are put in the fillet round the head, one at each side of the forehead, and two behind. Those already initiated decorate themselves in a similar manner but less elaborately. Two men then sit on the kwer or platform outside the salagoro and prepare to wereivere while a procession is formed. Each man in the procession carries a bamboo on his left shoulder, and in his right hand a stick of hibiscus wood, stripped of its bark so that it is glistening white. The order of the procession is very remarkable. The neophyte is in the middle with the last initiated man immediately before him. The man who has been longest a member heads the procession and is followed by the third in order of seniority. At the tail of the procession comes the member second in order of seniority preceded by the fourth and so on, so that the whole procession is like a curve concave upwards, in which the neophyte forms the lowest point, and the two senior members the two highest points\

The werewere begins before they leave the salagoro and when the sound is first heard, all shout and continue to call out as they walk, the werewere also continuing. They go to the village in which the women and uninitiated persons are assembled, and as soon as they reach it the shouting ceases and the procession passes through the village in complete silence, only broken by the sound of the iverewere coming from the distant salagoro. As soon as they have passed through the village, they again begin to shout, listening carefully to the werewere. When they hear the variations beginning they turn, passing through the village as before, and arranging their pace so as to get back to the salagoiv exactly as the wereivere comes to an end. If they do not arrive at the right moment they have to pay a fine to those not in the procession. We were told that owing to the possibility of this fine, many men are now refusing to take part in the procession. When the procession reaches the salagoro the ceremony is at an end.

All the food cooked for the talasa feast must be eaten ; if anyone allows this food to drop from his hand, he has to pay a heavy fine of a pig or money. Salagoro food is never given to the pigs ; if on any occasion there is more than can be eaten, it is put aside and recooked on another occasion, and if after several recookings it becomes dry and shrivelled, it is

' John Pantutun drew such a curve to illustrate this.

io6 Melanesian Society

burnt. The ashes are put in a special place from which they are only removed on an appointed day called kwong gir tuvus, when after a feast they are taken to the place where the ashes from the village are deposited.

After the talasa feast the skulls or jawbones of the pigs which have been killed are put up in trees or on stakes called palako, as memorials of the occasion.

I have already said that sometimes a person entered the salagoro never to reappear alive. This used to happen if a man entered the salagoro without having paid the necessary money. He was there cut off from any chance of earning money, and if he had no friends, or if his friends were poor, he might never gain the necessary amount and would have to live and die in the salagoro. A story is told of such a case in which a miraculous incident procured the release of the prisoner. A man in the salagoro of the Tasmate district had so few friends, and those so poor, that they barely brought him the food necessary for himself. One day he was sleeping in the salagoro when a black lizard bit his foot. At first it did not wake him fully, but after a time he realized what was happening and saw the lizard. He was surprised that the lizard showed no fear but only walked slowly away. As he was alone he followed the animal out of the building and into the bush, where it led him to a small tree-fern called ganegavi, round which it went several times. The man followed the lizard for a time, and then seized the plant which came up in his hands, showing below a flat stone. He raised the stone and found a large amount of shell-money, the string of which had rotted. He put back the stone and returned to the salagoro and began to prepare string. When men came they saw what he was doing and asked the reason, but he put them off with the pretext that he was only doing it to pass the time. When he was alone at night he went out and brought back the money, strung it, and hid it behind his pillow, the rule that no one might touch his sleeping place enabling him to do this without risk of discovery. When the money was strung he told the people to make ready for the talasa feast which takes place at the end of the period of seclusion. He had been in the salao-oro so lonQf that at first the people would not take him seriously, but when he persisted they prepared the food for the feast, telling him that if he was deceiving them and did not produce the necessary money, he

story of Tamate liwoa 107

should be hanged, the customary mode of inflicting death in the salagoro. The feast took place, and the man was able to give (kakere) the necessary amount, and had still a little over for his own use. The hero of the story was known to one old man still living on Mota.

The following story of the Tamate liwoa of Petanpatapata (a place on Vanua Lava) was told to Mr Durrad.

There was a very lovely girl, a tnotar^. A chief of Merlav heard about her and came bringing pigs, rawe"- and money. The relatives of the girl saw the canoe approaching the landing-place. Then the chief said to the girl's father, " I am a man of Merlav and I have heard of your daughter, the girl motar, and I have come to purchase her with these pigs and this money." Then the girl's father went to ask her, but she did not want the Merlav chief and so he returned.

Then a chief of Gaua^ asked the Merlav chief, "Whence come you?" and he repUed, "We went about the motar of Petanpatapata to buy her, but she was not willing." And he replied, "I myself will go." So he took a well tusked boar and a rawe with tusks and a hundred fathoms of money and he set out with his companions and they blew several trumpets together when they arrived, and went up. The father of the woman said, " How is it that you have come here ? " And the chief replied, "I have heard about your daughter, and I have come hither to buy her." The woman's father went to ask his daughter, but she was not willing. So the chief of Gaua returned.

Then a chief of Mota heard about her, and he made ready pigs and rawe and money, and they rowed and landed at the landing-place. The father said, "How is it that you have come here?" And he replied, "I heard about your daughter, the motar, and I have come hither to buy her." And the woman's father replied, " Very well, I will ask her first." So he came and asked her, but the motar was unwilling. Then they went home again.

Then a chief of Motlav heard about her, and he prepared pigs and rawe and money, and they set off and went ashore at the landing-place. The woman's father advanced to meet them and asked, " How is it with you ? " And the Motlav chief replied, " I heard about your daughter, the motar, and I have come hither to buy her." And he answered, " Let me ask her first." And he went and inquired, but the motar was unwilling, and so the Motlav man returned.

Then a chief of Rowa heard about her and said, " I myself will try," and so they collected pigs and rawe and money and set off, and went ashore at the landing-place. And the woman's father said, " How is it that you have come hither ? " And the Rowa chief replied, " I have heard of your daughter, the motar, and I have come to bring her purchase-money." He answered, "Very well, I will ask her first." So he asked his daughter and she was unwilling. Then he returned.

Then a chief of Ureparapara heard about her and he made ready pigs and rawe and money, and set forth and landed at the landing-place. The

^ Mr Durrad was uncertain whether this was a personal name or a' name for a woman of high rank, i.e. of high rank in the women's Sukwe (see p. 139). I shall assume that it is the latter.

2 A special kind of pig (said to be hermaphrodite) with large tusks.

2 Santa Maria. Gaua or Gog is properly a district of this island.

io8 Melanesimi Society

father of the motar advanced to meet him and asked him, "What is it?" And the chief repHed, " I have heard about your daughter, the motar, and I have brought hither the purchase-money for her." And he answered, "Stay a moment till I listen to her first." So he asked her, but she was unwilling.

Then a chief of Ravenga ' heard of her, and made ready pigs and rawe and money, and they set out and landed. And the father of the 7noiar said, " How is it that you are come hither ? " And the chief of Ravenga replied, " I have heard of your daughter, and I have come bringing her purchase money." And the father said, " Very well, I will listen to her first." So he asked her, but she was unwilling and they returned.

Then a chief of Kwakea* heard of her and he prepared pigs and rawe and money, and they set forth and landed. And the father of the motar went to meet them and asked them, " What is this ? " And the chief of Kwakea answered, "I have heard of your daughter, the motar, and I have brought hither purchase-money for her." And her father said, " Let me ask her first." So he asked the motar and the motar was unwilling. Then he returned.

So those three remained. The father and mother of the motar went always to work ; the motar did not go about at all, but remained all day in the village weaving I Thus she did continually. But one day when her father and mother were gone out, the motar was weaving, and was seated outside at her weaving in the middle of the day. And as she was weaving she heard a noise and stopped her weaving and listened, and a second time the noise came, and she heard it yet again a third time, and saw as it were a rainbow drawn above her and the light of it upon the opening of a pit that was open. She threw away her weaving and ran and saw Wetmatliwo sinking down. The ??iotar seized hold of his head and said to him, " Whence come you?" And Wetmatliwo said, "I abide here." And the motar replied, "Come hither, for my husband are you. I have refused all those that brought hither purchase-money for I did not want them, but now I have found you, you indeed are my husband." Then Wetmatliwo answered, "Very well, go sweep the village square and chop a coconut leaf, and screen off a portion of the house at the end, and dig and fix a wooden drumming-board in the village square and decorate the square." So the motar went and did exactly what Wetmatliwo told her to do. And the motar took a bamboo drumming-stick and sat down by the drumming-board, and struck the drumming-board and sang a song, and Tamateliwoa^ danced to the tune. He went on and on and on and on, right into the house, and both of them entered the house and lay together and the motar did not go out again. Then her mother and her father returned, and when they had not yet entered the house, the motar rushed out quickly to meet them and said to her mother, " Let me take down the basket from you." She rejoiced over them, and her father and her mother spoke to one another and said, " What is this that she rejoices thus over us? She does not do so every day." Then her father and mother entered the house and saw the room screened off. And they said, " What is this room ? " And the motar said, " O Father and Mother, I have found my husband." And they said, " Indeed?" And the motar answered, "Yes." Then her father and mother said, "We want to see our ktvaligaJ"

1 A small island near the east coast of Vanua Lava.

'^ A pair of small islands on the east coast of Vanua Lava.

•* ? plaiting.

'' Note here the change of name from Wetmatliwo to Tamateliwoa.

story of Tamate liwoa 109

And the motar replied, "Very well." Then she went to the partition and lifted it up, and his glory as a great light filled the house. And they said, "It is enough, in this beautiful youth you have found our kwaltga.'" So it was that they two lived in that room and lay together and did not go out. The motar lay always with Wetmatliwo and clasped him, and did not turn away from him at all.

Then Weav (that is " fire ") ' saw that the motar did not go out at all any more and he inquired about it. One night at midnight he went quietly in to see the motar and he found her and Wetmatliwo asleep. Then said he, " Yes ! I have found out, to be sure, that the motar does not go out because she and Wetmatliwo sleep." So he sought how he might injure them both. Then he went softly in and lay quietly behind the motar in the night, for he is Warm Fire, and he made her warm and she felt him and lay round towards Weav and embraced him, and in the morning he got up and went out from them. He did this every night. Then Wetmatliwo thought of it, and on another night he said, " Let us remove the bed-place," and they removed it^ And on another night those two were sleeping and Wetmatliwo failed to sleep, but lay still and felt the motar turn round and embrace Weav and they embraced one another. And when they were already asleep Wetmatliwo got up and saw the motar and Weav lying together and embracing one another. He did not say anything, but then he knew for certain why the motar did not think of him any more, because Weav cleaved to her. Then one day later her father and mother went out to the garden-ground and Wetmatliwo said to the motar, "Go and sweep the village square as you swept it for me before, and chant my song that I may dance to it." Then the motar swept the village square and sat down and chanted his song, and Wetmatliwo danced to it. (He had already determined to forsake the motar.) And he went back, back, back, back. He danced towards the motar, and he danced back, and he went back so that the decoration was divided from him. He passed on and ran and ran as far as the pit from which he had come out, and he stood in it and sank down in it. Then the motar saw him sinking in the pit and she ran to him, but only the top of his head still remained, and she stretched her arms to draw him out. But Wetmatliwo said, "Let me go." And the motar replied, "Why are you forsaking me?" And Wetmatliwo answered, "Alas! you have deceived me. I had supposed you loved me. But no! For indeed you cleave to my friend ; he is a good man, even Weav, and I am evil and he my friend is better than L" Then the motar ran for the pig and the rawe, and gave them to Wetmatliwo, and he took the pig and she thought he would return to her ; but no ! Then Wetmatliwo sang this song to the motar: —

" Vanavanoi, waise va ve, Rowir take hold of thy hand Rowir take hold of thy face Rowir take hold of thy foot I, WetmatUwo, born of Panoi, You two have brought Weav hither and have abandoned me^"

When he had sung this song he sank down into the pit. And. the motar

• Said to be the spirit or personification of fire.

2 Mr Durrad was uncertain about the correct translation of this passage. ^ The first line of this song which forms the beginning of many songs was said to be untranslatable. None of the words belong to colloquial Mota.

no Melanesian Society

seized the top of his head which broke off and he sank down into Panoi. For this reason if anyone wishes to be initiated into the Tamate liwoa he must always first give a pig and a rawe. It is not permitted that he shall give only a pig and not a rawe.

And the black volcanic stone that stands now at Petanpatapata, so very tall, is called Tamate liwoa. The spray dashes over it, and they call it the decoration of Tamate Ikvoa.

List of Tamate Societies.

When I was in Melanesia I obtained fragmentary accounts of certain other societies, and Mr Durrad has since sent me a list of the Tamate societies of Mota with some account of their badges and masks. All this information was obtained from John Pantutun, and when no information is given about the mask or badge, this is usually due to the fact that John was not a member of the society in question. The societies are grouped in the three classes mentioned on p. 87, and I shall follow John's order, beginning with the less important societies, into which even young boys who have not yet eaten in the gamal can be initiated.

Initiation into many of these is the occasion of a kolekole performance, but it seems clear that these bodies into which children are initiated are regarded definitely as Tamate societies ; the hats or masks are called tamate, and the badges are used to protect property.

I. The Tajnate Matawonowono.

1. Kwasanomenamena. Very small children, even those just born, can enter this. The hat is made of the leaf of a breadfruit tree, split and tied in a peculiar way. Entrance can be bought very cheaply, a pipe full of tobacco even being enough.

Badge. — Leaf of breadfruit tree.

2. Kwat ngai. Can be entered while very young and can be bought as cheaply as the above. The hat is made of the leaf of a chestnut tree {make). It is rolled up in the form of a cone and pinned together, and bound round with sago-palm leaves. Models of these hats can be worn as playthings by the children in the house.

Badge. — Leaf of the tavisoviso.

3. Marawa (spider). May be entered by a baby in arms and bought with a short piece of money, four or five inches being enough. A kolekole is held in which the father carries his child. If a hat is worn, it is tied tightly on the child's head lest it should fall to the ground (see p. 91). If a hat is not worn, a half coconut shell is decorated as a hat and is carried by the baby or by his father.

Badge. — A very large croton. Fun is made of its bigness.

4. Sarasara (a fish). Bought with a few inches of money. Badge. — Leaf of the sarasara tree.

Tamate Societies 1 1 1

5. Takwagato (a fish). Bought with a few inches of money. (A girl named Romatewelgan, child of a great chief in Motlav, was once initiated into this Tamate.^

Badge. — A croton, short and white.

6. Funia. It is said that in old times the members built a shed, and all who had been initiated into this Tafuate went inside and the door was tighdy fastened. Some one outside then set fire to the shed which was burned to ashes, but the men inside had disappeared and were presently seen coming along the path.

Another story is that a pit was dug in the open ground in the village and filled with water. Fish appeared in it, and the men stood round and shot them with arrows,

Badge. — A very yellow croton.

7. Seperevangona (a little fish). Purchased with a few inches of money. There is no hat for this Tamate, but an image of the little fish is made and carried on a cycas leaf. To be initiated in this Tatnate is a very simple matter. If a few men are sitting in the bush talking, they will perhaps suddenly decide that a little boy with them shall be initiated. Some one makes an image of the little fish and it is carried into the village. If any one of the men is not already a member they will initiate him at the same time, even though he may be a grown man.

Badge ?

8. Wopsapaso paso. Bought with about ten green coconuts iyusa). Badge ?

9. Wovusmalrhal. Bought with coconuts {mal) at a stage later than the vusa.

10. Pepetakwatakwa (a little fish that resembles a withered leaf). Bought with a short bit of money.

Badge ?

11. Mes (parrot). Bought with half a fathom of money.

Badge. Two crotons, one memea (red), the other turturuga (green).

12. Wovele (a nut tree)^ Bought with half a fathom of money or less.

13. Roropei (dragon fly). Bought with three-quarters of a fathom of money. Clement made it for John's little boy Arthur recendy while out in the bush. John gave Arthur the money and Arthur gave it to Clement. In this case three pigeons were given in return by Clement to Arthur, these taking the place of the pig in more important societies. Clement made the tamate (not a hat but an image of the roropei) and laid it on a cycas leaf and then all those who were already members shouted. A boy named Barnabas carried the tamate. When they were near the village they shouted the name of the tamate, and then remained still while Barnabas danced with the tamate in the middle of the village. The women had heard of their approach and had swept the dancing ground ^ Afterwards the tamate was carried to another village near and burnt. It was said, "<? tamate me mate" "the tamate died."

14. Kwatnian (a bird, a black-bodied, red-headed honey-eater). Bought with three-quarters of a fathom of money. An imitation of the flower of the tvotaga tree is made and an image of the bird is hung to it, as the living bird hangs on the flower. Initiation is as in the case of Tamate roropei.

Badge. — A croton from Vanua Lava.

' Properly the fruit {wo) of the vele, a barringtonia.

'^ It is doubtful whether they left the village; presumably they did so.

15. Tasia — female of the Kwatman bird. Initiation fees and ceremony exactly as for the Roropei and Kwatman.

16. Tawene av (a live ember). Bought with three-quarters of a fathom of money. There is no hat for this Tamate.

John is not a member, and could not explain the decorations for the dance which takes place in the night. He says that from the waist to the foot on the front of each leg is seen a strip of fire ; also round the waist and down the front of each arm.

Badge. — A firebrand.

17. Worormelau. Bought with green coconuts (zu/sa). There is no hat. One man will not be initiated by himself but a number become members at one time. The initiates sit together by themselves in the bush and cover their eyes. Those who are already initiated remain in hiding and taking sticks, dirt, stinking water, they approach the initiates with a shout. They walk round them and pelt them with the things they have brought, and if the initiates are much hurt they will run away and escaped

18. Mwara (dove). Bought with three-quarters of a fathom of money. A hat is made and an image of a dove is put on the top. It is brought into the village as is the roropei.

Badge. — A short broad red leaf (from Vanua Lava).

19. Teremawora. Bought with a short piece of money, about threequarters of a fathom. The proceedings of initiation are the same as for the Roropei. There is no badge, but a hat the nature of which was unknown.

20. Wurene. Those initiated beat the uninitiated, both men and women. Nature of hat and badge not known.

21. Wotomava. The hat is worn only in the night, the dance not taking place by day.

22. Wopoava. Bought with three-quarters of a fathom of money. There is a hat used when the dance takes place.

Badge. — Leaf of the ora tree.

23. IVoninpei (a kind of water snake). An object is made to hold in the left hand. No badge.

24. Womasanara. This is the name of a part of Vanua Lava opposite Mota where the Tamate may have been " born " {wota). There is a hat and the dance takes place at night. Initiation costs three-quarters of a fathom of money.

Badge. — A white croton.

II. The Tamate talo maea.

I. Mweretang. This is one of the societies at which formerly only one man would he initiated at a time : now several can enter at once. The hat is "of great beauty." When a man is initiated he stays in the meretang salagoro, and the nieretang is played, after which the man goes out wearing the hat, and the irieretang ceases. Then later on some men who have already been initiated will say, " Now we will go and eat some green coconuts in the meretang salagoro.'' All the initiated go together and blow the meretang iox a long while. When they have eaten, the playing of the iheretang will cease.

This society has several badges, viz., four crotons, a red, a white, a long and striped and a very short variety ; the leaf of the rangorango tree ; a reed (toga) with the leafy top still on.

' Badge not known.

The Kwat Society 1 1 3

2. Tiwia (a small bird that lives on the shore). This is a fellow society to the Mweretang. A man who joins the Mweretang can eat with a man of Tiwia.

Badge. — A speckled croton.

3. Wosus. Bought with pigs and much money. This society, formerly kerkerei, can now be entered by a number of men at once. The dance takes place at night, probably to conceal the method of manufacture of the hat, which is made of the young white leaves of the coconut, plaited over a coconut which is then withdrawn. It was not known where the tamate "was born and grew up."

Badge. — A large long red croton.

4. Wotngerevno. An expensive initiation though the expense is not as great as for the Mweretang or Wosus. There is a hat and the dance takes place by day.

Badge. — Crotons, white and turturuga.

5. Wotletansil John knows nothing of this Tamate. No initiation has taken place in his memory, nor even since his father was quite a youth.

Badge. — A small red croton.

6. Kwat. A very important Tamate and pigs and 40 or 50 fathoms of money are required to purchase initiation. It is possible for one man to become a member by himself, but it is not considered correct. Membership of the Kwat is much coveted as the dance is specially fine. The decorations upon the top of the hat resemble a cock's comb.

Badges. — Many crotons including silsiliga (black) and viemea (red), and a reed with the leafy top pinched off which may easily be mistaken at a distance for the badge of the Mweretang, the only difference being that the leafy top is taken off in one case and not in the other.

The Kwat has many divisions or includes many minor societies, but separate initiation into them is not required Whoever is initiated into the Kwat is ipso facto a member of all its subsidiary societies. They are named as follows :

I. O uwa (the turtle) : 2. O ihae (the sea-snake), called both O mae talo salagoro and O mae talo kwat : 3. 0 lele (a flat-fish) : 4. 0 pagoa (shark) : 5. O wumeto (bowl) : 6. O kwat maremare (a fish) : 7. O takwalekwoe (a large fish with a big mouth) : 8. O ut (a long fish that lives in the deep sea). Of these the shark {O pagoa) is the most important but not preeminently so. It has no special sanctity and anyone can mention its name, but he would speak respectfully and not shout it out loud.

When the initiating ceremony takes place, which can happen at any gamal, a very high fence is built round the gamal to prevent the uninitiated from seeing the proceedings. The door of the fence is spoken of as the vale pagoa or the shark's mouth. Those who are to become members remain in seclusion within the fence, and blacken themselves all over every day with soot and ashes. They remain a long while inside and are taught a very difficult dance which they perform to the accompaniment of a little bamboo drum. Whenever they are not practising the dance, they must lie or sit down. Those who are already members of the Kwat go out daily for food and chase anyone they see, and catch and bind them and beat them with nettles. They sing as they go, and burst open hollowed leaves to make a noise to warn people out of the way. But in the garden-ground they will creep by stealth through the bush, and if they find anyone, will seize, bind and beat him. When he is bound one of the party sings a chant. If the man who is caught

is submissive, they will deal gently with him, but not if he struggles to escape. John says he helped to catch his brother Henry. John went ahead of the party and persuaded Henry to help him to gather some breadfruit, and while so occupied John pinned him with his arms and the rest came up and bound him.

When the new members go out, they do not organise a separate dance for themselves, but go to a kolekole and dance there. On the first occasion they all wear very tall black hats of the same pattern and no other decorations. At a second kolekole they dance again but without any hats or decorations. At a third kolekole they dance and wear their special hats. Each man can choose the hat of any one of the subsidiary divisions, and of these the pagoa (shark) is perhaps the most popular. The hats are shaped to represent the various animals, fish, etc., after which they are named. In recent times they have changed the wumeto (bowl) into a model of a boat.

7. Mautereteremake (the bird pecks the chestnut). Initiation requires much money and pigs. The decoration of the hat is like the head of a bird. John was initiated when quite small but does not know much about it, for no one has been initiated since.

Badge. — A long red and white croton.

8. Viov or Oviov. There is some uncertainty about the correct name of this society. It was sometimes called Viov and sometimes Oviov or Oviovi, but the first letter of the latter forms is the article. The Tamate is very important and was said to be the most important and expensive of all the societies. It is one of those which can only be entered by one man at a time, and only by those whose relatives are powerful as well as rich. What is meant by calling it the most important is doubtful ; it did not seem to be essential for progress in the higher ranks of the Sukwe as is the case with Tamate liwoa, but its membership is certainly highly valued. It was said that the men of Viov and Liwoa are rivals and compete with one another in all things. A man can belong to both, but it was said that the Vioz> would take the first place in the mind of him who belongs to both. Viov is a male Tainate, having a tall mask with a long beard. Within it there are minor societies into which a man can be initiated separately if he wishes, but it appeared that this separate initiation was not necessary.

The following are the names of the sub-societies :

Lele. A flat-fish.

Akatakwanag. A Tatnate which chases people. Badge. — A shell of the cuttle-fish stuck on a reed.

Tawastarara. Badge. — A sprig of the shrub of this name.

Kwasapule. A pigeon. No badge.

Ura (crayfish). A man cannot be initiated into this unless he is also a member of others. Badge. — The antennae of the crayfish.

Tolama. A Tamate which chases people. Badge doubtful.

There is another society also connected in some way with Viov called Togotil. It is also called valun 0 Viov or "the companion of Viov'' ; its mask has four eyes, and its two badges are the leaves of the nanara and kwatkwol trees.

In addition to the badges of these subsidiary societies, there are also badges of Viov as a whole. These are: — the womarakwarakwa, a bright orange fruit like a tomato, and when used as a badge of the Viov, several, at any rate more than four, are spitted on a reed which is stuck in the

The Viov and Pir Societies 1 1 5

ground as a soloi ; the wowarwar, another red fruit ; the skin of the breadfruit ; a stem of hibiscus with branches left on but stripped of its bark ; the leaves of the mamaota, wura and noto trees, the kalato or nettle-tree and the las, a plant even more irritating and poisonous than the nettle.

In addition to these a Viw man may decorate himself with a red croton called kariura, but only if he has also been initiated into Tamate liwoa. If he belongs only to Viov he may not touch this leaf, and the objection to his doing so was given in the form "(? nonomia ape Kwatgato" i.e. thought about Kwatgato. The kariura seemed to stand in some relation to ura, the crayfish.

In decorations Viov is represented by the lozenge-shaped figure with two eyes, shown in PI. IX.

9. Pir. A female Tamate with a short hat. Pir means terrible, anything with a terrible appearance being called pirsus.

Badge. — An oblong red fruit.

The Viov and Pir societies are closely associated. A man who has been initiated into Tamate viov can eat from the oven of a member of the Tamate pir and vice versa. The Viov is regarded as male and the Pir as female. This may refer to nothing more than the shape of their hats, but I was also told that Pir is regarded as the wife of Viov. This information was given to me, but was not repeated to Mr Durrad when he was obtaining an independent account of the two societies from the same man. It may be only a fancy of our informant, but it is possible that it has some significance. Both Mr Durrad and I were told tales concerning the relation between the two societies which agree in their main details. This story is interesting in that it provides an account of the origin of some of the badges of the two societies.

Before Viov was born the skin of the breadfruit (one of the badges of Viov) was not important. Tariaka, a man of Vanua Lava, took Viov for his own, while a friend (whose name could not be remembered) took Pir. Tariaka and his people one day arranged a dance, and the friend, the owner of Pir, was roasting a breadfruit when the dancers appeared. He was so intent on watching the dance that he forgot the breadfruit, and when the dance had finished, he found on looking at the breadfruit that it had been badly burnt. He called to Tariaka who was inside the tamate, and showed him the burnt breadfruit, and said to him, ^'■Pulaih o tamate nasasana Pegveng," "Your tamate is named Pegveng." Pegveng'xs a Vanua Lava word, and it is now a salagoro word for burnt breadfruit {peg, breadfruit, and veng, burnt) and only members of the Viot^ society may use the word veng for burnt, and even they only use it among themselves. A man has only to say that a breadfruit is vingo, the ordinary word for " burnt," and no one may eat it but a member of the Viov.

The story continues that after a time the friend of Tariaka who took Pir, arranged a dance and invited Tariaka to see it. On the appointed day Tariaka was about to make a pudding. He had already roasted and pounded his breadfruit, and had broken and scraped some nuts (ngai, Canarium nuts) and after wrapping them in some leaves, had put them on the fire when Pir came out. Tariaka stared at Pir and forgot his nuts and when the dance was finished, he found that his nuts were badly burnt. He called to his friend within Pir, "Look this way; your tamate is named Ngirveng," ngir being the salagoro word for the nuts. Then the two friends took conch-shells ; Viov took fifty and Pir a hundred, and now everyone who is initiated into Viov has to bring fifty shells, and a man initiated into Pir has to bring a hundred. It

8—2

ii6 Melanesian Society

was said also that Fegveng might be used instead of Viov as the name of the Tamate, and Ngirveng instead of Pir.

Viov is spoken of as the property of Tariaka and Pir as the property of his friend.

10. Manvusvusmala. A society about which nothing was known.

11, Mantelel. Takes its name from a shore-bird, its hat having the form of this bird.

Badge. — A very small croton.

III. The Tamate of the Salagoro.

1. Tamate liwoa or werewere. This is the society which has already been fully considered. It has many badges which have been given on p. 96. There are connected with it certain minor societies : Wis, Kwatwasawasa, Saka and Serekor, all of which are Tamate which chase people. Tamate saka is perhaps a separate Tamate, for it has its special badges, a small red croton and a sprig of a long grass.

2. Tamate nivat (the four Tamate). This includes the four following societies :

(i) Menmenide, a male Tamate.

Badges. — Fruit of the marakwarakwa (as for the Vioi^ but green and unripe, and the leaf of the kwagkwage.

(ii) Kwatgatotapanoi, a female tatnate (the explanation of the word is "(9 ura ta Panoi, we kurkur aneane 0 tanun,^' "The crayfish of PanoiS a great devourer of men."

Badges. — (a) The ripe fruit of the marakwarakwa as for the Viov, but only four are used instead of the larger number of Tamate viov; they are spitted at regular intervals on a reed, which is stuck upright in the ground ; {b) leaves of the naraga, nutmeg ; {c) the magoto (a grass) ; {d) a yellow flower ; {e) several kinds of crotons — red, white, silsiliga and turturuga.

(iii) Novnau, a female Tamate.

Badges. — {a) Two or three fruits of the marakwarakwa spitted on a reed ; {p) leaf of a kind of naraga different from that of the Kwatgatotapanoi; {c) several kinds of croton.

(iv) Pepe (a fish), a female Tamate.

Badges. — {a) One fruit of the fnarakwarakwa ; {b) leaf of the pepeure shrub; {c) several kinds of crotons but not a croton of the silsiliga kind.

There is one badge for the Tamate nivat as a whole, viz. a croton called the tamatpaso.

These four societies grouped together as the Tamate nivat have one salagoro which in Vanua Lava is merely the clearing in the bush to which I have already referred (p. 89). The association between the four is also shown by the fact that members may eat at each others' ovens with the exception of Tamate pepe, the members of which can only eat at their own. These four Tamate have different dispositions ; Menmenule is always fighting, and never appears without a bow and arrow in his hand ; Kwatgatotapanoi chases and eats people, while lYovnau and Pepe are always gentle. There are special customs connected with the fish from which the Tamate pepe takes its name. If a party of people including a member or members of this society are fishing and one of the party catches Sipepe fish, the people will

' Panoi is the abode of the dead.

Tamate of the Salagoro 1 1 7

feel embarrassed, and if it has been caught by others, the fish will be given to the member or members of Tamate pepe and they will eat it. If no members of this society are present, there would be no embarrassment and the fish would be eaten by those who caught it. All four societies have the marakwarakwa fruit as a badge but used in different ways, Menmetiule people using it unripe. The following story gives the origin of this custom (it was all that could be remembered of a long tale) : Three women were decorating themselves with the marakwarakwa fruit when Menmenule arrived from an expedition. He asked for some of the fruit with which to decorate himself, but the women had used all those brightly coloured, and so he had to use the green unripe fruits which are now his badge. He also wished to adorn himself with pigments, but the women had used all but the black, with which Menmenule had to content himself, and he painted his whole body black. This story comes from the island of Ureparapara, and probably the whole society had this place of origin.

3. Gasoso. A society which can only be entered by one man at a time. Badges. — Crotons, red, white, yellow (but not silsiliga) and a leaf of

another kind.

4. Tamatpilagi, the companion society to the Gasoso, the members of both eating together. The pilagi is a bird and is speckled, and the hat of the society is speckled in imitation of it.

Badge. — A white croton with a spot of red at the base.

5. Lakto. The nature and construction of the hat is concealed in this word, though the uninitiated do not realise it. A reed (togo—oi which the last syllable is dropped) is split open and the pith taken out and cut longitudinally in strips. These are bound together {lak-tata is an alternative word) to make the hat. The uninitiated look at the hat and wonder, for they cannot understand of what material or in what manner it is made.

Badges. — Crotons, white and red.

6. Mesigolo.

Badges. — White and red crotons, leaves of the pawiira and tawora, and the stem of the hibiscus as for the Tamate liwoa but shorter and thicker.

7. Igtawosor, a fish of a place on Vanua Lava (or Ureparapara ?). Badge. — A red croton.

John was initiated a little while ago but did not wear the hat. It is really a Vanua Lava Tamate, and only the people of this island know how to manufacture the hat properly.

8. Igtavalpei {iga tavala pet, literally " the fish of the other side of the water "). A Vanua Lava Tajnate. The hat, which is female, is decorated with a fish.

Badge. — Several kinds of croton.

9. Iglesao, a fish of Lesao, a place on Vanua Lava where the Tamate "was bom."

Badges. — Crotons, white and red.

10. Seglama (a long snouted fish: also the albatross (?), literally "the kingfisher of the sea " stgo lama).

The hat of this society follows the bird in shape and not the fish. The society can only be entered by one man at a time. Robert Pantutun^ and Matthias are the only two men now alive who belong to it. No one living has ever made the hat but a man named Carpenter saw one in his youth

^ This man has since died.

ii8 Melanesian Society

and would make it if anyone wished to be initiated, but no one has yet decided to do so.

Badge. — A red croton.

Seglama has a companion Tamate the name of which John had forgotten ; the members of this society and of Seglama can eat together.

11. Igagangan (a fish). The hat is after the fish in pattern. Matthias is the only living member of this society. It cannot be entered by more than one man at a time.

Badge. — A red croton.

12. Manlulid. bird). Abbreviated from manu luium — "a beautiful bird." Badges. — Red and white crotons.

13. Wotrono. John has been initiated into this but did not wear the hat. He was initiated by James {tiro ?nun James) and James did not know how to make the hat, and did not pay Carpenter to make one.

Badge. — A red croton.

14. Mansapruk (a sea bird). The hat follows the bird in shape. Badges. — Red and white crotons.

15. Malop.

Badges. — Red and white crotons and a green coconut leaf stuck upright in the ground.

16. Wiriu — a mischievous Tamate {Tamate kiskislag) the members of which shoot people. John's father and Carpenter alone are left, but John wishes to be initiated and has made arrangements about it.

1 7. Tamat wopewu — Ta?nate of the fruit of the pewu, a root which people eat in times of famine.

Badge. — Fruit oi\hepeivu stuck on a reed.

18. Igtamategmal. Interpreted by John as Iga talo matai gamal, the fish of the door of the gamal, perhaps indicating that the Tamate "was born " at the door of the gamal.

Badge. — A red croton.

19. Lastur. The hat was not worn at recent initiations. Badge. — A red croton.

20. Gapilwatiga. A flower. Badge. — A leaf of the gapgapil.

21. Gasuwe (rat). An imitation-rat is put on the top of the hat, but the rat is regarded as a common thing, and the word can be spoken or the rat killed by anybody — man or woman. It is not in any way or under any condition sacred.

Badge. — A small white croton.

22. Sigo (kingfisher). The hat is made to resemble the head of the bird. The kingfisher is not really sacred, though people do not shout the name with a loud voice but utter it in moderate tones.

Badges.— A piece of a white ants' nest, stuck on the branch of a tree as a soloi, and a white croton.

23. Langtanua (a species of fly, rather long). An image of the fly is made on the hat.

Badges. — White and red crotons.

24. Soro. Badges. — Red and white crotons.

25. Wareimi. Badge. — A red croton.

26. Igtangeresota (a fish). The face of the mask is modelled after the face of the fish.

Badge — A white croton.

Tamate Societies 1 1 9

27. Manterewut (O manu we teretere o wut, " the bird pecks the louse "). The members of this society have been chary of allowing others to initiate.

Badge. — Red and white crotons.

28. Wewe (a shell). This is a female Tamate. John thinks the shell may be painted on the side of the hat, but he does not know as he has not been initiated. This society has many members at present.

Badge. — A beautiful red croton.

29. Igatalotung ("fish of the pool "). A fair number of people are members of this society. John has not seen the hat or the badge, a croton which grows on Vanua Lava so that the Mota people have to get it from there.

30. Wotlewona {Me wota alo one, " born in the sand "). A woman was sitting on the sand making drawings, and she drew a design which a man saw and used as the pattern for the hat. A little while ago the Mota people made one of these hats, and went about with it (me ngarag) on Vanua Lava.

Badges. — Red and white crotons.

31. Wotlomal {Me wota alo gamal, " born in the gamal"). Badge. — A red croton.

32. Walitira. The croton for the badge does not grow on Mota but on Vanua Lava.

2,2,- Singiowo, with a withered banana leaf as badge.

34. Wer (a bird that lives in the swamp) has a hat which is tall and male. Badge not known.

35. Mala (hawk). A year ago some people were initiated into this Tamate on Vanua Lava. It is not fitting that youth should belong to this society because of its importance.

Badge. — A long red croton which grows on Vanua Lava.

Two other societies about which nothing was known are Igon and Tokove.

A large number of the Tamate societies take their names from animals and in these cases the hats or other sacred objects either represent these animals or have representations of the animals upon them. In a few cases as that of the gasuwe or rat, the animal so connected with a Tamate society has no special sanctity either to the members of the society or to anyone else, but frequently the animals from which the societies are named may be regarded as definitely sacred. What seems, however, to be clear is that the use of such an animal as food or otherwise is only forbidden to those who do not belong to the society, the prohibition ceasing after initiation.

There is, however, one definite sign of respect paid to these animals. The members of a society will not utter the name of the bird or fish connected with it. Thus, a member of the Tamate seglama will not call the seglama by this name but will speak of it as manu talo lama, the bird of the sea. I do not know whether this periphrasis would be used by men who are not members, but women would not utter such names at all except under their breath.

I20 Melanesian Society

The use of animals such as fish from which Tamate societies take their names should probably be limited to the members of the society, though as has been seen in the case of the pepe fish (p. 1 1 7), others will eat the fish if no members of the pepe society are present. There is evidently a close resemblance with the rules affecting the use of the badges of the societies which is stricdy limited to the members of the societies (see p. 93).

Mr Durrad has sent to me the following list of animals and plants the use of which is in Motlav limited to members of certain societies: —

The pepe fish may be eaten only by members of the Tamate pepe and Tamate kwat and kwatgato. The til or sword-fish may only be eaten by the members of the society of that name. Two fishes called talasar and tahanwau may only be eaten by the members of societies named after them and of Tamate kwatgato and novnau.

A kind of taro [kweta) called tamate liwoa may only be eaten by members of the great Ta^nate. The bird called menme7iule may only be eaten by the members of that society. The owl is not eaten by anyone but its feathers may only be worn as a decoration by members of the Tamate wis, though they may also be worn by women on whose behalf a kolekole in connection with a house has been performed (see p. 151). In this case women in general can use an object which is forbidden to men except those who have been initiated into a certain society, and this would seem to be part of a more or less general rule, for the different kinds of fish the use of which is limited to the members of certain societies may be eaten by women of full age, though not by unmarried women or girls. Thus, the wife of a man who does not belong to the Tamate pepe may eat the pepe fish while her husband may not eat it.

The following story illustrates certain features of the attitude towards the less important Ta^nate societies. A little while ago a youth named Clement tried to make the hat of one of the Tamate matawonowo7io in preparation for a great dance. When it was half made he felt the task was beyond him and went to the salagoro to ask help of Carpenter, the most expert hat-maker of the island. It so happened that though Carpenter belongs to scores of societies, he was not a member of the particular society whose hat Clement

Tamate Societies 121

was making and he refused his help. Clement then asked Carpenter to go out and wait for him. He went out and Clement followed, and asked him to become a member of the society so that he could make the hat. Carpenter consented though only half-heartedly, for he was afraid of the ridicule of the other men. Clement then gave Carpenter the sum of money necessary for initiation and told him to tiro to a small boy standing by. He did so and made the hat. When Carpenter left the salagoro, there were many men there making hats for the dance and their curiosity had been aroused by his departure and when he returned he was accused of having made some mistake and rendered himself liable to a fine. Carpenter denied it but no one believed him and after much heckling he confessed what had happened and was ridiculed unmercifully for having had to ti7^o to a small boy for so insignificant a Tamate.

All the societies included in the foregoing list are now to be found in the island of Mota but many of them have been introduced from elsewhere. Most of these introduced societies have come from Vanua Lava and the badges of some of these grow only on that island, while in some cases the masks or other objects can only be made by Vanua Lava people. There is some evidence that other societies have been introduced from elsewhere, though it may be only that special features have been borrowed. Thus the Tamate liwoa is said to have come from Ureparapara' while the black pigment which adorns the hats and is so much admired is called pei ta Vava or "water of the Torres Islands" and may have been introduced from these islands.

The feature common to all the societies is that they are entered by a ceremony of initiation which may, as we have seen, vary from the mere giving of a pipeful of tobacco and the making of a hat or other object to the highly complex and prolonged ceremonial which has been described in the case of Tamate liwoa. There seem to be many gradations between these two extremes. Of the details of these ceremonies of initiation into the different societies other than Tamate liwoa we at present know next to nothing but the few notes which have been given show that there are variations in detail.

1 On the other hand, Dr Codrington (Af., 8i) says that the great Tamate is not of much importance in Ureparapara.

Another feature common to nearly all the societies, is the possession of a badge or badges and apparently all have the same degree of efficacy in the protection of property.

A feature common to most of the societies is the possession of an object called a iamate, either a mask, hat or other object, and it is probable that it is this possession of a tamate which is in the minds of the natives a most important feature of a Tatnate society. As will be seen later there are, however, other objects called tamate which may be worn by women and in order to wear these a person has to go through a process of initiation, and yet these tainate seem to be definitely distinguished from those of the Tamate societies.

The variation in the size of the societies and the reasons for it have already been considered in connection with the account of the function of the badges. The usefulness of these badges in the protection of property is the greater, the smaller is the society in numbers. It would therefore seem that it is in the interest of the members of a society to keep out new members, but on the other hand the payments made by the initiated as well as the interest and honour derived from the initiation of a new member act in the opposite direction. When the obstacles to initiation have been overcome and a member is about to be initiated, other men will at once apply to be initiated at the same time, except of course in the case of those societies in which only one man can be initiated at a time. Once an initiation has been set on foot it is for the immediate interest of a candidate to have as many companions as possible for it diminishes the amount which he will have to pay, but on the other hand a later consequence of the entrance of many others will be the disadvantage of sharing with them the protective power of the badge.

There are other conditions which influence the act of joining the Tamate societies, some acting as obstacles and others as motives. Of the former one very important is that a man must liquidate his debts {pug) before he can be initiated and this must be done with money, pigs not being sufficient. We have seen that if a man wishes to enter the Sukwe he gives his friends pieces of money which have later to be returned with cent per cent interest. I do not know if the obligation to repay such forced loans would be included among the debts of a man in the above sense, but if so it is

Titt tatas 123

easy to see how debt can be a most efficient hindrance to entrance into Tamate societies. On the other hand, debt may act as a motive to promote initiation ; a man who is in debt, or even one who only desires the money necessary for initiation, may try to induce men to join societies of which he is already a member in order to obtain the money he wants.

It was said that difficulties owing to mistakes, as in using a badge wrongly, may also form an obstacle to initiation, and if the fine is exacted this is evidently the case, but, as we have seen, one of the results of such a mistake is that a man sometimes enters the society whose badge he has used wrongfully, so that this may assist as well as hinder initiation.

A very efficient obstacle is the difficulty of getting anyone already a member to undertake the duty of acting as introducer. When a man enters a society he has to obtain the services of a member to whom he gives money, the native expression being that he has to ''tiro mun" this man. Though the introducer receives money from the initiate he has to give a pig, or pigs if more than one man is initiated. If a man has only one good pig, he will not consent to act as introducer because a consequence would be the loss of this pig. This is so well recognised that a man who is known to be so situated will not be asked to act. If, on the other hand, a man has a pig which is a fence-jumper, yam-eater, or a public nuisance, he will be only too glad to act as introducer and thus get rid of the animal with profit.

Sometimes disputes will arise as the result of reluctance to admit new members. In such cases the objectors usually carry the day, but if there is disagreement the process called tiu tatas may be set in action, and members may decide to join the societies which are called upon in the course of this tiu tatas.

Tiu tatas is a process arising in connection with quarrels which leads to men joining new societies. If two men quarrel about anything and one belongs to, say the Ta^nate mantelet, and the other does not, he will say to the other " Tiu tatas goro ko mun 0 Mantelel." The literal translation of this is doubtful but the general sense is "You are bound under the curse of the Mantelel society." If the other man is afraid, he will at once end the dispute, but if angry and defiant, he will call down a counter-curse, naming some

society to which he belongs and his opponent does not, and the first man may retort with a third curse, naming a third Tamate, and the quarrel becoming known, the whole island is involved in the dispute, men coming from all parts. The disputers would have to pay fines of pigs to those societies the names of which had been taken in vain. The tiu tatas curses can only be used in disputes about food and are not allowed in quarrels about women. Food which is the occasion of such a quarrel must not be eaten but is allowed to rot on the trees.

Tiu tatas may also be used when quarrels take place between members of the same society (see p. 123) and in this case the man who starts the quarrel invokes the name of some other Tamate to which he belongs. Those who have been cursed in the name of some society may retaliate by saying "Very well, we will now become members of the Tamate with which you tiu tatas us," and it is in this way that the custom of tiu tatas increases the membership of the societies.

In a case in which a quarrel arises owing to a man having taken the breadfruit of another, the effect of the tiu tatas was put in a different way. If in such a case the owner of the breadfruit belonged to the Viov he would say " Tiu tatas 0 patau,'' thus assigning the patau or breadfruit to the Viov society, i.e. to be the property of the members of that society only. I do not know whether this is an alternative to the ordinary soloi or whether it is merely another way of expressing the prohibition, but in either case it shows the importance of the Tamate societies in such disputes.

In the case just cited the owner of the fruit may take other measures. He may curse the offender in his absence, saying to those present, " Tur van gaganag munia was nia i 0 Viov,'' i.e. "Go and tell him that he is a man of Viov.'' This is taken as an insult which will make the offender wish to join the Tamate viov in order that he may be on an equality with his adversary. An alternative procedure is for the owner of the breadfruit to curse the offender to his face, saying, ''Iniko i 0 Viov," i.e. "You are a man of Viov," whereupon the pair would draw their bows to fight, and the bystanders would hurl themselves upon the pair to prevent them from killing one another. Should they fight and one be killed, the whole community would be involved in the fight.

Size of Societies 1 25

It is probably some such occasion as this which has been recorded by Dr Codrington'.

It was said also that rivalry {0 varlai) between different villages or districts may influence the entrance into societies. Thus, there is constant rivalry between the districts of Veverau and Luwai in Mota and whatever one does, the other attempts to do, so that if a society in one district has an initiation, it may act as an incentive for a similar accession of strength in the corresponding society of the other.

It is evident that the conditions leading to the growth or diminution in size of the different societies give rise to an exceedingly complicated mechanism. A social factor which acting directly would prevent new members joining a society may indirectly cause its growth and it would seem probable that the function of the badges in the protection of property must lead to a kind of up and down progress. A small society will become popular because it is small and when through this popularity it becomes large, its size will make it unpopular and no new members will join till it almost disappears, as in the case of so many Mota societies at the present time. With the more important societies the case is different, for here the motives for joining are of a different kind. The special obstacles connected with expense or the loss of pigs only apply to these more important societies, and the chief motives for joining them are the honour of belonging to them and the right of taking part in certain portions of the kolekole performances. In one case at least, that of the Tamate liwoa, the fact that membership is essential to progress in the Sukwe certainly forms a most important inducement to increase its numbers.

It is clear that many of the societies of Mota at the present time are on the verge of extinction but it may be that this is only preparatory to a new growth. In one or two cases it was said that people are now thinking of joining these almost extinct societies.

It was said that a Tamate society might even become wholly extinct and yet be revived. In such a case the right of reviving it would belong to the son of a deceased member. In reviving the society he would be said to "' tiro pule tamate,'' the last word here referring to the dead man, so that the phrase means "to be initiated into the property of the

1 M., 75.

deceased." It is noteworthy that the right should fall to the son and not to the sister's son of the deceased ; it seems improbable that this is a recent innovation.

One change seems to be definitely taking place at the present time in the societies, viz. the disappearance of their masks or other tamate objects. In several cases in which John Pantutun belongs to a society he has never worn or even seen its tamate. In some cases this is because the tamate belongs to another island and Mota people do not know how to make it properly, but in other cases it appeared to be the result of long neglect to make the masks so that the secret is lost or known perhaps only to one man. This is probably because the interest in these smaller societies depends chiefly on their function in the protection of property so that other features such as the masks have been neglected.

One of the most difficult features of the whole organisation which has been described in the preceding pages is the exact nature of the relation between the Sukwe (in the narrow sense) and the Tamate societies. Dr Codrington has stated^ that the Sukwe is not connected with the secret societies of the ghosts but it seems clear that this statement needs modification. Though in general the two institutions seem to be largely independent, it is quite certain that admission into one Tamate society, the Tamate liwoa, is necessary for progress to the higher ranks of the Sukwe. It is very unlikely that this is a recent modification. It might mean merely that a man who is not a member of Tamate liwoa would be thereby of so little importance that he would have no chance of progressing in the village organisation, but the information as given to me was too explicit to be interpreted in this way; it was not that progress to the higher ranks in general was barred but the line was definitely drawn at the Tavatsukwe division as that which could not be entered without membership of the Tamate liwoa. On the other hand, it is possible that this limitation is connected with the fact that the werewere sound forms a necessary part of the ritual of the divisions of the Sukwe above Tamatsiria and that this is the real link between the two organisations. I may point out here that previous admission to Tamate liwoa is not the only distinguishing feature of the Tavatsukwe rank. It is only people of this rank and above who in the old days were allowed to

iJ/., 1 02.

The Tamate Societies and Death 127

drink kava at all and even now it is only they who may drink it in the gamal. It is clear that there is a distinct line of cleavage at this point in the ranks of the Sukwe and this increases the probability that the connection between Tavatsukwe and the Tamate tiwoa indicates some fundamental relation between the two organisations.

Mr Durrad was told by John Pantutun of a special relation existing between a division of the Sukwe and a Tamate society but could not succeed in finding out its significance. During a visit to the village of Luwai in Mota Mr Durrad noticed that the bright red fruit of the marakwarakwa was much in evidence in some decorations. This is one of the badges of the Viov, as well as of other Tamate societies, but on this occasion the fruit had been used in connection with a ceremony of initiation into the Lano rank of the Sukwe. When he inquired about this, Mr Durrad was told that ''Lano belongs to Viov!' I cannot help suspecting that a thorough inquiry would reveal many more connections between the two organisations. It is difficult to understand the possession of hats by different ranks of the Sukwe without some such connection.

It is quite clear that ideas concerning death are closely associated with the Tamate societies. Not only does the word tamate mean 'ghost' or 'dead man' but in the ceremony of initiation there is evidence of the representation of death and return to life. Thus, the beating of the novice and the destruction of his house during initiation is very suggestive of a ceremonial death, and so is the wailing of his female relatives when the candidate leaves them. It was said that they mourned as if he was leaving them for ever because he might never return (see p. loi), but though it is possible that this may partly account for their grief, it seems probable that it may be taken as another sign that the occasion is regarded as a ceremonial death. The seclusion again and the feasts during its progress and on its termination, have a very suggestive resemblance to the common Melanesian custom of holding a feast every five or ten days after death up till the hundredth day. Just as the feast after a hundred days of seclusion is especially important, so often is the feast on the hundredth day after death. The absence of washing and the taboo on touching the sleeping-place of the initiate may also be interpreted in this way. It seems as if there is during his seclusion

a kind of ceremonial make-believe that the initiate is dead, though at the same time he has to carry out the duties of feeding the members.

Closely related to this is probably the use of the expressions about the birth and death of the Tamate. In the account of the different societies there were references to the Tamate having been born at the door of the gamal or elsewhere. It is very doubtful what these expressions mean but there is a reference to the death of a Tamate which is more explicit. This is in the account of the Ta^nate roropei in which an image of the dragon-fly called the tamate was made at the initiation of a boy and burnt after a dance in which it had been carried, and this burning was said to be the death of the tamate. We seem here to have a representation of birth, life and death. I do not know how far this burning or destruction of a tamate is general. We know very little of any ceremonial connected with the Tamate societies beyond that of initiation and it is possible that this account of the Tamate roropei is only one instance of cere- monial which plays an important part in the ritual of the societies.

One of the most extraordinary features of the Tamate societies is their large number. As has been seen, even in the island of Mota alone there are over seventy societies at least, while if the branches of the main societies are included, the number would probably reach a hundred. Certain facts leading to this multiplicity of which we have definite evidence may be mentioned here, though the full discussion of the subject must be left till later.

In the first place it is clear that the societies of one island may be introduced into another and there is even reason to believe that societies may be derived from distant regions, and that the Banks Islands have received certain of their societies from the Torres Islands or the New Hebrides. Dr Codrington has pointed out' that new societies may be started by anyone who gathers round him certain co-founders and gives as an example the foundation of a society named "the Norfolk Island bird" by some one who had seen a strange bird while on a visit to Norfolk Island. It would seem probable that many of the less important societies have been founded in this way. One example is the Tamate

' M., 76.

New Societies 129

wotlewona in which a design drawn at random by a woman was adopted for the name and hat of a society (see p. 1 1 9).

These two causes of multipHcation will not, however, take us very far and the existence of so many societies in the Banks Islands raises a problem which will have to be more fully considered later.

R. M. s.

CHAPTER V

BANKS ISLANDS

Social Importance of the Sukwe and Tamate Societies.

In this chapter I propose to consider certain aspects in which the organisations described in the two preceding chapters influence the general life of the people, of the uninitiated as well as of the members themselves.

I will begin with the ceremonies called kolekole which are not merely rites of the Sukwe or Tamate societies but are participated in by the whole population, male and female, young and old, initiated and uninitiated.

The special examples of kolekole of which I obtained accounts are closely connected with the Sukwe and Tamate societies, but it is possible that in many cases the connection is less close, though probably there are always certain features of the ceremonial which can only be performed by those who belong to one or other of these organisations.

After considering the kolekole performances I shall consider briefly certain other aspects of the Sukwe and Tamate societies in their bearing on the general life of the people and especially on the determination of social rank and importance, the distribution of wealth, and the protection of property.

The Kolekole Performances.

I was not able in the short time at my disposal to obtain any complete account of the very important institution which a Banks islander calls kolekole. This name is given to performances, often of a very elaborate nature, which are, at any rate in many cases, closely connected with the organisation of the Sukwe and in which much of the ceremonial can only be carried out by those who have been initiated into certain

The Kolekole 131

ranks of this institution or into the Tamate societies. The corresponding verb is kole and there are certain objects in connection with which it is customary or obHgatory to have a kolekole in which a man is said to kole the object. Among objects so treated are ornaments of various kinds, especially the hats or masks of the Sukwe, houses, trees, or stones, while it would seem that a man might kole with no other notion than that of acquiring the reputation of having done so. There is probably, however, far more meaning in such a procedure than my scanty investigations were able to elicit. The greater the number of things a man has been able to kole, the greater is his prestige and the higher his general position in society. A kolekole can be performed either by men or women, or probably more correctly, a kolekole can be carried out by a man on behalf of a daughter or other female relative, but in this case it is the woman who is said to kole. There are certain objects which it is the privilege of a man to kole, while there are others, and especially houses, which only become the objects of a kolekole on behalf of women.

Among the most important objects which a man should kole are various articles connected with the Sukwe, and especially the hats or masks, and it seemed that it is from these performances that the chief pleasures of the Sukwe are derived, for there are many features of the performances which are only open to those who have been initiated into certain departments of the ritual of the Sukwe and Tamate societies. Various objects worn at the kolekole performances can only be used after special initiation and the same is true of several methods of decorating the head to be described later. In all these cases payments of pigs or money or both appear to be the most important feature of the initiations.

Initiation into the more unimportant Tamate societies is often the occasion of a kolekole. A kolekole is not necessary in these cases but if held will increase the importance of the occasion and thereby of the new member.

One of the chief objects which a woman should kole is the house. This is not necessary for men, but is incumbent on every woman unless she is to be continually subject to great inconvenience. A house which has been subrriitted to this ceremony is called gavur lava and a woman who has not herself performed such a ceremony would be prohibited from entering or even approaching a house of this kind. A more

9—2

elaborate kolekole ceremony gives a house the name of tamate woroworo and here again such a house may only be approached by a woman who has performed the appropriate ceremony, i.e. only women whose own houses are tamate woroworo may approach a house of that character. A woman who had not performed these ceremonies would be unable to use a path which passed by a house of the kind which she might not enter.

The ceremony is often performed on behalf of children, for though a girl who has not performed the ceremony may live in the gavur lava of her father and mother, she could not go to any other house of the kind. Even although this ceremony may have been performed while the girls of the family are young, it is often convenient to have a second house in which those who have not been through the ceremony may live and many people have in consequence two houses, one of which has and the other has not been the subject of a kolekole of this kind. A pointed bamboo purlin cut into a mouth and called valvalai is a sign that a house has been the subject of a kolekole and if a man has carried out the ceremony a number of times for different women, he puts in his own house the corresponding number of valvalai. There is much competition between different people in carrying out the kolekole ceremonial. Each man whose daughter or niece has to kole will try to outdo the performance of his neighbours. The chief features of a kolekole are the dance, the killing of pigs and the payments to those who participate, and every one will try to excel his neighbour in the splendour of the dance, the number of the slaughtered pigs and the liberality of payment. The whole behaviour of the people seems to be exactly of the same kind as when among ourselves people endeavour to gain social kudos by the splendour of weddings or funerals which may occur in their families.

Among objects worn by women at kolekole performances which require special initiation, at any rate so far as payments are concerned, are certain designs called tamate, which are imprinted on belts. Some of these are made out of leaves while others called pari are plaited, the latter resembling both in nature and name the ordinary woman's belt. A number of these belts of both kinds are shown in PI. XI and in the case of one of them, the tamate worawora, I am able to give an account of its manufacture and use which has been sent to me

Plate XI

A

B

C

D

A, taniate wcvawora ; B, tamatc tironiii ; C, tainatc kwat ; D, pcwi or plaited belt with kwat kwoe design ; E, pari with gapagapa design.

IVomens Belts 133

by Mr Durrad. Recently a man named Gapal who lives in Vanua Lava went to Santa Maria and saw the design called tamate worawora which can only be worn by women who have paid for the privilege. Gapal, who really belongs to the island of Santo (Espiritu Santo in the New Hebrides), bought the privilege for his sister's daughter, Catharine, and her two children, paying for it pigs and forty fathoms of shell-money. Not only can Catharine wear this design herself, but she can give or rather sell the privilege of wearing it to any female relative and she proposes shortly to sell it to the two daughters of her brother to be worn at a kolekole which is to be held when her brother has finished building a house. The brother will pay pigs and money which will go to Catharine, to her two children and to Gapal. The privilege was given by Gapal to Catharine as a sign of his affection for her, as something by which she will keep him in remembrance. Further, it is of the nature of an investment both for his own and the woman's advantage, for anyone who wishes to use the design will have to pay Catharine and, as we have seen, Gapal would get part of the money.

The belt on which the design of the tamate worawora is imprinted is made from a leaf of the gavinegae, a species of pandanus from which mats are made. The leaf is bound round a stick and dried in the sun. The pattern of the tamate is cut out of the bark of a tree (several kinds being used for this purpose), and this cut out bark is tied on the leaf. Some bark of the breadfruit tree is then laid on the ground (other kinds of bark may be used) and on it is laid some lakwe, 2l plant which provides a crimson dye, and the roll oi gavinegae leaf, with the bark-pattern bound on it, is then laid on the lakwe and the whole rolled up so as to form a bundle with the breadfruit bark outside. A dry coconut leaf is lighted and the roll is laid on it and scrapings of various trees are dipped in water and squeezed over the hot bundle and the lakwe within the bundle stains the leaf except where it is covered by the bark-pattern and by the fibre binding the bark-pattern to the leaf\ This manufacture is carried out by women who must fast while so occupied.

^ The transverse lines due to this binding are also to be seen in the other leaf belts and on one of thenar/ or plaited belts. In the other plaited belt, the^a/rtgapa, such unstained portions are only to be seen here and there, especially at the edge of the belt, and the procedure in dyeing this belt must have differed from that used for the others.

In the case of the tamate worawora only part of the design, viz, the body of the tamate, is on the belt. The rest of the design, viz. the head, is tattooed on the wrist, a part of the body highly thought of.

The design shown on another leaf belt (PI. XI B) is a modern tamate called tironin or glass. It is said to have been copied from the glass of a lantern which took the fancy of the maker. I have no information about the other three belts shown in PI. XI. It may be noted that the design on one of the plaited belts is called gapagapa, a word which also denotes the pattern in the form of a swallow used in the decoration of the head (see p. 137).

Tke Kolekole Ceremonial.

I obtained an account of a kolekole which had been performed not long before my visit. This was a ceremony in connection with the hat of the Kworokworolava rank of the Sukwe. In this case the man who was carrying out the kolekole did not belong to this division of the Sukwe and it seemed clear that a man might kole the hat of any division, even of one superior to his own. Further, it seemed even that a man who wished to give a princely entertainment might kole all the hats of the Sukwe at one time and such an occasion had occurred not long before in which the hats of all the ranks had been made at one time and all the rites peculiar to the different hats had been performed on one day.

Whenever it is proposed to hold a kolekole a number of objects, decorations, etc. have to be prepared and men abstain from sexual intercourse while so employed.

A man who makes up his mind to kole a certain hat would go to one called mao who knows how to make it and would order it, saying whether he wished it made with a handle to be carried or whether it should be scooped out so as to be worn on the head. It would be made in the salagoro. When it is ready the werewere sound is made and all women and uninitiated persons leave the region in which the dance is to take place. They go to a spot near the beach, a given path forming the boundary beyond which the uninitiated may not pass, and a procession is formed at the salagoi^o whence the hat will be carried to the place of the dance. The man who makes the wei'ewere sound heads the procession, followed by

Kolekole Ceremonial 135

the man carrying the hat. Then comes the band of initiated persons while the rear is brought up by the man who made the hat, who is regarded as the Hon of the occasion. The man who is giving the kolekole waits in the village and when the procession arrives, a pig is killed for those who have taken part in it, and a live pig and money are given to the man who has made the hat. The hat is put in the gamal and the procession returns again to the salagoro with the werewere sound while the man giving the feast remains in the village. As soon as the werewere ends, the covering of the stick {sur) is burned and someone tells the uninitiated that they are free to go about again. As the procession passes along, its members may cut down any trees on their way, whether coconuts, breadfruit, or bananas, but they do not touch yams or taro. When they are thus destroying property they express pity for the owners, but the latter never complain and profess to regard the matter as a joke. The day following the procession is the great day of the occasion on which the proceedings are open to all. The people of the island come to the appointed place, many of the men having decorated their heads, some in white, others in red, yellow or green, using processes which are secret and only known to those especially initiated into the methods of preparation (see p. 137). The thickly interwoven hair of the men is cut into varied shapes, the cutting and decoration often employing them throughout the whole night intervening between the procession and the actual dance. When the people have assembled, the man giving the kolekole orders the bamboo drums to be beaten and the people sing. A woman dances round those who are singing and continues to do so till the people have sung a hundred songs. While this is going on the decorated men will not appear but will remain hidden, and when the hundred songs have been sung, the men who have been beating the drums go and the women continue to dance but without the drums. The women of each village dance in turns and as soon as all have danced they sit down and the decorated men come out carrying in their hands the bladders of pigs which have been rubbed with ashes and put in the sun to dry and then blown up and tied with string. As the men come out in pairs from different directions they dash the bladders against a tree making a great report. When they meet, all stand in a row and dance. When this dance is over,

all go to their respective salagoro to rid themselves of their decorations except those decorated in white who go to the sea-shore to wash off the powder with which they have covered their hair. Then all return to the village and are paid by the man giving the kolekole ; people of other villages in money, those of the man's own village with a feast. All then assemble and wait to see the hat which has so far remained in the gamal. The big drum called kore, made from the hollowed trunk of a tree, is beaten by three men and all again dance and sing. Then those connected with the hat go to the gamal and fetch it out worn on the head of one of them, if it has been made as a hat, or carried in the hand if made with a handle. As it comes out the drum is again beaten and pigs' bladders again dashed against the trees and the people who have their bows and arrows in their hands shoot at fowls or any other animals they wish to kill for food. Every one should kill his own food and then all take part in a dance in which the man with the hat dances round and round with it on his head or in his hand. Finally the hat is put on a decorated wona or platform and the people feast.

The hat is left on the platform till it rots ; if the man who has made the hat sees that it is becoming unsightly, he may break up the image but must leave the debris there. The day following the kolekole is tapu and no one may come into or go out of the village but must eat the food already there. This tapu lasts for five days while for five more days after that no stranger may come to the village. No leaf of coconut, whether green or dry, and no part of the pandanus called gavinegae may be brought into the village during this time nor may the women make mats. The man who has given the kolekole remains in the gamal for ten days and when he comes out he bathes in the surf and the place where he has bathed is made tapu for one or two years. People may neither bathe nor fish there, but if it is seen that there is a shoal of fish in the place, the man who has bathed there may remove the tapu on receipt of a payment of money.

The kolekole proceedings for the different hats of the Sukwe seem to have the same general character and differ chiefly in the nature of the decorations and the sound which is made. For the hats of Lano and of all ranks above it, the werewere sound must be made ; for Kerepue, Mzvele and Tetug they may have either werewere or iheretang ; for Tavatsukwelava

Decorations 137

they make the rheretang sound and for Tavatsukwe they use the wetapup decoration, white or dyed fowls' feathers, worn round the neck or ankle.

Decoration of the Head.

In the preceding account I have mentioned that the heads of those taking part in the kolekole dance are elaborately decorated, and these decorations play a large part in the interest and occupation provided by the kolekole. The knowledge of each kind of decoration is the possession of certain people, and anyone wishing to learn any part of the procedure has to go through a ceremony for which the term tiro is used, which also denotes the processes of initiation into the Sukwe and Tamate societies. No woman can ever obtain any knowledge of this kind.

One of the most popular pigments called urai kwat is red in colour and is prepared from a plant called tatamera, the coleus or Moreton Bay nettle. Lemons or limes are cut in half and scraped out and sea-water is put in them and placed over a fire. The leaves or stem of the tatamera are chewed and the contents of the mouth which have become of an intense black are ejected into a leaf of the via or giant caladium. The lemons are then carefully taken up with wooden tongs and the boiling sea-water poured into the leaf, turning its black contents at once to a brilliant red. The outer bark of the mamalau tree is then scraped off carefully so as to avoid bruising the inner bark, and this inner bark is scraped off in very fine fragments. These are mixed with the red juice and smeared over the head. When this pigment is used part of the head is usually left untouched on each side, the unadorned region having the form of a bird called gapagapa {Collocalia urupygialis).

Another pigment called urai salagoro has a colour which was compared by my informant to dull gold. It is prepared from the husk of the unripe fruit of the cycas which is spread out, dried and powdered, and the powder is blown into the closely interwoven texture of the hair.

Another yellow pigment is called urai ango, which uninitiated people suppose to be turmeric. This is not so, but our informant did not know the method of preparation.

A green pigment called urai mantap is prepared by

mixing the bark of a green bamboo with coconut cream, i.e. with the expressed juice of the white of the coconut heated to boiHng point by means of hot stones. The mixture does not look well at first but becomes a brilliant green as it dries.

A white pigment called urai gara is made on the seashore, all the others being prepared in the salagoro. Pig- ments not made in the salagoro may only be taken into it by a man who can werewere.

A supply of the cycas fruit used to make urai salagoro is always kept ready for use. People notice the fruit and if the tree belongs to a man who can werewere^ he is asked for it. In the case of anyone else, the fruit is stolen, and even if people knew, they would never tell the owner by whom his fruit had been taken. The gapagapa pattern is only used with the red and yellow pigments, and in the latter case only by one who has been initiated into tirai kwat, i.e. a man must be privileged to use it with red pigment before he is allowed to make it with yellow. When waiting for the dance the people who have used urai kwat and urai ango sit together, but apart from the rest, while those with urai mantap and urai salagoro have not only to keep apart from the other two but also from one another. When the people go to wash off the pigments after the dance, all go to the salagoro except those with urai gara, who go to the sea-shore, the place where the pigment has been prepared.

The hair is not only coloured but made to assume various shapes, partly by cutting, partly by pressing. This is done by those who are skilled but no especial initiation is necessary.

It is doubtful how far the kolekole performances form an integral part of the organisation of the Stikwe and Tamate societies, and it is possible that they form a separate institution which has only become connected with that organisation owing to the custom of associating with the performances certain features of the secret ritual. It is probable, however, that the association with the Sukwe and Tamate societies is very close and deeply seated, and that the kolekole performances are means whereby the general population is allowed to participate to some extent in the ceremonial of these bodies. Further, the foregoing account has included many

Social Importance of Sukwe 139

features of ritual in which women are especially concerned. In several of these cases the objects round which the ceremonial centres are called tamate and it would seem possible that these are occasions on which women are allowed or encouraged to imitate features of the men's organisation in what seems to the men to be a harmless way. It is a question whether there has not come into existence some kind of women's organisation analogous to the Sukwe of the men. Dr Codrington states^ that women are admitted to grades of honour by the payment of money and the making of feasts, but I am not able to add anything to his account. The special point to be noted is that though women have no place in the Sukwe and Tamate societies and are even excluded from those parts of a kolekole performance which involve certain features of the ritual of the men's organisation, there are other parts in which they and other uninitiated persons participate freely, being allowed, for instance, to see the hats of the Sukwe.

I can now proceed to consider several still more important ways in which the Sukwe and Tamate societies affect the general life of the people.

One of the most important functions of the Sukwe and Tamate societies is that the determination of social rank and importance is largely dependent on them. It is not merely that a man can acquire increase of rank and general social importance in the eyes of the community in general by rising in the Sukwe, but also that such a high place in social estimation can be acquired by the possession of certain special accomplishments, such as that of being able to make the werewere sounds in all their completeness. Further, as has just been seen, a man can rise in social estimation merely by his lavishness in the carrying out of kolekole performances.

As Dr Codrington has pointed out", the part taken by the Sukwe and Tamate societies in the acquisition of social importance has introduced much obscurity into the problem of the existence of hereditary chiefs in the Banks Islands. His account seems to make it clear that there are in these islands true chiefs, etvusmel or tavusinele, but he found himself unable to distinguish definitely between such men and those who are called "great men" owing to high rank in the Sukwe. It was said that the dignity of etvusmel descended from father

» M., 1 10. 2 j/_^ 54,

to son, but it was doubtful whether this meant anything more than that the father put his son in the way to rise high in the Sukwe and the derivation of the name for a chief, ta vus ihele, the man who kills for the cycas, certainly suggests that the idea of chief is very closely connected with rank in the Sukwe. Owing to preoccupation with other topics I made no inquiry into this subject and its elucidation must be left for further inquiry. The point which is quite clear is that, whether there be hereditary chiefs or not, social rank and importance are so closely connected with the Sukwe that if a true chief were not a member of the Sukwe and Tamate societies or only of low rank he would possess little authority or importance in comparison with one standing high in these organisations.

Another most important function of the Sukwe and Tamate societies as well as of the kolekole performances is in the distribution of wealth. Every initiation into a rank of the Sukwe or into a Tainate society is part of a process whereby money passes not only from the new member to those already initiated but from his relatives and from certain members already initiated, such as the introducer, to other persons. Further, the account of the initiation into the Wometeloa rank in Vanua Lava (see p. 68) shows that the contributions made towards the price of initiation are largely if not altogether in return for previous gifts or payments. It is evident that there has thus come into existence an immense body of vested interests. A man who has spent large sums of money in order to rise high in the Sukwe has partly done so in order that by receiving money from those initiated later he may acquire wealth which will enable him to help his children and other relatives to follow in his footsteps. That such an idea is clearly present in the minds of the people is shown by the reasons given for the gift of the Tamate worawora by Gapal to his niece (see p. 133). I think there can be no doubt that one of the motives which leads a man to advance in the Sukwe or to enter his children is an idea corresponding very closely with that which underlies our practice of investment of capital. From one point of view, then, the Sukwe and Tamate societies and other associated institutions form a complex organisation by means of which wealth is acquired, and since it is only the rich or those with rich friends who can advance far in these bodies, the

Economic Aspect of Sukwe 141

organisation is a means for the perpetuation and even the accentuation of differences of social rank in so far as this rank is dependent on the possession of wealth.

It is, however, clear that there are other deeply seated ideas in the minds of the people which prevent wealth from becoming the chief means of social differentiation. The mere acquisition of wealth for its own sake is probably wholly foreign to the ideas of the people. The wealth derived from a high position in the Sukwe is probably regarded merely as a means to still further advance. It has been seen that the highest ranks in the Sukwe are often without members and there is always the possibility of further advance with the great expense which such a step will involve. There are also the elaborate kolekole performances already mentioned which are the recognised means of further advance in social estimation.

A man, however high in the Sukwe he may be, will suffer social depreciation if he does not undertake such expense. Mr Durrad was told that when a man reaches high rank in the Sukwe and has thereby the power of amassing wealth, he would be considered unworthy of respect and honour if he hoarded his gains. To retain his influence and glory he must distribute his money by paying people to work for him in his gardens and by giving splendid kolekole performances. The Banks islanders seem to have developed a sentiment which has removed a social danger to which the organisation of the Sukwe standing alone would have exposed them.

Though there are thus social factors which have prevented the Sukwe and Tamate societies from becoming too exclusively the means for the acquisition of wealth, there can be no doubt that the high degree of development of these bodies has brought in its train so complicated a mass of vested interests that their disappearance would produce very great confusion in social values. It may be noted that the whole of the monetary transactions of the Sukwe take place in the native currency of shell-money and it has been suggested that with the increasing use of European money these vested interests would lose their importance and lead naturally to the decay of the organisations.

The part played by the Tamate societies in the protection of property has already been fully considered. Their influence is of two opposed kinds. It is by means of the badges of the

societies that property is normally protected, these serving as taboo-marks analogous to those which act as the protectors of property elsewhere in Melanesia. At the same time there are several occasions on which the injury or destruction of property is directly due to the action of the Tamate. In one case there is reason to believe that this destruction has a definite ceremonial significance, viz. the destruction of the house of the candidate for entrance into the Tamate liwoa. This is probably symbolic of the ceremonial death of the candidate, and if we knew more of the other occasions on which property is destroyed we might find that these have in every case their meaning. At the same time they are very suggestive of the occasions for general relaxation of law which are so often found among rude peoples. Just as there is reason to believe that at certain festivals the most fundamental laws regulating the relations of the sexes are not merely broken but that an excessive degree of relaxation is allowed or enjoined, so may it be that we have on the occasions when the Tamate destroy property in a wholesale way a general relaxation of the laws regulating the protection of property in which the Tamate societies normally take so prominent a place.

It is clear that the functions of the societies in enforcing respect for property-rights are both numerous and important. It is probable that they are also the means of applying or enforcing penalties for other social offences. The only definite fact of this kind of which I know concerns a child. A father told his little boy to do something, but the boy disobeyed him and went down to play with some companions on the shore. The father told a youth who was a member of the Tamate saka to chase the boys with the saka and to single out his disobedient son for special punishment.

In conclusion, it is clear that the Sukwe and Tamate societies are not merely associations to awaken and then to satisfy curiosity or to terrorise those who do not belong to them, but have a very definite and useful function in the community. They are closely concerned in the respect for property and in the maintenance of social order. At the same time they give the people an absorbing interest which, so far as the kolekole performances are concerned, is shared by the whole population. The ordeals and trials of patience perhaps seem to present a less desirable side but even these

Social Value of Sukwe 143

have their merits and have probably contributed to form that habit of restraint and respect for authority which is so strongly developed in the people. To quote the words of a missionary who has observed the operations of the Sukwe, " I could not help feeling that the existence of the Suqe does much to foster the virtues which go to make up the successful citizen and man of business and so maintain the vigour of the community. To be a successful Melanesian citizen [i.e. one high in the Suqe\ means that a man must have unlimited patience, indomitable perseverance, excellent health, and a strong physique to enable him to recover from the perpetual succession of losses of which he runs the risk."

CHAPTER VI

BANKS ISLANDS

Birth and Childhood Customs, Relations with Animals AND Plants, Magic, Possession, Money, Canoes, Decorative Art.

Birth and Childhood Customs.

At childbirth the woman's own mother and sisters will not be present nor will the own mother or sister of her husband, but women who are called sister or mother through the classificatory system will attend and the husband's own sister will choose one of them to act especially as midwife.

The fact that a birth is about to take place is kept as quiet as possible, and it was said that this is done in order that there shall not be too many visitors, each of whom would have to be paid for attendance, while it is evident that the customs connected with adoption already considered often make it desirable to keep the event as much as possible from general knowledge.

When the child is born the cord called gaputoi is divided with a bamboo knife (even now a steel knife would not be used for this purpose). The afterbirth is buried under the fireplace of the house and a fire is kept burning above it, the idea being to assist in the drying of the remains of the cord on the body of the child, but if the navel becomes inflamed, the fire is extinguished. The child is kept in the house till strong, the idea being that the afterbirth which is part of him is in the house, and if the child is taken to another village before he is strong he will be sulky and cross because part of him remains in the house ; he has been taken away from his nigina (or nest) which is a name for the fireplace and the part of the house surrounding it. The after- birth is regarded as the house of the child before it is born.

Birth Customs 145

When the piece of cord left on the child separates it is tied up in a piece of umbrella-palm leaf and offered by the father to his own sister. He knows, however, that she will refuse it and he then gives it to another woman, whom he calls sister according to the classificatory system, who hangs it round her neck, the leaf containing the cord having been so completely covered with string that it is not visible. She keeps this on her neck till the child is about two years old when the father must give a feast, and the reason why the own sister of the father refuses the piece of cord is that she does not wish to have to give her brother the expense of providing the feast. It was said that sometimes when a woman has a bad disposition and has a grudge against her brother, she will accept the gift of the cord, rejoicing in the prospect of the expense and trouble to which she will thus have the opportunity of putting him later. After the feast has been given, the cord in its coverings is hung up in the rafters of the house above the fire. The father's sister will also ask for some of the nail-parings of the child and keep them on her neck, and it would seem that these also might be made the occasion of a feast. The father's sister may also keep her nephew's nail-parings when he is older. I was told of a case in which the father's sister of an adult man picked up some of his nail-parings when he was going to another island and when he returned he had to give her a feast.

It is regarded as a good thing to have twins if these are both of the same sex. If they are boy and girl they are regarded as man and wife, but my informant could not tell me what was done on such an occasion. It is probable that in old days they were killed.

Before the birth takes place firewood, torches and coconuts are brought into the house. When the pains begin, the husband opens with his own hands anything covered up ; he takes the string off any bowl, removes the string with which the door is fastened and takes down anything which is hanging up. While bearing a child, a woman must not eat fish caught with a net or with a hook, though she may eat that shot with an arrow. The idea is that the child will be entangled as in the net or on the hook and will not come forth freely*.

^ I was told of these customs in connection with the birth of a first-born child but they are probably practised before any birth.

The parents must remain in the house for some time after the child is born. The father must not do any work for five days (till the cord has separated) and he must not do any hard work for a hundred days. He must not eat food which comes from a place where the people have been doing an oloolo rite
nor must he go to such a place himself. If by any chance this happened, he would have to submit to a ceremony to remove the source of danger to the child. For some time after childbirth any visitor who has come from the beach or from a distance must not go straight into the house but must wait outside for a time.

Directly after the birth of a first-born male there takes place a ceremony called kalo vagalo. A little bow is put into the hand of the child and a woman stands with the child in her arms at the door of the house. All the maraui or maternal uncles of the child collect outside and shoot at the woman and child with blunted arrows or throw lemons at them. The woman moves the child about so as to diminish the chance of its being struck, and after a time the father puts an end to the business by paying money to the uncles. When the woman brings the child back into the house she hands it to the sister of the father who holds the child with her arms straight out till it trembles, and then says: — "You and tawarig go up into the cultivated land ; you with your bow and tawarig with the basket, digging yams ; you shooting birds, tawarig breaking up firewood ; you two come back into the village ; she will take food and carry it into the house ; you will take your food in the gajual." When she has said these words she lifts up the child. The tawarig of this speech refers to the woman who will marry the child when he grows up ; the wife will be the tawa7ng of the father's sister and the passage should probably read "you and my tawarig T

When a male child goes out for the first time the father and mother go with it and tie leaves together and throw them down on the path so that when the people see them they will know that a male child has been added to the community.

Before the birth of a first-born child, probably at an early stage of pregnancy, the husband gives a feast, and a rite called valugtokwa is performed. The wife's brother chooses four male relatives each of whom gives him half a fathom of shell-money and to each the brother himself adds half a ' See pp. 156, 160; also Codrington, M., 140.

Birth Customs 147

fathom, so that there are altogether four fathoms, and the father of the child provides eight fathoms. A yam pudding has been made which is taken out of the oven and put in the open place in the middle of the village and the wife's brother then puts his four lengths of money on the pudding so that they lie from east to west. He then goes for water to a particular spring which is used for this purpose only (there is only one such spring in Mota) and brings the water back in a leaf of taro, which he has tied up carefully to hold it. He must do this alone and in absolute silence. The expectant mother then stands at the west end of the strings of money and her brother takes up one of the four fathoms and puts it over her right shoulder, and so with the other three fathoms. The woman then takes the money off her shoulder and holds it hanging down over the pudding while certain words are said by the father's sister. The husband puts his eight fathoms on the pudding lying from north to south ; then takes them up again and gives them to his wife who hands them to her brother. The latter takes the water he has brought and, standing behind the woman, slowly puts the leaf containing the water over the woman's head till it comes below her chin, the woman standing quite still. He then pinches the bottom of the leaf-cup and if the water squirts out the child will be a boy ; if not, a girl. The whole proceeding must be carried out in absolute silence and the rite is often done at night to ensure absence of noise.

The sex of a child is also believed to be known before birth by means of dreams, and when the sex is already known in this way or through the rite which has been described the child may sometimes be named before birth. My informant could not say what would be done if the prenatal diagnosis turned out to be incorrect. Whether the naming is done before or after birth there is no special ceremony or feast.

In all cases there is a feast ten days after birth, when those who have been present at the birth receive their payment. The women paid on this occasion give coconut oil which is drunk by the parents of the child.

The foregoing account applies to the island of Mota and the following was obtained from a native of Motlav.

When a woman gives birth to her first child all the women of the village, sogoi of both parents, assemble in the house. They bring their mats and sleep there, the gathering being

10 — 2

called lengvisi. On the fifth day after birth there is a feast, the men eating in the gamal food which has been cooked in the house. Yam puddings are made and cut up and the pieces are put in two baskets, one for the men and the other for the women. The women who have assembled may sleep in the house for twenty days. The whole of this time is regarded as a festival occasion. The women will have a different kind of food each day, and in the old days it would be decided that on one day they would all paint their eyebrows red, put scented leaves in their armlets or wear hibiscus flowers as fillets for the head. The important thing was that all wore the same decoration on any one day. Now that imported garments are worn all decide to dress in the same colour, one day in red, another in blue and so on. While the women are living in the house of the newly delivered woman in this way they may go and dig yams anywhere and it will not be regarded as stealing, this custom being called tigtigopur. They decide where they will tigtigopur yams and may go to another village for the purpose.

On the twentieth day the husband pays all the women, each being given a short piece of money called taraky and if a woman does not receive this she will refuse to go. The wife will be especially helped by four women, and two of these will get full fathoms and the other two short fathoms from the fingers to the opposite shoulder. A pudding is made and distributed and the pieces of money are put on the top of the pieces of pudding. When the money has been put on the pudding the women all sit down in a ring outside the house and the sister of the father of the child brings the baby out of the house and hands it to the first woman in the ring who passes it to the next and thus it goes round the whole circle, each woman holding the child for a time. When the father's sister again receives the child she carries it round the whole circle four times and then takes it to the mother who has remained in the house during the ceremony. If the father's father is alive he divides out the pudding and the money, but if he is dead this is done by the father himself. As soon as the distribution is over, each woman takes her mat and returns home where she eats the portion of pudding which she has received.

When the mother is strong and is able to go about a young man is told to build a little house called paito in the

Birth Customs 149

bush near the village. The mother then goes to it with the child and with a little girl who may be of either sogoi. The three take with them many packets of ashes wrapped up in leaves and carried by the little girl in a basket. The mother hands the child through the little house to the girl, who then goes round and hands the child back to the mother, again passing it through the little house. Then they go on farther towards the garden-ground, and after going about twenty yards they open one packet of ashes and scatter it over the pathway and then go on the same distance and scatter the contents of another packet and so on till all the ashes have been used when they return to the village. This handing through the house is only done for a first-born male child. No house is made in the bush for a girl, but otherwise the whole ceremony is the same. The idea of handing the boy through the house was said to be that it signified the child going through the Sukwe later. If anyone sees ashes on the path he will know that there has been a tarak^. After this ceremony the mother will go to work.

In Motlav they have a custom called virep virep similar to that called kalo vagalo in Mota, but there are many differences. The child has no bow in its hand and remains in the house, the father only standing at the door, when he is pelted with fruit and not shot at with bows and arrows. When the uncles are paid by the father each is given a small piece of money about a foot in length, measured from the tips of the fingers to the wrist which is called leme gambak. There is nothing corresponding to the ceremony in which the father's sister holds the child and utters the formula about the tawarig.

There is a feast called sau in this island before the birth of a first child, for which they scrape nuts (ngai) in sea water and cut breadfruit in half. They prepare the feast in the evening after the sun has set, cook the food during the night and eat in the morning. It is a festival for the whole village and all contribute to the food for the feast. The food provided by the father and his sogoi must be equal in amount to that provided by the mother and her sogoi, and this food is exchanged, the exchange being called ravaleitleit and then both parties take the food away and eat it in their houses.

Another ceremony called hirmarmartok takes place shortly

• The name of the ceremony is apparently the same as that of the short lengths of money given in the earlier rite.

before delivery is expected when the husband gives to the wife's brother two fathoms of money, the idea being that the woman's womb may be large and the child strong. This again is only done for the first-born.

If the husband catches an octopus during his wife's pregnancy neither he nor his wife will eat it, the idea being that the child will cling to the womb and not come out. As in Mota the mother must not eat fish caught with a hook, it being believed the child would be held back as with a hook. If during pregnancy a woman wishes to roast breadfruit or yams in the bush she will not make a fire on the ground but will put the fire on a platform of stones, the idea being that the child may grow tall.

In order to produce safe delivery the woman's mother will pound the leaves of the romlis tree in a bowl of the kind called wumeto and rub the juice of the leaves on the abdomen and will also give a draught to drink. If there is delay the people will see if the husband has anything — belt, armlets, etc. — bound on him, and if so he will remove them, but they do not trouble about unbinding things in the house as in Mota.

If a woman wishes that her child shall be either a boy or a girl, it is believed that she can produce the result she desires by eating the fruit of certain trees. One tree will cause the birth of a boy and another that of a girl. It is believed that boys stay in the womb longer than girls, and if birth does not take place at the expected time, they know that it is going to be a boy. After the child is born the father does not work or go far away till the child is strong, the motive given for this being that the man should be at hand to help his wife. My informant did not know of any proceedings in Motlav corresponding to the ceremony by which the sex of a child is foretold (see p. 147).

The ceremonies which have been described for a first-born child are performed whether the child is male or female, and only once, so that if the first child is a girl and the second a boy the latter would not be handed through the little house.

Relations with Animals and Plants.

In the last section I have omitted a group of customs followed in Mota and Motlav in which certain events which

Animals and Childbirth 151

occur before childbirth greatly affect the life and fortune of the child, and bring about a definite relation between it and an animal or plant.

In Mota there are many persons, perhaps as many as half the population, who are not permitted by custom to eat the flesh of certain animals nor to eat certain fruits nor to touch certain trees. The ground for the prohibition in most cases is that the person is believed to be the animal or fruit in question, his mother having received an influence from an animal or plant of the kind before his birth.

The course of events is usually as follows : a woman sitting down in her garden or in the bush or on the shore finds an animal in or near her loin-cloth. She takes it up and carries it to the village where she asks the meaning of the appearance. The people say that she will give birth to a child who will have the characters of this animal or even, it appeared, would be himself or herself the animal. The woman then takes the creature back to the place where she had found it and places it in its proper home ; if it is a land animal, on the land ; if a water animal, in the pool or stream from which it had probably come. She builds up a wall round it and goes to visit and feed it every day. After a time the animal will disappear and it is believed that that is because the animal has at the time of its disappearance entered into the woman. It seemed quite clear that there was no belief in physical impregnation on the part of the animal, nor of the entry of a material object in the form of the animal into her womb, but so far as I could gather, an animal found in this way was regarded as more or less supernatural, a spirit animal and not one material, from the beginning.

It has happened in the memory of an old man now living on Mota that a woman found an animal in her loincloth and carried it carefully in her closed hands to the village, but that when she opened her hands to show it to the people, the animal had gone, and in this case it was believed that the entry had taken place while the woman was on her way from the bush to the village.

I could not find out what interval usually elapses between the disappearance of the animal and the birth of the child, but this did not seem to be regarded as a matter of importance. It seemed clear that the people are now aware of a relation between sexual intercourse and pregnancy, and that cases

of childbirth which follow the appearance of an animal or plant are not regarded as exceptional in this respect.

When the child is born it is regarded as being in some sense the animal (or fruit) which had been found and tended by the mother. The child may not eat the animal during the whole of its life, and if it does so, will suffer serious illness or death. If it is a fruit which has been found the child may not eat this fruit nor touch the tree on which it grows, the latter restriction remaining in those cases in which the fruit is inedible. Thus, a fruit used as a taboo-mark would be useless for this purpose to one who owed to it his origin. A case has occurred quite recently in which a girl unwittingly offended against the prohibition. She was an eel-child, and though quite young had gone to fish with some companions on the shore. They caught some fish including an eel, and all were cooked by them on the shore in the same pot, and were then eaten. A few hours afterwards the child began to rave and became quite mad. The people inquired into the doings of the child and found that she had not eaten any part of the eel, but only the fish cooked in the same pot, and this was held to be sufficient to have produced her condition.

The belief underlying the prohibition of the animal as food is that the person would be eating himself. It seemed that the act would be regarded as a kind of cannibalism. It was evident that there is a belief in the most intimate relation between the person and all individuals of the species with which he is identified.

There is also the belief that a child partakes of the physical and mental characters of the animal with which it is identified. Thus, if the animal found has been a sea-snake, and this was said to be a frequent occurrence, the child would be weak, indolent and slow ; if an eel, there will be a similar disposition ; if a hermit crab, the child will be hot-tempered ; if a flying fox, it will also be hot-tempered and the body will be dark ; if a bush-turkey, the disposition will be good ; if a lizard, the child will be soft and gentle ; if a rat, thoughtless, hasty and intemperate. If the object found has been a fruit, here again the child will partake of its nature. In the case of a wild Malay apple {malmalagaviga) the child will have a big belly, and a person with this condition will be asked, " Do you come from the malmalagaviga ? " Again, if the fruit is

Animals and Childbirth 153

one called womarakarakwat the child will have a good disposition\

The animal or plant nature believed to be possessed bycertain persons has no influence in the regulation of marriage. There is nothing to prevent an eel-man marrying an eelwoman if they are otherwise suitable mates.

In the island of Motlav there is a similar belief that if a mother has found an animal before childbirth, the child will be identified with that animal and will not be allowed to eat it. Here again the child is believed to have the characters of the animal ; thus, a child identified with a yellow crab will have a good disposition and be of a light colour, while if a hermit crab has been found, the child will be angry and disagreeable. In this island a woman who desires her child to have certain characters frequents a place where she will be likely to encounter the animal which causes the appearance of these characters. Thus, if she desires a light-coloured child, she will go to a place where there are light-coloured crabs.

I inquired very carefully whether a case had ever been known in which the prohibition of an animal as food due to this belief had been passed on to a child or other descendant, but it seemed clear that such an idea was quite foreign to the beliefs and customs of the people. The prohibition is purely an individual matter.

There are other conditions in these islands under which persons abstain from the use of certain kinds of food ; thus, a Motlav man who has certain powers of a medical or magical kind may abstain from eating some animal, believing that his rites would otherwise lose their efficacy. Again, it is habitual in mourning to abstain from certain kinds of food. Sometimes the mourner will himself or herself decide to naro or prohibit certain kinds of food, and this custom is so definite in the case of a widow or widower that the term naro is used for the condition of these persons. When a man dies his sister's son may naro any kind of food he chooses and the same right is possessed to a less degree by other relatives. It was said that bananas and breadfruit are not the subjects of this naro

1 The disposition of a child may have other causes. A child born in a place where the grass sways from side to side will be like it, undecided, turning from one purpose to another, while one boy now on Mota is fond of playing because an old man dreamed shortly before his birth of children playing round a pool.

custom and are never forbidden as food. I could not find that these prohibitions are ever transmitted to a later generation.

The Tamaniu.

Another example of a special relation between human beings and animals is to be found in the tamaniu. This relation differs from that already considered in that it concerns a person and an individual animal rather than the whole of a species, though there is this degree of resemblance that a man who has a given animal as a tamaniu will not eat any animal of that kind.

A man obtains a tamaniu by means of a definite rite carried out by a man who possesses special power (mana) for that purpose, or perhaps, more probably, owns a stone or other object which has such mana. This man pounds the leaves of several kinds of tree into a pulp and either rubs the pulp on the man who wishes for a tamaniu or gives him some of the juice to drink. The pulp is then deposited in some cleft of the rocks, in a place where the leaves cannot be reached by salt water. At the end of five days the place is examined to see " what creature has been brought forth " and this, whatever it is, becomes the tainaniu of the man on whose behalf the rite has been performed and is given by him to his father, mother, or grandparent to put in a secret place. The man says, "This is myself; hide me." The man himself will be told where the creature has been put and if he has a son, he will also be informed. The following are animals which have been known to be tamaniu : the crab {^gave), rat {gasuwe), eel [maj'ea), sea-snake [mae or mai), fishes called takwale kwoe and takwagato, the shark (pagoa), sting-ray (vaiH), octopus (wirita), turtle (uwa), centipede {surataramoa), a black lizard [puasa), coconut crab {naeru), and two sea birds called seglama and mantoanei^.

The tamaniu are said to grow sometimes to a great size. They are quite tame towards their owners and if small enough may even sit on the hand. Even if injurious ordinarily, they will not hurt their owners.

The tamaniu has functions of two kinds. If its owner wishes to injure anyone, he will speak to the man who has

• It is evident that the rite by means of which some of these tamaniu are procured must differ from that which I have recorded.

The Tamaniu 155

procured the tamaniu for him, saying that he wishes the animal to injure his enemy and the tanianm will do so in the manner peculiar to itself ; if it is an eel or centipede, it will bite him ; if a shark, it will swallow him.

A more important function is to act as a kind of life-token, a representative of the man himself whereby he can discover when he is ailing or likely to die. It was said that it is the thought of death which makes a man wish to have a tamaniu, as by its means he can know whether he is going to live or die. When a man is ill he will send his son saying, " Go and see whether there is anything harming me or no." The son may ask where to find the tamaniu and will then go and call out, " Father ! come out here ! " He will hear a noise and the animal will issue from its hiding place and the son will look at the body of the tamaniu to see whether it has been injured. He may find that some object is sticking in its skin, in which case he will remove the object and his father will recover. If he can see no injury, he will ask the tamaniu whether it is going to live or die. If it nods its head it is a sign of death ; if it shakes the head, the man will get well. If the man dies it will be found that the animal has died too.

I was given a striking account of a recent occurrence in connection with a tamaniu, which had in this case been brought to Mota from another place. The owner, a blind man, went to the small island of Merig and asked a man there to carry out the appropriate ceremony. A lizard of the kind called puasa appeared and was brought to Mota and put near the village in the roots of a big banyan tree, where it grew very large. Long after, the blind man fell ill and told a friend to go and see the animal, using the words " Look at me," referring to the lizard as himself. The man went alone to the tree, but when he got there, was too frightened to call upon the animal. He was sent again, and this time took some companions to keep up his courage. When they reached the tree the man called the name of the animal, Rosasangwowuty and the tamaniu came out. It was a very large lizard, much larger than any other the men had seen in Mota. When it came out it seemed sluggish and walked as a sick man would walk. The son of the blind man, who was one of the companions, then asked the tamaniu if it was ill, and the creature nodded its head and returned into the roots of the banyan tree. They went back and told the blind

man that he was ill, and not long after he died. At the same time the banyan tree fell and is still lying on the ground across the pathway, and this was taken as a sign that the tamaniu was also dead.

The abode of the tamaniu is kept very secret. If it were known, it would be possible for an enemy to kill it and thus kill the owner, for if a tamaniu is killed, the owner dies at the same time. A man who has killed the tamaniu will tell his enemy " I have shot you dead," and the man thus addressed will die instantly.

It will have been noticed that on several occasions the tamaniu is addressed or spoken of as if it were the man himself. It is clear that there is a definite identification of personality between the man and the animal, and this idea of identity is carried so far that the injury or death of one necessarily involves the illness or death of the other.

Magic.

The following account is very fragmentary and gives merely a few examples of the accumulation of beliefs and practices which a full investigation would reveal in these islands. The amount of material which I was able to collect in the short time at my disposal is sufficient to show that the magic of the Banks Islands presents much of interest and one feature which, so far as I know, has not been described elsewhere, viz. the existence of special associations, bodies of men who, while learning a magical method, at the same time protect themselves against the magic of others.

My information comes from two islands, Mota and Motlav, and as there are definite differences in the practices of the two places, I will deal with them separately. In both islands the name for the rites by which the magical effect is produced is oloolo. Both beneficent and harmful forms of magic exist ; there are the rain-makers and promoters of fertility on the one hand, and the producers of injury and death on the other.

Mota. The more beneficent forms of magic are the appanage of certain men who have usually inherited their powers from the maternal uncle and only exceptionally from the father even now when, so far as other property is concerned, inheritance from the father is becoming frequent. The

Magic 157

efficacy here is inherent in certain objects and it would seem that this efficacy or mana belongs to the object, such as a stone, rather than to the possessor of the stone. In this island a person who steals a stone with mana may acquire the power of carrying out the rite effectively, and I was told that it was not uncommon for stones to be stolen for this purpose. Formulas are uttered but these appear to be regarded as of minor importance.

Increase of the supply of fish is brought about by means of stones of certain shapes which are put on the reef, certain leaves being, in some cases at any rate, put on the stones. Turtles do not come to Mota but in the island of Vanua Lava there are similar stones to increase the supply of these animals.

To promote the growth of yams, stones resembling yams are buried in the middle of the garden.

In rain-making a big clam-shell is dug into the ground and a bright orange fruit is strung on a creeper to imitate the rainbow ; a new and glittering clam-shell is opened and shut rapidly to imitate lightning ; thunder is simulated by beating the shell of a coconut on the ground and a fire is lighted so that the smoke may represent clouds. Words are said in each case but my informant did not know them.

There are also methods for producing sunshine and for making a big surf, the latter being put into activity when some one from another island is coming to collect debts.

An interesting feature of all these beneficent methods is that they are carried out by those who own the efficacious objects without reward. The owner of a form of magic uses it on behalf of another not for any material reward but in order to add to his reputation for the possession of mana ; that he may attain to a greater extent the attribute called rongo.

A rite was carried out at the sacred place of the Talosara division of the Takwong (see p. 23) which was designed to promote length of life. At this place there are two rocks, one called I Togaul (long life), the name of a vui or spirit ; the other Tangil, the name of the digging-stick of Togaul, while near the two rocks is a pit reputed to have been dug by Togaul. The place belonged to a man who carried out the rite necessary to prolong life, charging five fathoms of money for the privilege. The seeker of life stood by Togaul

158 Melanesimt Society

and put the five fathoms upon the creeper covering the rock and the owner of the place then removed the creeper so as to show the rock beneath and then looked hard and long at the other man, saying at length, " You will be an old man ; if you fight, no man will kill you or hit you. Here is Togaul ; here is his digging-stick ; here is a grave where he buries those who die." In one case a man on whose behalf this rite had been performed was given a sign in order to strengthen his faith ; he was told that if he lighted a fire he would find later a fungus called kwero growing on the spot where the fire had been.

On the maleficent side probably the most frequently used form of magic is that called garatai which depends on the use of fragments of the food of the person whom it is wished to injure. Another form called talmnatai depends on some object being laid on the path so that the person to be injured shall step over it. I was told of a recent instance in which a man intending to injure a woman who had refused to marry him produced an epidemic which affected a large proportion of the population. The man put together a number of objects including leaves of a certain tree, a lizard, the bones of a man and something from the sea. He mixed these by crushing them together and put them in a piece of bamboo and climbing into a tree he held the bamboo so that the wind blowing from the island of Merlav would reach it and pass on to the village where his enemy was living. Soon after the sickness began, and after a time he was entreated to stop the epidemic. In order to do so, he prepared a number of young coconuts, so young that they could be torn off with the hand, and by some means imparted mana to these. He then called his sister's son and told him to pour some juice from the coconuts in the middle of the village and at the door of every house in which there was a sick person while other juice was to be blown into the air.

Another method was given as one especially used when a man wishes to kill an adopted child who insists on returning to his real father (see p. 52). The man gets a marete or sea-slug from the shore and squeezes from it a reddish fluid and mixes this with the scrapings of the bone of a dead man, with a crustacean called vinvin and with certain leaves. The mixture is put into a piece of bamboo and corked so that air cannot get to it. The man then rubs his hands with the juice

Magic 159

of certain leaves in order to protect them from the noxious influence of the mixture he is about to use and, taking some of the mixture on his hands, he pats the shoulders or cheek of the boy. In about twelve hours sores break out on the places which he has touched and these kill the boy\

So far there is nothing exceptional in these methods except perhaps that the motive which induces the possessor of magical powers to put them in action should be merely to increase his reputation. A further feature is more unusual. It sometimes happens that a man who has carried out a magical rite with the idea of killing his enemy repents when he sees that his purpose is about to be effected. He may try to save the life of the man who is dying at his hands, and in order to do so he commissions his brother or some other man of his family to go to the house of the dying man and give him to drink a mixture of the milk of a coconut and the juice of certain leaves. The sick man will drink it and recover but will at once know the cause of his illness and will seek to discover at whose hands he has suffered. The intermediary must never confess by whom he has been commissioned to bring the healing mixture and rather than do so, he will pretend that he has himself been the worker of the spell, even at the risk of the injury or even death which his pretence may bring on him. A man who undertakes this duty knows that he does so at the risk of his life. When the sick man has completely recovered, his relatives present a large sum of money to the intermediary who is said to matovul this money. The intermediary will usually give some of the money to the person who has imposed the spell and he will break off a small piece of the money and buy with it some food which he will eat. This food must be bought and eaten on the same night that he receives the money and if this is not done the whole of the money must be buried, when it will be called som tamate, the money of the dead man.

I can now pass on to the peculiar associations generally known diS parmal. Each association ox parmal is named after the kind of magic employed, as mal garata, talamatai mal, etc. Each of these associations has a head man who is usually its oldest member and one who wishes to join an association is

Mt is possible that there is something really poisonous in the mixture and that it should not be included under magic, but if so, I doubt whether the distinction is one appreciated by the people themselves.

i6o Melanesian Society

taken to the head man and is given to drink the milk of a young coconut in the stage called ihal, the head man or someone else already a member drinking of the same coconut. If several men join at once, they all drink of the same coconut, the common drinking being called oloolo. The initiates learn the magical method which is the special property of the association, such as garatai or talamatai, and have in future the power of putting it into effect, but they may only do so on those who do not belong to the same association, on those who have not drunk of the same coconut. It is this drinking from the same coconut which is regarded as the bar to any harmful action and it is extended to all those who have become members of the association by means of the rite of drinking a coconut together.

The old members of the parmal learn the name of the initiates who have now become protected from their magic and the initiates are told the names of the other members of the band. Occasionally a man asks that his name shall not be revealed to new members, his motive being that he is afraid of its being known that he has the power of carrying out magical rites, and to keep this secret he is prepared to risk the chance that the new members may put the magical machinery in action against himself. I was told that if anyone attempted to put the magical powers in action against one of his own parmal, the punishment would be death.

The initiate makes a payment on his entrance which goes to the man who admits him and the amount of the payment is decided by the head man, but it is not usually large. If a man makes an unsuccessful attempt to carry out the magical rite of his pa7'mal, he may ask the head man to try if he can be more successful, and in consequence the head man of di parmal is the first to be suspected if people have any reason to believe that they are suffering from its special form of magic.

As a rule the members of a parmal are related to one another and belong to the same veve or moiety of the community. Not uncommonly a man belongs to more than one parmal of the same kind and thus protects himself against the magic of a large body of his fellowmen and when, as usual, there are only two associations on the island, he may thus protect himself against the magic of any member of his

Magic i6i

own community. Further, it is the fashion to join \hQ parmal of other islands and thus obtain a still wider protection.

Often there are only two parihal of a certain kind on an island and in this case the second /a;rw^/ can only be entered at the expense of the rule that members of a /<a:r;^^/ belong to the same veve. In consequence it is customary to make the entrance fee much larger for a member of the other veve. Thus, when the brother of Virsal wished to join t\Q parmal to which Leveveg and Rivlava belonged he was only allowed to do so on payment of a large amount (see pedigree on p. 27).

As I have already mentioned, there are many varieties of parmal, each for a special kind of magic, and it does not follow that a man belonging to one association, say the mal garata, will have the same associates in another, say the talamatai mal, but so far as the members of his own island are concerned this is so to a great extent, owing to the rule that members of the same veve belong to the same association.

It seemed that the relations between members of the same parmal are not limited to mere protection against each other's magic but the members help each other in different ways and form a band for mutual assistance. It would seem as if a society for mutual protection against magic was becoming one for mutual help generally. It seems quite certain that these associations are wholly independent of the Stikwe and Tamate societies though it is possible that the formation of the protective associations has been suggested and their growth promoted by the existence of the Sukwe and Tamate organisations.

Mot lav. In this island associations exist for mutual protection against magic like those of Mota and, as I have already mentioned, inhabitants of one island not uncommonly belong to the associations of another. The Motlav name for the associations is vivtig leva karat. I learnt more in this island of the details of the methods which are practised by the members of these bands. If a member wishes to injure an enemy, he goes to the head man who commissions one of his followers to obtain something from the man to be injured, whether fragments of his food, his hair, nail-parings or excrement. It is usual to use food and the method employed if fragments cannot be otherwise obtained is the following : the man commissioned to obtain food roasts a yam and breaking it in half he gives one half to the man who is to

R. M. S. II

be injured while he eats another piece of yam himself which the victim is led to believe is the other half of his own. As a matter of fact the piece of yam which is eaten has been substituted for the real other half of that eaten by the victim which is kept for the magical rite. Sometimes the device is discovered, when a terrible fight will ensue.

The fragment, whether obtained in this or some other way, is called niliao and is crumbled up and mixed with the scrapings of the bark of a scented tree called nurnem and the mixture called lakwor is wrapped in a leaf and the package roasted over a fire. After being well turned over as it is roasted, the package is taken to the beach, buried in a hole dug in the sand and covered with a stone. It is buried in a place where the sea will flow over it. At the end of a month the package is taken out, the leaves of another tree are mixed with the contents and it is re-buried to remain now for two months. Then it is taken up again and the bone of a dead man is roasted over a fire till it is friable. The bone must not be touched with the hand but is broken up with a stick and the powder mixed with the other contents of the package. On this occasion words are chanted, nothing having been said on the previous occasions. The package is not now returned to the sand but is buried in the ground inside the house of the man carrying out the rite and a fire is lighted on the top of it. It stays here for four months and it is during this time that the victim first begins to feel ill. He will be ill for a time but will recover and begin to go about again. At the end of four months the package is taken up and its contents are mixed with small fragments of a substance called nerlam, probably red branching coral, which has a special reputation for mana. The package is then hung up in the house and a fire lighted under it while words are again chanted. My informant had heard the chant but did not know the words. The package is hung up in a part of the house partitioned off from the rest but it is taken down and hidden when any stranger is present. By this time the victim will be seriously ill and unable to go to his garden and he continues in this state till the practice of a further rite which causes his death. For this purpose the magician goes to a special coconut tree with a reputation for mana and takes five coconuts at the stage called mel (corresponding to the mat of Mota). The package, which has by this time reached

Magic 163

a considerable size, is opened and a deep depression is made in its contents. Into this depression are poured the contents of the coconuts, this process being called halmat, and as this is done the man dies. If this final rite is not performed the victim will live on indefinitely in his enfeebled condition. At the end all the objects which have been used are thrown away. It will be noted that this process lasts for at least seven months and that the victim does not begin to suffer till the process has already been in action for three months.

Of beneficent forms of magic several examples were given. To increase the crop of yams or other food supply, the process called serwehigi is put in action. A man wishing to ensure a good crop goes to a man with the necessary power which depends on the possession of a stone with mana for this purpose. On this island he pays money, first ten lengths called letnovivhe, this corresponding to a cubit measured from the elbow to the tips of the fingers. After giving this he asks, " What next .-* " and is asked for three lengths from the tips of the fingers to the opposite shoulder. He gives it and again asks, " What next ? " Two full fathoms are then demanded and when these are given the worker of the rite is satisfied. The pair then go together to the stone, on which the man lays the money which has been paid ; the worker of the rite takes up the money and the other man lays his hands on the top of the stone. The worker of the rite then addresses the vui residing in the stone, asking or demanding that the crops of the visitor shall be blessed. " Do thou go with this man to his house and bless his food so that it may be plentiful ! "

There are similar rites to increase the supply of pigs, fish and flying foxes. In order to promote the fertility of pigs a special stone is buried ; to increase fish the appropriate stone is put in the surf while formulas are uttered, and to increase the supply of flying foxes the juice of certain leaves is squeezed over a stone shaped like a flying fox while certain formulas are said.

The method for making rain was described much more fully than in Mota. Here the rite is done in the house in the middle of which a great clam-shell called nabeng is placed, and on either side of this shell large stones are put to represent the big clouds rising from the horizon which are called gerger. Other stones are then added to support these so as

to make the clouds thicker and thicker. Between the stones and the shell umbrella-palm leaves are put to represent the sky full of clouds. The rain-maker then goes to the bush and scrapes off pieces of the bark of the dirbini tree. These are brought to the house, mixed with water and put into the clam shell, together with leaves of the rolemblemb tree, the two things signifying rain-clouds. A red fruit called vil is put into the shell to represent lightning and a special kind of stone to represent thunder. The shell is then covered with the leaves of the umbrella-palm, leaving an opening through which the rain-maker chants into the shell the words, ''nok wengilgili, nok wengilgili,'' and when these words have been said, the opening is completely closed. Wengilgili is the name for rain which comes down so hard that it furrows the soil and nok is the first person possessive pronoun. This completes the rain-making process, and when it is desired to stop the rain the leaves of the umbrella-palm are taken away.

I was told that in Motlav the stones used for these various purposes might be sold, the appropriate formula being taught at the same time. It was clear that the passage of money and of the formula was essential for the proper working of the rite in the hands of its new possessor and that it would therefore be useless to steal the stone as seems to happen in Mota. The way the matter was put was that the stone would have no mana unless it was bought.

Possession.

The only information I have on this topic has been sent to me by Mr Durrad who obtained it from John Pantutun. Certain forms of illness are due to possession by a tanmte (a ghost) or a vui (a spirit which has never been a human being). The effects of the two kinds of possession are quite different. Illness may also be due to capture by a vui of the ^/az which, following Dr Codrington, may be translated "soul."

Possession by a tamate. I f a person meets a hostile ta7nate it may enter into him, and if several tamate enter in this way the man becomes raving mad, of enormous strength and of incredible swiftness of foot. He is caught with difficulty, mastered and held down by a number of men. A fire is lighted and bunches of a scented shrub called sav are heated over the flames to increase their smell and then laid over the

Possession 1 65

sick man's nose. This treatment is applied again and again and the sick man calls out the names of the tamate within him, being violently convulsed as he utters them. At last he becomes calm and limp, and after a few hours in this condition, he revives and becomes well.

A child possessed by a tamate does not become mad but is taken with sudden and violent illness and cannot open its eyes. In this case it is uncertain whether the illness is due to possession by a tamate or vui or has some other cause. The people summon an expert called gismana who detects a case of ordinary illness at once, but if he decides that it is a case of possession, he strokes down the body of the child with both hands and blows upon its eyes, spitting and yawning as he does so. If the child is possessed by a tamate, the gismana is able after a time to call out its name and extracts it from the child on the same day.

Possession by a vui. In this case the symptoms are the same whether an adult or a child be affected. The treatment takes a longer time than in the case of possession by a tamate, for the gismana has to find out the place inhabited by the vui which has entered into possession of the sick person and this may take several days. When the gismana has discovered the place of the vui, he seeks out the man whose " sacrificing place " it is and asks him to recall the vui to its usual haunt. This man goes to the place and calls on the vui (there may be more than one) to return and the vui leaves the sick man who then recovers.

Capture of the soul of a vui. If the gismana decides that the atai or soul has been taken from a sick person by a vui, he will undertake to bring it back. He makes a potion of leaves and drinks it before going to sleep. In his sleep he dreams, which indicates that his own atai has left his body to go in search of the vui. The soul of the sick person may have been hidden, perhaps on the branch of a tree or in a hole in some rock, and the vui makes every effort to retain it, blocking the way of the soul of the gismana by making a barrier of stones or raising some other obstacle. The soul of the gismana, on the other hand, makes every effort to seize the soul of the sick person and will sometimes be assisted by a friendly vui who will take the soul of the sick person from the hostile spirit and hand it to the soul of the gismana. Sometimes the gismana may try in vain, and in this case he

i66

Melanesian Society

announces his failure when he wakes in the morning. If he succeeds the soul is brought back to the village and returned to the sick person. While the gismana is asleep no one must disturb him, for even if his atai had succeeded in seizing that of the sick man and bringing it back as far as the village, the awakening of the gismana would make his atai relax its grasp and the soul of the sick person would slip away and return to the hostile vui. In this case the gismana drinks the potion again on the following night, but on this occasion his soul does not need to search for the hiding-place but goes straight to it to recover the soul of the sick person.

Money.

The money of the Banks Islands consists of shell-discs strung on the fibre of hibiscus. There are two kinds, one now rare, consisting of finely ground and thin discs (Fig. 6 a) ;

Fig. 6. a, the finer, and b, the coarser shell-money of the Banks Islands ; c, the shell from which the money is made ; d, side view, and e, under-surface of fragment to be ground down; /, the completed discs; g, the cut surface of the stick showing the depression made by the discs. (All nat. size.)

the other (Fig. 6 <5), much coarser, which is in general use. The discs are made from a small shell called som (Conus ceylanensis Hwass^) (Fig. 6 c), and the money itself is also called som. It is manufactured chiefly on the small coral islet of Rowa, but some is also made on Saddle Island or Motlav. The shells from which the money is made are gathered by the men, put in baskets, and taken to the village where they

• I am indebted to Mr Hugh Watson for this identification.

Money 167

are put in the sun to dry till the molluscs they contain are dead. Then the men's part of the work is finished and the actual manufacture is carried out by women. They break the shells by hammering them with a piece of the giant clam-shell called chele (the talai of Mota), this process being called vil pulpul. The fragments of which the pointed ends of the shells form part are dried in the sun, washed in fresh water and then chipped with a sharp piece of shell till they form small pieces, pointed on one side corresponding to the outer surface of the shell, and hollow on the other side corresponding to its interior. The side view of a fragment is shown in Fig. 6 d and the hollow under-surface in Fig. 6 e. The chipping is called vit tartar.

The next stage is to rub down the pointed surface of the fragment on a stone, already well polished and ribbed by long usage. The stone is sprinkled with water and strewn with black volcanic sand brought for the purpose from Vanua Lava. Then the woman fetches or cuts freshly for the purpose a stem of the kwagala (a variety of hibiscus), about as thick as the finger and about four feet in length, and putting a fragment of shell on one end of the cut surface of the stick with its hollow side towards the wood, the end of the stick and the piece of shell are rubbed a few times to and fro on the stone so that the tip of the shell becomes worn down and the fragment converted into a disc (Fig. 6/") with a hole in its centre due to the hollowing of its other aspect. As each fragment is thus finished it is dipped in water and put with others in a coconut cup ready to be strung together when dry. The pressure of the piece of shell produces a depression on the end of the stick, shown in Fig. 6^, and after the first use this depression is ready to receive succeeding fragments of shell. The thread on which the discs are strung is called gavw and is made both by men and women from the fibre of the tree called vir (the var of Mota), a variety of hibiscus. A piece of the stem of this tree about six feet in length is cut off and the bark taken off gently and put in the sun to dry. The outer surface of the bark is then pulled off and the inner aspect is found to be made up of strong fibres which are used without further preparation.

I saw the latter part of the process of manufacture at Rowa and nothing could have been simpler and more expeditious than the conversion of the fragments of shell into

the discs, advantage being taken of the natural hollowing of one side to rub down only the other side of the fragment. Dr Codrington has described^ a somewhat more elaborate method in which a pointed stick is used and the fragment of shell ground smooth first on one side and then on the other. I cannot say whether this is the method of making the finer money or whether it is only that the method has been simplified since Dr Codrington's observations were made.

The money thus made in Rowa is exchanged for yams and taro obtained from Vanua Lava and from the other islands of the group with the exception of Merlav, Merig and Santa Maria. All the vegetable food on Rowa is, or was, obtained from elsewhere, the other chief article of diet being fish, of which a good deal is eaten.

I endeavoured to obtain some estimate of the rate of exchange and of the amount of labour necessary to obtain food, and though my information is very incomplete I give it here as an example of the kind of inquiry which further opportunity might have rendered possible. A woman working for one day from sunrise to sunset will be able to make a fathom of money. In exchange for yams the people generally use lengths of money called ckiregi, measured from one shoulder to the tip of the outstretched fingers of the other hand, i.e. about four feet, and for this length they will obtain ten yams which will provide vegetable food for five persons for a week. This seems a small amount, but I was told that the people of Rowa eat much less than those of other islands.

People often like to have new money and in order to obtain it will give double the quantity of old, and it would appear that the Rowa people profit by this custom to obtain a larger number of yams than they would get for old money. I was told that the new money was preferred in order to make more impression when giving it in ceremonial connected with the Stikwe, but it is doubtful whether the money is given when it is new or is kept till it is old before it is used. In the latter case, it would seem as if the preference for new money is due to a man liking it to be known that he has it and to his pride at its being seen in his house.

Various units of length are used but there are certain complexities in the native table of measures which I was not

• M., 325.

Money 1 69

able to understand satisfactorily. When describing transactions of the Sukwe, it was continually stated that so many rova or fathoms would be given, but in some islands, not apparently in all, this meant that double the number were really given, a man giving twice the amount which had been arranged. So far as I could tell this ambiguity is connected with the method of measuring shell-money. The usual way is to put two pegs in the ground a fathom apart and to wind the money round them, one loop of money being called a tal (tal, to go round), and there seems to have come about a sort of understanding that when a rova of money is required, it is a /^/ which must be given ; double lengths are measured but only one side is counted, the usual method being to take the loops off the pegs and hold them up by one end, only those of one side being counted. Sometimes the double rova is called a rova totogoa (Mota), or rov totgo (Rowa), literally a full stretched fathom.

In monetary transactions other than those of the Sukwe it would seem that the term ial is used in a way not exposed to the ambiguity just considered, but is definitely a double fathom ; thus, if the owner of a pig asks for taltavelima (chilchiviliem in Rowa) or five tal, the buyer will give him ten fathoms. It would seem as if in the Sukwe transactions of some islands, there has come about the practice of giving a tal in place of a rova which is properly the term for fathom, but that the double length has still continued to be called a rova. It seems quite clear that the confusion is not due to misunderstanding on my part, but that there is a definite ambiguity to the natives themselves, for I was told that the people have long had the idea of altering their practice so that they would give one rova when they say they are going to give one rova. There are still many who prefer the old style and oppose the change, and so far nothing has been done.

It is possible that the whole matter is in some way connected with the practice of giving interest on loans. In the account of the ceremonial of the Sukwe it has been seen (see p. 64) that there is an organised system of giving or lending money to friends with the idea of receiving it back with cent per cent interest, and this rate is given for loans of the more ordinary kind, the amount of the interest being the same whatever the interval which elapses between the loan and its repayment. Here again we have a doubling of an

amount of money which has a suggestive resemblance to the doubling of the rova in the Stikwe ceremonial.

A measure of length already several times mentioned is that from one shoulder to the tips of the outstretched fingers of the other hand. This was often spoken of as a short rova but it is also called vu^o or leme gambak in Motlav and chiregi in Rowa. Another less frequently used measure is from the elbow to the tips of the fingers of the same hand called in Motlav letnovivhe. Another shorter measure is that which in Motlav is called tarak, probably the length from the wrist to the ends of the fingers.

In a ceremony after death (see p. 59) strings of shellmoney are drawn away from the corpse and the name of this ceremony, rave epa or " draw mat " suggests a survival of the time when mats were money in the Banks Islands as they still are in the northern New Hebrides.

Canoes.

I obtained an account of certain interesting occurrences which in Mota and Motlav and probably in other islands accompany the manufacture of a new canoed People from any part of the island may help to cut down the tree from which the canoe is to be made and the tree is drawn to the village, those who have assisted being rewarded with a feast. The tree is hollowed out to make the body of the canoe in the village and the canoe is then drawn down to the beach where the people cut the three yoke-pieces (iwatid) of the outrigger (sama) and fix them in holes in the side of the canoe and those helping are then feasted. After the feast the people tie on the outrigger and make the platform (kweakwea), and then they launch (salsale) the canoe. The men who have made the canoe put it in the sea and wade in with it till the water reaches their waists. After turning the canoe round three times, the men stand in two rows, one on each side of the canoe, and those on one side push the canoe towards those on the other side who push it back towards the former, and this is done three times. (It may be noted that the composition of these two rows of men has no relation to the two moieties of the community, each row being composed of men of both veve.) Then all the men stand round the canoe and

• For an illustration of a Banks Islands canoe, see Codrington, M..^ i(yi.

Canoes lyi

each in turn puts his hand on it and passes it on to his neighbour till it has passed round the whole circle, and this again happens three times, the ceremony having a striking similarity to that of handing round a newly born child in Motlav (see p. 148). Then the people put sand on leaves and place the leaves on the platform of the canoe and light a fire on this sand.

Then men take their places in the canoe for a trial journey. The canoe is paddled to the chief landing-place {tursao) of the next village where all the people of that village assemble with bows and arrows. If a man has no bow he should make one for the occasion, or he may arm himself with stones. As the canoe approaches all stand either with drawn bows or with stones in their hands till one of the crew raises a paddle as a sign and then all in the canoe paddle their hardest to reach the land, while those on the shore discharge their arrows and fling stones at the approaching boat.

Directly the land is reached the attack ceases and food is given by those on shore to the men in the canoe. The canoe is then paddled to the landing-place of the next village where a similar scene is enacted, and so round the island till the people reach the landing-place of the village from which the canoe started, where for the last time the canoe is shot at and stoned. The shore of Mota is raised in most places, even at the landing-places, so that those on the land often stand at some height above the water and are able to shoot or fling down at the canoe, and it is said that severe injuries are often inflicted on the crew of the canoe. The object of the whole affair is said to be to make the canoe sink to test it and see how strong it is, but this native explanation must be received with caution.

Canoes are generally named after the trees from which they are made, such as siwo (breadfruit), salte (Catappa) or tora. Thus, a man asking for the loan of a canoe will ask for the tora or the salte. The man with the steer-paddle in the stern of the canoe acts as captain.

One striking feature of life in canoes at sea is that all men become equal. A chief or "big man" on land is no more important than anyone else when he is at sea. Thus, if a canoe sinks at sea and some one in a canoe which remains afloat says, "Take in that 'big man,'" others will say, "Let him swim. He is not a 'big man' when he is out at sea."

A man who is going for a voyage of any length will take only his own sogoi with him, for if he took members of the other veve and anything happened to one of them while away from the island, the sogoi of the sufferer, i.e. the members of the other veve, would demand compensation and there might be much trouble. An exception to this rule is made in the case of the sister's or daughter's husband, who should always go in the canoe of his wife's brother or father. Thus, if John or Robert Pantutun were going out in a canoe, Simon would go as a matter of course. It would not be necessary to ask him and, if he did not go, he would be generally despised.

The fighting which has been described only occurs at the island to which the canoe belongs. There would be nothing of the kind at the first visit to another island. Even in the old days no very long journeys were made by canoe. Thus the people of Mota would go as far as Santa Maria or Ureparapara but not to Merlav or other distant islands in that direction. The people never travel by night and in consequence never use the stars in navigation and in correspondence with this I could obtain no definite information about the names of stars or constellations : a few names were given but my informant could not point out the stars which they were said to denote.

Decorative Art.

In the account of the Sukwe and Tamate societies I have given illustrations in several places of patterns and ornamentation which may now be considered as examples of decorative art.

The pudding-knives shown in PI. V, VI and VII give a very good example of transition between geometrical and anthropomorphic or zoomorphic figures. The designs of most of the knives are evidently based on the human figure, and those marked A i to A 6 appear to furnish a fairly continuous series. A 4 would seem to be most naturally explained by the disappearance of the head, and if this transition be accepted the other two knives would show still further conventionalisation*. The fact, however, that these knives belong to different ranks of the Sukwe makes one hesitate in supposing A 4, 5, and 6 to

' For the justification of the use of this term see Chapter XXXI.

Decorative Art 173

be directly related to A i and 2. The knives numbered B i to 5 go together not only in their general resemblance but also in that all belong to the Mwele rank of the Sukwe. The fact, however, that B i has a distinct covering on the head which is absent in the others raises an element of doubt. The shape of the head is also much the same in B 6, so that this knife is probably one of the same series with which there has been blended a different pattern.

The knives numbered B 7 and 8 and C i to 5 show sufficient resemblance to one another to make it probable that they also have been based on the human figure, though there must be doubt about C i which possibly belongs to an altogether different series. Of the others C 2 perhaps departs most widely from the human pattern, but its anthropomorphic nature is rendered probable by a study of the belts D and E in PI. XI. There is a fairly definite resemblance between the middle portions of the designs of the belts and those of the knife.

Four knives figured by Mr Edge-Partington^ are shown in PI. VII, E. The knife E 3 has a general resemblance to A I and A 2, but the very different shape of the headcovering makes it probable that it belongs to a different series. E 2 almost certainly belongs to the same series as B 8. E I probably goes with A 4 to 6, while E 4 cannot be definitely linked with any of the others. Taking all the knives together there are at least four different forms of head-covering and there can be little doubt that these are to be connected with the tamate or hats of the different ranks of the Sukwe. It is these different head-coverings which make it probable that, if we had a much larger number of knives, we should be able to arrange them in a number of series corresponding to different ranks of the Sukwe, and it is possible that other features of the designs, such as the rounding or angularity of the conventionalised arms and legs, would be found also to be connected with the different kinds of head-covering.

One or two more points may be mentioned. A comparison of A 2 and 3 shows very clearly the transition from the outturned hands in A 2 to the condition in A 3 which might be taken merely for a ridge designed to separate the handle from the blade of the knife. B 6 shows a condition which approaches

^ Man, 1 909? IX, 182.

very nearly to the out-turned hands of A 2, and in B 5 traces of the hands or feet still remain above the ridge. Similarly, in Mr Partington's E 3 the formation of the ridge by the feet is well illustrated. The first of his knives shows a doubling of the ridge and this feature is also present in B 3.

There is a definite resemblance between the faces on these knives and those of the stone figures which form the part of the gamal shown in PI. Ill, Fig. i, though there is a very striking difference in the posture of the arms. In all the knives the hands rest on the hips while in those of the stone figures they are held above or beside the head, or with the elbows bent at a right angle with the hands in front of the body. Another point of resemblance is that the stone figures have definite coverings on the heads which are of at least two different kinds.

Though the designs of most of the knives are certainly anthropomorphic, this is not so in all. The knife of D i in PI. VII was definitely said to represent the til or sword-fish S the serrated handle representing the snout of the fish and the hole between handle and blade its eye. Another of the knives D 2 resembles it so closely that it must also represent a fish, and probably the til, in spite of the protuberance at the end of the blade. As already mentioned, it is possible that some of the other designs, and especially that of C i, may have a motive other than the human figure.

The designs on the belts of PI. XI are all representations of the human figure. The anthropomorphic nature is least obvious on the tamate worawora, but this is partly due to the fact that the design on the belt only represents the body of the tamate, the head being tattooed on the wrist of the wearer (see p. 134). The tironin belt is modern and its special feature of interest is that while the design at each end is a simple human figure, those in the middle are complicated by the addition of conventionalised designs between the head and arms, one representing an additional head, while the other probably represents additional arms. The kivat kwoe belt shows a design in the position of the hips which probably represents the out-turned hands and fingers, and then it would seem as if circles had appeared in the situation of the hands, which may possibly furnish the explanation of the circles

• It may be noted that the /// gives its name to one of the Tamate societies of Motlav, the members of which may not eat this fish (see p. 120).

Decorative Art 175

between the figures of the tamate worawora. The tamate kwat belt is noteworthy for the design separating the figures placed feet to feet, which is probably derived from the genital organs of the two figures.

The figures shown in PI. IX representing Tamate viov are certainly the results of a conventionalisation of the human face which has proceeded still further than in the figures of the belts. It is clear that the design of Tamate viov represents a face with the two eyes. The design of the Tamate liwoa suggests the human figure strongly, but I have no definite evidence that this is so, and it may be that the resemblance is misleading. This plate also shows a number of small human figures, most of them placed in pairs with the heads in apposition, and these illustrate in a very interesting manner the result of conventionalisation.

CHAPTER VII

THE TORRES ISLANDS

The people of the Torres Islands form two distinct groups with definite differences both in language and custom. One group inhabits the islands of Loh and Toga and the southern part of the island of Tegua ; the other group occupies Hiw and the northern part of Tegua. Though the people of Loh and of the southern part of Tegua form one group so far as language and custom are concerned, their relations with one another are not friendly. The island of Loh has two parts separated by an isthmus, and the story runs that a man of the northern half once produced a drought by acting on the sun, and all the people of the southern part of the island had their gardens scorched and, in consequence, drove out the people of the north to Tegua, where they have since lived. The hostility between those who remained on Loh and those who were driven out has continued ever since and is constantly breaking out in warfare. Till recently no intermarriage took place between the people of Loh and Hiw, but since the return of labourers from Queensland this rule has been breaking down.

Though there are many points of similarity between the cultures of the Banks and Torres Islands, — a similarity which will be obvious in connection with relationship, — the social organisation of the two groups of islands differs considerably. Instead of the two divisions of the Banks there are three in the Torres Islands, and each of these is again subdivided. The three main divisions are called Terhar, Gameljat and Gamelteriiata, and each is divided into sections called 7netaviv. Four metaviv were given for the Temar called Tumun, Telemot, Tetaleta and Tepepukul. The Gameljat also had four metaviv called Tahar, Tenemera, Teletaluwo and

Social Organisation 177

Takalala. The Gameltemata, on the other hand, had only two metaviv, the Nihega and the Riwangjap or Tarovor. The exact social significance of these subdivisions could not be ascertained, and I do not know whether they play any part in the regulation of marriage.

The equivalent of the Banks word sogoi was said to be tutuma, but I did not ascertain whether this term is applied by a person to the members of his own main group, or whether it is limited to members of the same metaviv.

It was said that the Gameltemata are more nearly allied to the Gameljat than to the Temar and this statement and the similarity of name suggest that the smaller of these divisions, the Gameltemata, may have split off from the Gameljat, and thus produced a threefold division, instead of the usual dual organisation of this part of Melanesia. On the other hand, it was said that, when a man of the Gameltemata marries a woman of the Gameljat, he thereby acquires some tutuma or clansmen among the Teriiar but the significance of this statement could not be ascertained.

I was given the names of a large number of animals which are not eaten, and these included the shark, the sea eel and the octopus, but in every case a definite motive was given for the abstinence. Thus, the shark is not eaten because it is believed that those who eat it would be caught by a shark when in the sea, and the eel is avoided because it is believed to be poisonous. Further, it seemed clear that the abstinence is observed by the whole population, and is in no way different for the members of the different social groups.

The System of Relationship. Loh.

The following are the terms used in this island : —

Ma. Father, father's brother, mother's sister's husband and more distant relatives falling in the same categories according to the usual rules of the classificatory system, such as the father's father's brother's son, etc.

Reme. Mother, father's brother's wife, and mother's sister corresponding to the use of ma. It is also used for the father's sister, and for the wife of the mother's brother. Like ma, it is used in the usual classificatory sense.

lyS Melanesian Society

Magola mino, n'etuk and nak magola. These are terms for child, male or female, and are used for the actual son or daughter, for the son of the brother whether man or woman be speaking, for the sister's child of a woman, and for the child of the husband's brother, of the wife's sister and of the husband's sister.

Tigik. This is a general term for brother, whether older or younger, and contrary to the frequent rule is used by both males and females, but its use for a brother when a woman is speaking is probably irregular. The term is also used for the son of the father's brother and of the father's sister, as well as for the son of the brother and sister of the mother, so that it is used for a first cousin of any kind as well as for more distant relatives according to the usual classificatory usage. Tigik is also a term which may be used by a woman for her sister's husband.

ChiocJiiok. This is the term for sister used by a man, and is applied by him to the four kinds of female cousin, corresponding to the use of tigik for males. The term rriay also be used for the wife's sister.

Takalek. This is a term used by sisters to one another. I did not definitely ascertain that it was also used for cousins of various kinds, but probably this is so.

Meruk. A reciprocal term applied to one another by mother's brother and sister's child.

Tukwuk. The term for the father of the father and of the mother, i.e. for both kinds of grandfather ; also for their brothers and more distant relatives of the same generation as the grandfather, according to the usual classificatory rules.

R'epu, The corresponding term for the mother of the father and of the mother and more distant relatives of their generation.

Tukwada. The general term for grandchild, the reciprocal of both tukwu and r'epu.

Kwiliga. The term for father-in-law and son-in-law. It is applied to the father of both wife and husband and is used both by men and women. It may also be used for the mother of the husband.

Tivina. This is the term for the wife's mother, and for the son's wife when a man is speaking.

Recha or reja. This is a reciprocal term for husband's mother and son's wife, i.e. it is applied to women by one

another. This term is also used between husband's sister and brother's wife, woman speaking. The husband's mother may, however, also be called kwiliga. The term recha or reja is also applied by a woman to the father's sister of her husband.

Wuluk. This is the reciprocal term for the wife's brother and the husband of the sister of a man, i.e. it is used between men. It is also used for the husband of the father's sister, but in this case its reciprocal is magola or child.

R'ewoluk. This is a reciprocal term for the husband's sister and the wife of the brother of a woman. Thus, it is applied by women to women just as wuluk is applied by men to men. These relatives also call one another recha or reja. The husband's brother and the wife of the brother of a man are addressed by name, and personal names may be used between the wife's sister and the sister's husband, who also call one another chiochiok and tigik (sister and brother). The wife of the sister's son is also addressed by name by a man.

Tukwutog. A term given by the grandfather of a man to the grandfather of the man's wife.

Retukwu. Applied to one another by the respective grandmothers of a married pair.

Wotagavu. This is a term applied to relatives who belong to different clans ; thus, if two brothers of the Teniar married women of the Gameljat and Gameltemata respectively, the group of four people would be wotagavu.

The Loh system has several features of interest. First, the frequent use of tigik for a brother whether man or woman be speaking. Secondly, there is no distinction in terminology between the four different kinds of cousin, a fact to be considered more fully later in connection with the presence of the cross-cousin marriage in the other part of the islands and its occasional occurrence in this part. Thirdly, the terms chiochiok and tigik are used for the wife's sister and the sister's husband respectively, although these relatives may also address one another by name. The significance of this will also be more fully considered later. Fourthly, sex is distinguished in the terms for grandparents and parents-in-law. Lastly, the system is unique, so far as our information goes, in having special terms for the relationship between the grandparents of a husband and wife.

There is an evident relationship to the Banks Islands in many of the terms, such as kwiliga, wuluk, meruk and tigik.

12 — 2

i8o Melanesian Society

Hiw.

The following are the terms of this island : —

Mamakia. This term, meaning father or rather "my father," is also applied to the father's brother, and to those whom the father would call brother according to the classificatory principle.

Tatakia. Mother, mother's sister, wife of the father's brother and others in the ordinary classificatory way.

Moaiakia or megoiakia. Child, used generally as the reciprocal of mamakia and tatakia, and also for the brother's child by a woman, i.e. as the reciprocal of father's sister.

Teiik. Brother when a man is speaking ; also used for the son of the father's brother and mother's sister.

Tutuak. The reciprocal term for the brother-sister relation : also used for children of father's brother and mother's sister when of different sex.

Maruk. The reciprocal term for mother's brother and sister's son.

Weiuk or weyuk. The term for the child of the mother's brother and of the father's sister, i.e. for cross-cousins. Also used by a man for the wife's brother and sister's husband.

Pupukia. Used for all four kinds of grandparent and the corresponding four kinds of grandchild.

Tauwenakia or takwenakia. Husband.

Tukwinnekia. Wife.

Kwiiga. A reciprocal term used for the father and mother both of the husband and wife, and the four corresponding kinds of son- or daughter-in-law. The term is applied to the wife of the mother's brother who may also be addressed by name. The father and mother of the wife are also called pwoigak, but this may be only a dialectical variation.

Woiuk. This is a term for the husband's sister and reciprocally for the brother's wife of a woman, and is used for the husband of the mother's sister (m. s.). It is probable that weyuk and woiuk are simply variants of the same word in which case one term would be used for the children of mother's brother and father's sister, for the wife's brother and the husband's sister and their reciprocals, and for the mother's sister's husband. If this be so, the difficult point to understand is that in the case of relatives by marriage the term would be

used only between persons of the same sex, whereas in the case of cousins, it is also used between those of different sex.

The wife's sister and the husband's brother are addressed by name, and the personal name may be used for the children of the mother's brother and father's sister, who are also called weyuk.

There is no term of relationship ordinarily used for the father's sister, but she is addressed by her personal name except when her brother's son has married her daughter.

The Hiw system differs in several important respects from that of Loh. The customary distinction in the terms for brother and sister according to the sex of the speaker seems to be habitual in this island, and there is definitely present the distinction of the children of mother's brother and father's sister from other cousins which is absent in Loh. On the other hand, no distinction is made in Hiw between grandparents of different sexes, nor between parents- and children- in-law of different sexes. The application of the usual term for parent-in-law to the wife of the mother's brother is evidently a consequence of the cross-cousin marriage, and may only hold good when this marriage has occurred. It is clear that men address their sisters-in-law by name and that this is reciprocated, and it is probable that men address their brothers-in-law, and women their sisters-in-law, by the same term as is used for cross-cousins.

The foregoing terms were obtained by the genealogical method and with less ample pedigrees than I should have liked. Though they are doubtless correct in essentials, they will probably require some modification in detail for there is one feature of Torres sociology which introduces a difficulty in the use of pedigrees as a means of collecting terms of relationship. In these islands marriages between relatives are very frequent (see p. 185), and in consequence two persons are often related in more than one way. Thus, if a man marries the daughter of his mother's brother, his uncle becomes kwiliga or kwiiga instead of meruk or maruk though the latter term may also continue to be used, while the uncle's wife, previously called reme, becomes tivina. Similarly, in Hiw a brother may become kwiiga by the marriage of a woman with her brother's son. These double relationships make it necessary to use many pedigrees in order to obtain the exact connotation of the terms.

Functions of Relatives.

No very full inquiry into this subject was made but several points of great interest came out incidentally. The mother's brother stands in a close relation to his nephew and the latter may take any of the possessions of his uncle. The father's sister can forbid the marriage of her nephew in Loh, while in Hiw marriage may, and in the old days used frequently to, take place between these two relatives. This marriage is not now allowed in Loh.

Numerous regulations concerning behaviour towards relatives by marriage were recorded in Loh. In this island a man has a special regard for his wife's father. If the father of the wife passes by, the son-in-law need not stand up unless the father-in-law speaks to him but in general if a man is sitting down his wife's father will not approach but go out of his way. If it is the father-in-law who is sitting down, the younger man will not approach him closely, even to tell him anything, and if he speaks the father-in-law will not get up. If a man is carrying anything on his shoulder and calls to his son-in-law to take it, the latter will not take it directly from the other but will ask him to put it on the ground whence he will take it. It was said that the father-in-law would have no objection to the load being taken directly from his shoulder but that the younger man would be too shy to do so. If a man is sitting and his daughter's husband wishes to pass him, he will ask his father-in-law to get up, but if the latter does not wish to rise, he may meet the occasion by turning his face away as his son-in-law goes by. The behaviour towards the wife's mother is of the same kind. A man must not go near this relative nor say her name and he may only talk to her from a distance of five or six yards. If for any reason he has to pass her at a smaller distance than this he will do so in a crouching attitude and in no case will he pass her if she is sitting. If, on the other hand, the wife's mother has to pass her son-in-law even at a greater distance, she will go down on her hands and knees and will not pass him at all if he is sitting. It was said that formerly a man might marry the mother of his wife but that such an event would be the occasion of a fight.

The wife's sister may either be called chiochiok (sister) or may be addressed by her personal name, but the mode of

Functions of Relatives ' 183

address is a fact of great significance, for if the personal name is used, there is the suspicion, if not certainty, that there have been sexual relations between these relatives. If, on the other hand, the woman is called chiochiok this possibility is definitely excluded. The same rule applies to the use of the personal name and other regulations concerning the wife's mother. If a man calls his mother-in-law by name or does not observe the rules of avoidance which have just been recorded, the people will at once know that there have been sexual relations between them. It was said that if such relations have actually occurred a man would no longer call his wife's mother tivina but would call her by her personal name for the rest of his life. Sexual relations between these relatives are said to occur at the present time and with this consequence. If a man calls a woman chiochiok it shows that he has no intention of marrying her. It seems clear that the use of the personal name of a woman by a man implies the possibility or actual occurrence of marriage or illicit sexual relations between them. Another sign of sexual relations is the use to a woman of the word cha or ja which should only be used by a man when speaking to his wife. It has the same significance as the use of the personal name. On the other hand, a man may call his brother's wife by name though he will not speak to her familiarly nor use the word cha, but in this case it did not appear that the use of the personal name implied sexual relations. It is probable that this approach to polyandry is now so foreign to the ideas of the people that the use of the personal name is not regarded as important while the relationship with the wife's sister is one which is still often associated with sexual relations.

Those who call one another wuluk are allowed to say each others' names but they will be very respectful to one another.

A woman calls her husband's elder brother by name but if her husband dies, she then calls the elder brother kwiliga and he calls her tivina and they may not marry one another. The widow may, however, marry any of the younger brothers of her late husband. If none of them marry her they call her reme, a term used otherwise for the mother while she calls them magola mino or child. If the widow marries the brother of her husband who comes next to him in order of age, she calls his younger brothers magola mino.

A woman must not take a basket from the head of her husband's sister whom she calls r'eja or rewoluk. Either of these women will rise when the other approaches. The term recha or r'eja is evidently derived from the word cha or ja, implying familiarity, which may be used between these persons.

Those who call one another tukwutog and retukivu in this island are mutually helpful to one another; if one is sick or unable for any reason to leave the house the other will bring food or other necessaries. This relationship is very unusual, if not unique. It resembles that of the Banks Islands in which the respective parents of a married couple call one another gasala, but exists between persons one generation farther back.

In Hiw a man does not say the name of anyone he calls kiviiga. If a man wishes to pass a kwiiga he asks for permission and after giving it the kwiiga will turn aside his or her head as the son-in-law goes by. Any infringement of these rules is punished by a fine of arrows, pigs' jawbones and kava. The mother of the wife will not eat anything which has been carried on the shoulder of her son-in-law but there is no restriction if it has been carried in the hand. If anyone touches a man's head his kwiiga will object, saying that the head of his kwiiga is sacred.

Regulation of Marriage.

The three main social divisions of the islands are exogamous ; a man of the Teriiar is not allowed to marry a woman of that group and so with the Gameljat and the Gameltemata. I have no evidence that the metaviv are concerned in any way with the regulation of marriage. The Torres Islands, and especially Hiw, are noteworthy for the number of cases in which those related by blood may marry or are even expected to marry.

In Hiw there appeared to be a definite connection between marriage with the father's sister and marriage with her daughter. It was said that if the father's sister was young enough a man would marry her, but if not, he would marry her daughter. The cross-cousin marriage was spoken of as a kind of substitute for marriage with the father's sister. If a man marries the daughter of his mother's brother he does not pay

Marriage 1 85

anything for her. A man desires especially to have his sister's son as a son-in-law and will not look for payment. In this island also a man may marry the widow of his mother's brother.

A very unusual form of marriage practised in Hiw is that of a man with the daughter of his brother, and naturally of his elder brother. One such case was found in a pedigree and it was said that it was only an example of a custom which is generally recognised as orthodox if the brother's daughter is old enough.

When a man marries the daughter of his father's sister, the father's sister becomes his kwiiga and the man will not speak to her nor approach her, though there had been no such restrictions before the marriage. So, after marriage with the daughter of a brother, the brother becomes kwiiga and similar restrictions apply to him.

In the island of Loh a man is not now allowed to marry his father's sister though there seems to be no doubt that this form of marriage was once practised, and the marriage with the daughter of the brother appears to be unknown now, whatever may have been the case in the past.

I n Loh the cross-cousin marriage exists but is probably far less habitual than in Hiw, and it was subject to certain restrictions. It was said that a man would only marry the daughter of his mother's brother or father's sister if she had two elder sisters. In other words, if a man has two daughters only, they will not marry their cross-cousins but if he has more than two, the third may marry the son of her father's sister or of her mother's brother. The relationship of brother may be set up artificially between two men in the Torres Islands and in such a case one may not marry the sister of the other.

Kava-drinking^ .

In the Torres Islands kava is called gi and the act of making and drinking it wiena. Only men may drink it in the gamel of the Hukwa, which corresponds to the gamal of the Sukwe in the Banks Islands. The special feature of the Torres procedure is that each man prepares kava for himself.

' This section is based, partly on an account obtained by myself, partly on a description by Mr Durrad, Southern Cross Log^ 1912, p. 42.

1 86

Melanesian Society

Three kinds of cup are used in the preparation. One called the la tutu viena, lit. the cup standing beneath, is a shallow cup (see Fig. 7 a) formed by half a coconut-shell polished and cut thin round the lip; another called la tutu raina, the cup standing above, or la kemhem raina, the cup hanging above, consists of a narrow and long coconut-shell (see Fig. 7 d) cut off at the top, also polished and cut thin round the lip; the third bowl called rung is a shallow unpolished cup which is used to hold water. These cups are placed on stands consisting of rings of coconut-shell and they should never be allowed to touch the ground. The men of each rank of the Hukwa prepare and drink kava in their own compartment of the gamel where there is a strip of ground,

Fig. 7. Kava-cups; a, the la tutu viena', b, the la tutu raina.

called the tema komkom or place of mouth-rinsing, running lengthwise with the hut, which is used for this purpose. No one will step across this part of the gamel but will go instead to the lower end of the building and up the other side.

When kava is to be drunk each man takes his place, sitting cross-legged, those within each division sitting according to the order in which they were initiated into this rank. When the men of highest rank take their places, the fact is announced by the loud blowing of a conch-shell. Pieces of kava root are distributed and each man scrapes his piece of root with a shell called koli^ (Fig- 8), saying as he does so, '' Hiv raka netek da ne veta ruar'-, "Scrape up (or back) my soul from the holy place." He then bites off and chews a piece of root, taking at intervals a little water into his mouth. When he has chewed

• A species of Dolium.

Kava

sufficiently, each man washes his hands and taking the chewed root on the palm of his hand, he places it in the shallow cup which he fills with water, and squeezes the mass thoroughly till the juice has mixed with the water. Then, in order to get rid of any sediment he pours from the shallow cup into the tall cup and back again several times, holding the tall cup higher and higher each time till on the last occasion it is held at arm's length above his head. A thick froth forms on the liquid while the sediment remaining on each occasion is thrown away. Then, holding the kava, now in the shallow cup, in his hands the man calls with a loud voice either to one of the dead who in his lifetime was unusually rich and powerful or else to the ghosts of the dead in general, saying in the latter case, ''Gomaga nok, hemwera temet; vahgo no,' "Here is drink for

Fig. 8. Shell [koh) used to scrape kava.

you, ye ghosts of the dead, take care of me," and drinks off the kava at one draught. The cup is not drained completely but the dregs are thrown on the ground with the word ''tuf' while either on this occasion or when throwing away the sediment left during the pouring, the man may say, " Tut vita ne nanaria da no,'' "May evil things depart from me," thus throwing away from him any malignant influence due to his having visited some forbidden place or offended in any other way.

As already mentioned men only should drink kava, though women are now beginning to make it secretly. Mr Durrad records a practice in which a man of any rank of the Hukwa will approach the ganiel while kava is being made and will tap on the roof as a signal that he wishes to drink and a cup of kava

will be passed out to him in return for an arrow, mat, or other object used in native business transactions.

As Mr Durrad points out' the use of kava is associated with all the important events of a man's life. At the time of childbirth the husband will drink kava while he prays to some departed ancestor to aid his wife and in throwing away the dregs the cry of " Tuf will indicate an appeal that harm may be averted from her. When the child is one hundred days old a ceremony is performed as one incident of which kavacups are placed in a miniature gamel. When the boy is first introduced into the gamel, he sits beside his father and is given a little water to sip from the rung or water-bowl, and when he is fully grown, every important act of his life is preceded by the drinking of kava.

Finally, when soon after death the body is laid first at the threshold of the gamel and then in its lowest division, the men who are present drink kava, and it is used at a feast on the fifth day after death when the ovens are opened and again on the tenth day when the head is removed from the body. On the latter occasion the men act in concert, lifting their cups simultaneously and, as they throw away the dregs together, all say with one voice, "Throw away the thing which made this man to die." They drink together and spit out the last drops together, after which the head is taken back to the body where an offering is made of coconut, yam and kava, with the words, "Here is food for you; here is gi for you."

» Op. cit.

CHAPTER VIII

THE NEW HEBRIDES

In this group the systems of relationship of two islands were obtained ; of Pentecost Island or Raga in the northern part and of Anaiteum in the south. A few facts were also gleaned about Lepers' Island or Omba.

Pentecost Island.

The system of Pentecost is the most complicated and extraordinary of all those recorded in this volume. It was obtained during a brief stay at the island, but the little information on other topics which I was able to obtain is derived either from a native who was on the Southern Cross, or from John Pantutun of Mota who had lived for some time in the island. Any general information I can give about the island must therefore be received with caution, but I shall endeavour to make clear the probable value of each item.

The natives of Pentecost from whom my information was obtained came from the neighbourhood of Lamalanga at the north end of the island. At least two different languages are spoken, and the system of relationship and other information to be recorded belong to the people occupying the northern end of the island. It seemed that these people call themselves and their language Tahau. I was told that formerly all the people lived in two districts, one at each end of the island, the northern called Olohala and the southern Talai, with an unoccupied central region called Utevuroi, meaning the empty part or space, and that it is only in recent times that new villages have been built in this formerly unoccupied central region.

The social structure is that of the dual organisation with matrilineal descent. The two exogamous moieties are called Tagaro and Malau, the latter being sometimes also called

Subwe. Each moiety is called a tavalui, and a man calls a person of his own moiety, his ataluna or havana. There are also groups called verana, the exact nature of which I failed to discover.

According to John Pantutun there are also subdivisions of the moieties of such a kind that a man of one subdivision is not free to marry any woman of the other moiety, but must take one of a given subdivision. Thus, he said that each moiety had three subdivisions which we may call A, B and C in one moiety, and D, E and F in the other. It was said that a man of A had to marry a woman of D, a man of B a woman of E, and a man of C a woman of F, while men of D, E and F divisions could not marry women of A, B and C respectively, but a man of the F division must marry a woman of A, and so on. The details he gave could not be confirmed by an examination of the pedigree of the only native of the island I then had the chance of questioning, and I was obliged to leave the matter unsettled.

Though the details of the statement of John Pantutun could not be confirmed by an examination of the accompanying pedigree\ there are certain forbidden marriages which suggest a complication of the dual organisation. Thus, it was stated that John, the son of Hugo, could not marry Catharine, Mercy, Muriel or Vuivira, his father's sister's daughters, his father's mother's sister's daughter, and his father's father's brother's daughter's daughter respectively, nor could he marry the daughter of Muriel, though all these persons belong to the Malau moiety while John himself is Tagaro. Some of these persons would be debarred from marriage owing to ties of relationship, but it was stated definitely that they were forbidden as wives because they were of the same verana as John though of a different moiety, so that it seems that there is a social grouping which crosses the grouping into moieties and that this secondary grouping has the effect of preventing marriage with certain kin, even though these belong to the moiety into which a man has to marry. The exact nature of this secondary grouping must be left for future investigation. Taken in conjunction with the independent evidence of John Pantutun it would seem to point to some very important modification of the dual organisation.

1 In this pedigree the moieties of its members are given in italics below their names.

Social Organisation 191

Pedigree III.

VlRAMATA =

Malau I

MOLBWOITU = Mesalamua Tagaro \ Malau , 1 ,

Timothy = Meitalaveveana Molharana = Mesaldali

Tagaro \ Malau Tagaro Malau

I — Kongo

^ ,

TARiMONA = Aileen Anika

I Tagaro \ Malau \ ' 1 1

VlRAVAGA= Julia HuGO = Bessie Mercy Waga | Denga

Tagaro \ Malau Malau \ Tagaro MoLKALO = Muriel

I 1 , , 1 1 1 Tagaro \ Malau

Allan Catharine John Wilson Silas No ch.

MOLTANGAROA= $

Malau I

I 1 1 ' 1

RoiROl I I TARlASU = Metarihuhu

Timothy = Meitalaveveana ViRALiu = Vevlosingo Tagaro \ Malau Tagaro \ Malau Tagaro I Malau Tarimuli

(see above) r— ^ 1

ViRATURU=Girilalau Girinamoli

Tagaro \ Malau Vuivira

There is some sort of correspondence between the two moieties and the main groups of the Banks and Torres Islands. A man from either place is assigned to one or other of the Pentecost moieties, but I could not discover by what principle the arrangement is governed. The three divisions of the Torres Islands create a difficulty but nevertheless I was told that it was known definitely to which moiety a Torres man who came to Pentecost would belong.

I could not discover that there were any special prohibitions connected with the two moieties or any special animals their members might not kill or eat, but I am indebted to the Rev. H. N. Drummond for the information that the members of the Subwe moiety call the Tagaro people matan dura (sow) and matan talai (giant clam) while the Tagaro people call the Subwe matan avua (turtle), matan tabwatabwa (a flower), and matan bweta (taro). These names are now generally used only in sport or anger but were said to have reference to the origin of those to whom they are applied. A stranger whose tavalui is unknown is called mutan gai halhala or floating stick.

The greater part of the system of relationship to be now recorded was obtained by the genealogical method during a brief visit to the island supplemented by information given by means of the genealogical method by the native on the Southern Cross already mentioned. Several terms which I had omitted or about which I was uncertain were obtained for me by Miss E. Wilson of the Melanesian Mission. Further, the Rev. H. N. Drummond of the same mission went into the whole subject very carefully after my visit, making inquiries at many villages. He kindly gave me the list of terms which he had compiled and, with one or two exceptions which will be recorded, his list agrees so closely with mine, that there can be no doubt that the system about to be recorded, extraordinary as it is, is accurate in all essentials. The following are the terms with their applications. They were usually obtained with the third person possessive, and I therefore give them in this form, adding occasionally that for the first person also. In the cases of the more anomalous relationships, I have added instances from the pedigree which is only one of several I recorded.

Tamana {tamanggu), vocative tata. The term for father, also applied to the father's brother and the husband of the mother's sister, elder brothers of the father being tamana lalaboa or great fathers and younger brothers tamana terigi or little fathers. In addition to these uses, the term is also applied to the son of the father's sister as in the Banks Islands ; thus John would call Allan tamana, i.e. a crosscousin is called by the same term as the father. Other relatives who are called tama by a woman are the daughter's husband and the husband of the sister's daughter, thus Meitalaveveana would apply the term to both Viravaga and Molkalo as well as to Molbwoitu, relatives two generations apart being addressed by the same term, so that men whom we should regard as the sons-in-law of a woman are classed by her with her father. A man may be addressed as the father of his child, thus Hugo may be spoken of as taman John.

Ratakina [ratahiku), vocative mua or niuani. The term for mother, mother's sister and wife of the father's brother. The father's sister is also usually addressed by this term though according to Mr Drummond she may also be called bilan barai. This term does not resemble in form the other terms of relationship, and may possibly be a term for the

status of this relative, or may belong to a more ancient system. In addition, ratahi is used in a sense corresponding to one of the uses of tama, viz. for the daughter of the father's sister, and the wife of the father's sister's son is also known by this term ; thus, John would call Catherine and Mercy ratahina, and also the wife of Allan when this boy marries.

Nituna (niiuk). The term for son or daughter, the reciprocal of tama and ratahi. It is used as the reciprocal to these terms in their anomalous senses, so that the son or daughter of the mother's brother is nitu, thus corresponding to the Banks Islands usage. Similarly, the wife's mother and the wife's mother's sister are addressed and spoken of by this term ; thus, Viravaga would apply the same term of relationship both to Meitalaveveana and Mesaldali and to Catharine. The child of the sister's son is also called nitu.

Mr Drummond tells me that terms are used to distinguish the sex of one called nitu, viz. mala for males and 7itei for females. I do not know if these terms are only used for own sons and daughters or whether they follow the anomalous uses of nitu ; probably the former is the case.

Tuagana. This is the term for elder brother when a man is speaking and for elder sister when a woman is speaking. It is also the term for the child of the father's brother or mother's sister when the person addressed is older than the speaker and of the same sex. In addition, the mother's mother and her sisters are called tuaga by both sexes ; thus, Allan and Catharine give this name to Meitalaveveana and Mesaldali. A peculiar feature of this nomenclature is that a woman is addressed by a male two generations below her own by means of a term ordinarily used between those of the same generation and the same sex.

Tihina. The younger brother of a man or sister of a woman, and thus the reciprocal of tuaga. It is used by a woman of her daughter's child, being applied by Meitalaveveana to Allan and Catharine. A woman also calls her sister's daughter's child tihina, Mesaldali calling Allan and Catharine by this name.

Tuana. This is a term including both tuagana and tihina.

Hogosina (hogosikuy. A reciprocal term for the brothersister relationship, being given by a man to his sister and by ^ I was also given this term in the form gorosina.

a woman to her brother. It is used in the customary way for the child of the father's brother and of the mother's sister. It is also used in tw'^o anomalous senses ; in one a man and his sister's daughter's daughter are hogosina to one another ; thus, Catharine and the brother of Meitalaveveana would stand in this relationship to one another, the usage thus falling into line with that of Catharine and Meitalaveveana herself. The other anomaly is that a woman and her husband's father are hogosina to one another, thus Bessie would address Timothy as hogosina and would be so addressed in return. Here persons only one generation removed use a term normally used between those of the same generation.

Tarabend^. The mother's brother. I was told by one informant that this term is also used for the father's father, but this information was not confirmed and is probably wrong.

Aloana. The reciprocal of the last term, used for the sister's son by a man.

Hurina. The husband of the father's sister. I could not discover definitely whether it is used reciprocally, i.e. for the wife's brother's son ; probably not.

Sibina. The sister's husband of both man and woman and also the husband's brother, being applied to Viravaga by both Hugo and Aileen and to Waga by Bessie. The term may also be used by a woman of her husband's sister, as by Bessie of Julia or Aileen, though there is also another term, habwena, for this relationship. Thus, it is a term for various relatives by marriage of the same generation as the consort. Further, it is applied by a woman to her husband's mother as by Bessie to Meitalaveveana, being thus used by a woman of another woman a generation above her own, and according to one account it is also applied to the husband's father, as by Bessie to Timothy, though, as we have seen, this relative is also kogosi. Lastly, sibi is used both by males and females for the father's mother and the mother's father ; by both Hugo and Julia for Molbwoitu and for the wife of Moltangaroa. According to some informants, including those from whom Mr Drummond inquired, it is also applied to the father's father, while others said that this relative was properly

' This is the word as written by Mr Drummond. I usually heard it tarapetta, but Codrington (3/.Z., 432) states that these differences are chiefly individual and I have adopted Mr Drummond's spelling as probably representing the more usual form.

addressed as atalaveraku, i.e. man of my verana. It will be remembered that one informant classed this relative with the mother's brother so that for this relationship there is definite discrepancy of evidence.

Mabina^. This is a term reciprocal to sibina in all its senses, so that it is applied by a man to his brother's wife and his wife's sister, and also to the wife's brother though there is another term for this relative. By a woman it is applied to the brother's wife, though here again there is another special term for this relationship. Further, the term is used by a woman for the son's wife and according to one account by a man also. A woman applies this term to her son's child and a man to his daughter's child, while according to some accounts a man also calls his son's child mabi. Lastly, the wife of the mother's brother is called mabi. For the relationship reciprocal to the last it was said that sibi is not used, but that the husband's sister's son is addressed by his personal name.

Bulenana, bulenanggu. The wife's brother ; also sometimes, as we have seen, called mabi. This term is not used reciprocally but a man calls his sister's husband sibi.

Habwena. A reciprocal term for husband's sister and brother's wife (w. s.), i.e. used between women.

Bwaligana. A reciprocal term for wife's father and daughter's husband.

Ahoana. Husband ; a woman will speak of her husband as "the man."

Tasalana. Wife. A man often speaks of his wife as a ghost (atmate) or as rubbish or excrement, and this depreciatory usage may also be used by others so that people who have seen the wife of a man will say to him " I have met your atmate'' or " I have met your rubbish."

Lalagi. A term used only in address for the brother's wife, the mother's brother's wife and the wife's sister, all relatives who, as we shall see later, are potential wives of the man who uses the term. According to one account this term is also used by a man for the wife of his sister's son.

Havana. A term for all those of the same tavalui or moiety.

1 I usually heard this word as mambina^ both the pure b and mb being interchanged according to Dr Codrington,

13—2

196 Melmtesian Society

Waruena. A term applied to one another by the different wives of a polygynous marriage.

Mr Drummond informs me that when a man has three or more sons, the eldest is tuagai, the youngest vabwehui, and the intermediate son or sons lologi.

Before considering the more unusual anomalies of this system, reference may be made to certain difficulties in connection with the terms for relatives by marriage of the same generation. In the usual Melanesian system of the less simple kind the terms used between those of the same sex differ from those used between those of opposite sex and the Pentecost system is anomalous in that the terms sibi and mabi, which would seem to be properly terms for use between persons of different sex are also used between persons of the same sex. For some of these relationships there are, however, other terms, bulena used by a man of his wife's brother and habwe used between women, and it would seem probable that these are the correct terms for these relationships, and that their denotation by the terms sibi and mabi is an innovation, the adoption of which has been assisted by the wide connotation of these terms. From the consideration of these terms as used for relatives by marriage of the same generation it seems most probable that primarily sibi is a term used by a woman for her husband's brother and sister's husband, while mabi is a term used by a man for his wife's sister and his brother's wife.

Some of the anomalous features of the Pentecost system are the same as those of the Banks Islands and are probably to be explained in the same way. Thus, the fact that crosscousins are regarded as standing to one another in the relation of parent and child, the child of the father's sister being the parent, and the child of the mother's brother the child in this relationship, corresponds exactly with the Banks condition, and, as in those islands, it would seem to have had its origin in a marriage regulation still in practice which makes the wife or widow of his maternal uncle the proper wife of a man. Another anomalous relationship which is to be explained on the same lines is that the mother's brother's wife is called mabi, which is otherwise used for the brother's wife (m. s.) and the wife's sister, i.e. for people who are potential wives (see p. 206). So far as these three relatives are concerned the term 7nabi is the exact equivalent of the Mota maieima.

I

Marriage with Wife of Mother's Brother 197

Further, these three relatives are distinguished from other persons who are called mabi in that they are denoted in address by another term, lalagi, and according to one account this term is also used for the wife of the sister's son, who would thus appear to be a potential wife. Another feature which will follow from marriage with the wife of the mother's brother is that the wife of the father's sister's son will be a rataki; for by this marriage the wife of this cousin will be classed with the mother and might even be the actual mother of the person who is speaking. In all these examples the special feature of the anomaly is that people apply to those of the generation immediately above or below them a term which is also applied to persons of the same generation.

There are two other relationships in which the anomaly is of the same order, viz. those between a woman and her husband's father and mother. A woman calls her husband's father kogosi, used otherwise for her brother, and possibly also sibz, which is otherwise used for a relative by marriage of the same generation, and the husband's mother is certainly called by the latter term. The result of marriage with the wife of the mother's brother is illustrated in the accompanying scheme.

I ' 1 ^

B=a A=b

I A» = b

Where there are only two divisions of the community the persons called B and b in this scheme, if members of the same generation and of the same moiety, would be kogosi to one another, and if b, the wife or widow of A, becomes the wife of A\ her husband's sister's husband B then becomes by the new marriage the husband's father of b, i.e. her husband's father is at the same time her kogosi. The use of the term sibi for this relative and also for the husband's mother is less straightforward. In the one case a person of the same moiety is called sibi, while in the other case a term which should probably be applied by a woman to a man is here used between women. The use of the term sibi for one of the same moiety is so unlikely that we may be fairly confident that this item of information is wrong. It is possible that the use of the term for the husband's mother is the result of a woman calling her brother's wife mabi (see p. 195), for a would normally call b mabi as the wife of her brother A, and when b becomes the

wife of A\ i.e. the son's wife of a, she will continue to be called mabi, and reciprocally she will call her husband's mother sibi.

It is more difficult to see why the child of the sister's son should be called nitu. The use of the term would follow naturally if the wife of the sister's son were classed with the wife. It is thus explicable as a result of marriage with the wife of the sister's son, but it is difficult to see how it has come about through marriage with the wife of the mother's brother.

There still remains a group of anomalies of the Pentecost system wholly different from those present in the Banks Islands. These anomalies differ from those already considered in that persons two generations apart apply to one another terms also used for those of the same generation. The following is a list of these terms : — A person and his or her mother's mother address one another by terms used otherwise between brothers and sisters ; both men and women call their mother's mother tuaga, which is the term given to the elder brother of a man or the elder sister of a woman, and reciprocally she calls them tiki, used otherwise for younger brother or sister. Closely allied to this usage is the fact that a woman calls the brother of her mother's mother kogosi, a term which she would otherwise apply to her brother. Next, the other three kinds of grandparent, the father's father, the father's mother and the mother's father are all called sibi, a. word which, as we have already seen, is otherwise employed for brothers- or sisters-in-law, the use of the term for the father's father being, however, doubtfully correct. Further, the mother of the wife is called nitu or child, while reciprocally she calls her daughter's husband tama or father, a person of a generation above thus using a term appropriate to the generation below, and m'ce versa. Lastly, and in line with the preceding, the husband of the sister's daughter of a woman is called by the term used otherwise for father, and correspondingly the child of the sister's daughter, a person two generations below the speaker, is called tuaga, tiki or kogosi, according to age or sex, though it would probably not often happen that the term tuaga would be appropriate.

When I discovered the existence of these correspondences in the system of Pentecost, they seemed to me absolutely inexplicable, and the only fact which made me think that they were probably correct was that they were obviously

Marriage with Granddaughter 199

of the same order as certain anomalies I had found a few months earlier in the systems of some inland Fijian tribes (see Chap. XI). While in Pentecost I heard of no social institution of the island which seemed to throw any light on their origin, but a clue was given later by a chance remark thrown out by a native of the Banks Islands. John Pantutun, from whom I learnt most of what I have had to say about relationship in the Banks Islands, had lived for some time in Pentecost and frequently compared the customs of this and his own island in a very instructive manner. One day when speaking of Pentecost he said more or less in scorn that it was a place where they married their granddaughters. Here it seemed to me at once was the source of the anomalies of the Pentecost system, for the natural result of marriage with a granddaughter would be to put people on such a double relationship to one another as is shown to have existed by the Pentecost terms ; people two generations apart would come to be related through marriage as if they were of the same generation, exactly as I had found to be the case.

Before considering the matter further it may be pointed out that with the dual organisation a man could not marry his son's daughter or one who would have the same name as the son's daughter by the classificatory principle, for these women would be of the same moiety as their grandfather (see p. 18). If therefore it had been the custom in Pentecost to marry a granddaughter, using this term in the classificatory sense, it must have been the daughter's daughter who was married.

By the time I received the clue provided by John's remark I had forgotten the details of the Pentecost system, and it occurred to me that before refreshing my memory it would be interesting to construct a theoretical system founded on marriage with a daughter's daughter and then see how far the constructed system agreed with that I had actually obtained. The following would seem to be the natural results of such a union so far as relatives by marriage and grandparents are concerned.

It is probable that in the marriage in question a man would not marry his own granddaughter but the granddaughter of his brother, using this word in the classificatory sense. For the purpose of simplicity, however, I will assume at first that a man marries his own daughter's daughter and consider the relationships set up by such a marriage. Thus, if in the

schematic pedigree A^ marries b^ the daughter of his daughter, one result will be that his wife's father (A*) will be the husband of his daughter (b") and would therefore naturally be denoted by the same term, and as a matter of fact both are in Pentecost called bivaliga ; secondly, his wife's mother (b'') will be his own daughter, and as a matter of fact the wife's mother is called 7iitu, otherwise the word for a daughter ; thirdly, his wife's sister (b^) will be his daughter's daughter, so both should be denoted by the same term, and as a matter of fact both are called mabi\ lastly, his wife's brother (B^) will be his daughter's son, and, as a matter of fact, both are called mabi, though there is a second term, bulena, for the wife's brother which is not used for the daughter's son. Thus, the four chief relationships of a man set up by his marriage with his daughter's daughter are denoted by terms agreeing exactly with those

B2=a2 Ai = bi A2 B3 = ai

I I I ' 1

A* B=a3

A3 = b I

Ai=b3 b* Bi

which were foreseen to be the natural result of such a marriage. In the marriage of a man with his own granddaughter the two relationships in each case are present in one and the same person, but it is hardly necessary to point out that by the classificatory principle, marriage with the granddaughter of the brother would have exactly the same effect on nomenclature.

The terms used by the wife (b^) in the marriage in question are not so straightforward. The result of the marriage of b' with her mother's father (A') will be that her husband's father and mother will be her great-grandparents, while as a matter of fact they are actually called hogosi and sibi respectively, but, as we have seen, the terms for these two relationships would be the natural result of another form of marriage which is practised on the island, viz. that with the wife of the mother's brother.

The relationships with the brother and sister of the husband fit in more closely with the granddaughter-marriage. The husband's brother (A') will be one and the same person as the mother's father's brother, and the husband's sister (a') as the mother's father's sister. The brother and sister of the

I

I

Marriage with Granddaughter 201

husband are both called sibi, though the husband's sister may also be called habwe, and it is almost certain that the relatives with whom they would be identified by the marriage in question are also called sibi though the terms for these relationships were not actually obtained. If, on the other hand, it be assumed that the marriage takes place with the brother of the mother's father, the agreement is more definite, for then the husband's brother will be the mother's father, and he is certainly called sibi.

The nomenclature for the four grandparents agrees closely with that to be expected as the result of the marriage in question. The relationship of mother's father has already been considered, this relative being also the husband's brother when a man marries his brother's granddaughter. In the case of the relationship of mother's mother the important point is what she would be called by the brothers and sisters of the wife, for to the wife herself the mother's mother would be simply another of her husband's wives. To the brother (B^) and sister (b*) of the wife the result of such a marriage is that b^ who would probably be their elder sister, will become the wife of their mother's father, and would come to rank as a mother's mother, thus providing the correspondence in nomenclature actually found between the mother's mother and the elder sister. Two of the grandparents are thus denoted exactly as it was foreseen they should be if their nomenclature had been the result of the marriage in question.

The terms to be used for the father (B') and mother (a^) of the father were not so obvious. It seemed at first sight as if with matrilineal descent they need not be related at all, but it must be remembered that, as a consequence of the dual organisation, these relatives would have a status which would make it suitable for certain terms to be applied to them by the husband. On reference to the schematic pedigree it will be seen that B', the father's father of b' and of B\ will be of their moiety and of the same generation as b\ the mother's mother. B' and b' would therefore, by the classificatory system, be hogosi to one another, and in consequence B^ would have the position of mother's mother's brother of b^ and B' and would probably be called by the same name as the mother's father, viz. sibi. On the other hand, it seemed possible that he would merely receive the name applied to a man of the same moiety. It is interesting to find that this

relationship, the right nomenclature for which is theoretically doubtful, is one about which there was much uncertainty among our informants ; according to Mr Drummond the father's father is sibi, which would agree with one theoretical deduction, while I was told that he is called atalaveraku, which is merely a name for a man of the same verana or social division as the speaker. Similarly, the father's mother (a-) would be a woman of the same moiety and the same generation as the mother's father (A^), i.e. the husband, and would therefore be called by the same term as the husband's sister. She is actually called sibi, and though the proper word for the husband's sister is kabzve, there seemed to be no doubt that this relative might also be called sibi.

There is thus the most remarkable correspondence between the terms actually obtained and those deduced theoretically, the only definite exception being the case of the father and mother of the husband, who might have been expected to be identified with the great-grandparents, but it is just these two relationships which have been accounted for in a different way, viz. as the result of the marriage with the wife of the mother's brother. In all other cases there is an almost exact agreement between the terms which should be expected as the result of marriage with the daughter's daughter, either of the man himself or of his brother.

In addition to the correspondences already described there is another rather more complex set which had not been foreseen. The husband of the sister's daughter is classed with the father by a woman and his children are classed with brothers and sisters by a man (see iama and hogosi). I was only told of these correspondences when a woman is speaking in the one case and when a man is speaking in the other, but both follow from marriage with the granddaughter whether men or women are concerned. Thus, if B marries a^ the brothers and sisters of her mother a' will come to stand to B in two relationships ; he will be at once their father and the husband of their sister's daughter. Similarly, if the children of B by the new marriage are classed with A^ and other children by the earlier marriage, the brothers and sisters of a^ will, according to their age and sex, be the tuaga, tihi or hogosi of her daughter's children, i.e. the children of the sister's daughter will be classed with brothers and sisters.

There still remains one correspondence to be considered.

Marriage with Granddaughter 203

In Pentecost a woman and the brother of her mother's mother call one another hogosi, the term for the brother-sister relationship. I have already suggested that theoretically we might expect that the brother of the mother's mother would be classed with the father's father, being of the same moiety and generation, but this does not help us, since the proper term for the father's father is uncertain. There is no obvious reason why the marriage of b* with A^ should make b* regard the brother of b^ as her brother, though, if b^ came to acquire the position of mother's mother, it is possible that this change of status might be transferred to B^ who would thus come to be regarded as a mother's mother's brother. Such a process does not seem very probable, but I am unable to suggest any more satisfactory explanation. It may be noted, however, that the correspondence in question rests on one statement only and may be an error.

In this demonstration I have assumed that a man married his own granddaughter but the correspondences would by the classificatory system be the same if a man married the granddaughter of his brother and, as we have seen, there is one correspondence, that of husband's brother and mother's father, not altogether straightforward on the hypothesis of marriage with the granddaughter, which is readily explicable if a man married the granddaughter of his brother.

It would therefore seem that the marriage which has had as its consequence the remarkable series of correspondences I have recorded was with the granddaughter of the brother rather than with the granddaughter herself and on general grounds the former kind of marriage is of course far more probable.

Further, the most definite evidence I obtained of the actual occurrence of marriage with one who would rank as a granddaughter points in the same direction. I have already mentioned that I only obtained the clue to the existence of the granddaughter-marriage after I had left the island and at this time my only means of inquiry was the native on the Southern Cross already mentioned. He was very explicit that a man might and actually does marry the granddaughter of his brother but said that a man would not marry his own granddaughter and might even have been killed for doing so. John Pantutun was certain of the existence of marriage with a granddaughter in the classificatory sense and was able to

illustrate from his own pedigree the kind of marriage that takes place. Mr Drummond, on the other hand, could not find that these marriages take place at the present time, and we must await further information, and especially a study of the subject by a more complete application of the genealogical method than was possible during my brief visit. In the meantime, however, we can be confident that even if such marriages are no longer practised, they have taken place in the past and in such an organised manner as to have produced the extraordinary correspondences which have been shown to be the natural results of this form of marriage.

Functions of Relatives.

Only a very fragmentary account of these could be obtained. A complete account would perhaps show that they are as numerous and elaborate as those of the Banks Islands.

Mothers brother. The functions of this relative are probably of much the same nature as in the Banks Islands but few details were obtained. It was clear that the normal inheritance is by the sister's son who takes any individual property (see p. 209). If a man dies while his children are young they are taken in charge by their mother's brother.

If a dying man gives any kind of food to his sister's son, this kind of food may not be eaten for a time, either by the sister's son himself or by the brother of the dead man, and this restriction may also be observed by other persons of the moiety if they like. The duration of the prohibition is connected with some observance of the organisation called Loli which corresponds to the Banks Sukive\ it probably comes to an end when the sister's son takes his next step in this body, but the exact conditions could not be obtained.

Father s sister. The functions of this relative are of the same general kind as in the Banks Islands. She chooses a wife for her nephew who will not hesitate to take the chosen woman and the father's sister will also be obeyed in general. If a woman asks her brother's child for anything it must be given and an order given by her must be carried out. A man and his father's sister may eat together but he may not address this relative by name. If a man wishes to attract the attention of his father's sister, he will call her muani once or twice and if she does not hear and is not far away, he may

Functions of Relatives 205

then sometimes call to her by her personal name and if they are alone she may not be angry but, if others are present, this conduct would be very much resented.

Father s sister s husband. This relative is obeyed by his wife's nephew but the two men can eat with one another and talk familiarly. Viravaga in the pedigree is also called Thomas, and John will address him as Tom. I could not discover that there are any special practices resembling those connected with this relative in Mota.

Father s sister s son. It seemed that it was the duty of this relative to dig the grave on the death of the son of his mother's brother. It was said that this duty might also be undertaken by the father's brother. In each case it will be noted that the person who digs the grave belongs to one moiety and the deceased to the other.

Wifes brother and sister s husband (m. s.). A man may not touch the head of his sister's husband nor go behind his back but the restrictions are not reciprocal, for a man may touch the head and go behind the back of his wife's brother. These men may, however, use each other's personal names. The unusual feature of these customs is that they are not reciprocal and this is probably connected with a special feature of the terms of relationship. It was clear that a man calls his wife's brother bulena but this term was not used reciprocally, a man calling his sister's husband sibi. The bulena of a man is therefore one who may not touch the man's head nor go behind his back. We have here a very good example of the close connection between terms of relationship and functions. A relationship for which it is usual to have a reciprocal term requires in this island two separate terms, because the two relatives have different duties and restrictions.

Wifes sister, A man may not speak to his wife's sister; he may not use her personal name nor speak to her. If a woman is up a breadfruit or coconut tree and sees her sister's husband coming, she must at once descend. Neither man nor woman must touch the head of the other. I do not know if this restriction is applied in the metaphorical manner of the Banks Islands (see p. 44), but since two people who may not see one another are not likely to touch each other's heads, the fact that I was especially told of this restriction suggests that it has a wider significance than appears at first sight.

Brothers wife (m. s.). It was said that a woman could

2o6 Melanesiait Society

speak to her husband's brother except when her husband is dead, when the brother must not speak to the widow unless he is going to take her as his wife.

Marriage.

I have already had to consider certain kinds of marriage which are prohibited or allowed in Pentecost. There is evidence that the ordinary restrictions depending on the dual division are supplemented by a secondary mechanism which prohibits marriage with a number of women of the opposite moiety, though the exact nature of this mechanism could not be determined. Further, there is evidence of the existence of positive regulations enjoining marriage with certain relatives. There seemed to be no doubt that marriage with the wife or widow of the mother's brother is still practised at the present time.

The marriage with the daughter's daughter (using these words in the classificatory sense) which has been inferred from the system of relationship, is probably also still practised and it is probable that a man still marries the sister of his wife in polygynous marriages; he certainly does so after the death of the wife. It is also the custom for a man to marry his brother's wife on the death of the brother. The practice is not obligatory but is usually followed. The brother has the first claim on the widow. It was stated by John Pantutun that these regulations concerning marriage with the brother's wife and wife's sister were the same in Pentecost as in the Banks but were followed more strictly in the former island.

There remains to be described the ceremonial connected with marriage. The following account was obtained from John Pantutun, a native of Mota in the Banks Islands. He had been for some time on Pentecost and had witnessed the proceedings he described. He is certainly an acute and trustworthy observer but it must be remembered that his account is that of a stranger and that there was no possibility of corroboration or of inquiry into special points from natives of the islands

• It will be noticed that the account shows a general agreement with that ^iven by Dr Codrington {M., 240). It must be remembered that my account is that of an actual marriage as witnessed by a stranger to the island while Dr Codrington's account is in general terms.

Marriage Ceremony 207

A marriage is arranged between the would-be husband and the parents of the girl, the latter not being consulted even when she is old enough for this to be done. Usually the marriage is arranged when the girl is quite young, the man who wishes to marry settling with the parents of the girl how many pigs he will give and it is arranged in how many days he shall be ready with them. On each day he plucks a leaflet from a leaf of the male cycas to keep a tally of the days and when he tells the parents that only two leaflets remain to be plucked and that it is time to prepare the property of the girl, they put in a big bag all the things which they and other relatives propose to give. While they are doing this the girl will be sent out to the gardens on some pretext in order that she shall not suspect what is about to happen. On this day also the future husband will bring a new mat which the girl's mother hides in the house so that her daughter cannot see it. On the next morning the father keeps the girl in the house on some pretext, such as roasting breadfruit or playing some game, and the future husband comes with his people to fetch her. Some man is chosen to go into the house, pretending that he has come to fetch fire, and he goes quietly to the girl and, seizing her by the wrists, says "Marry." The girl is very much astonished and begins to cry and before she knows what is happening she is wrapped up in the new mat brought the day before by the future husband. This is done by his sister who has come into the house. This wrapping up is done as quickly as possible after the girl has learnt that she is to marry, for if she has time to realise what is happening she will resist and it was said that a girl has been known to seize a knife and slash the arm of the woman who is attempting to bind her. The man who has told the girl to marry then seizes a stick which he knows will be wanted and goes out of the house to encounter a man belonging to the moiety of the father of the girl. This man has a club with thick thorns on it with which he strikes at the man who comes out, it may be so severe a blow as to break his arm. It was said that he does this to show that only the brothers and the sister's son of the father are to get the pigs which the bridegroom is about to present. As soon as the blow is struck all the men of the village produce sticks prepared the night before and begin to fight with the party which has come to fetch the girl. While this is going on the sister of the bridegroom takes

2o8 Melanesia^ Society

the girl out of the house, the latter being so completely wrapped up in the mat that she can see nothing. The girl is then unwrapped and her father gives her one of his own pigs to kill as a sign that it is the last of his property with which she will have anything to do. She kills the pig with a club and is then again wrapped up in her mat. One of the husband's party is then deputed to fetch the pig, this duty being regarded as a high honour. A relative of the father stands over the animal to resist its being taken but the man of the husband's party has only to succeed in touching its body for the resistance to cease when the pig is cut up and the parts distributed. The bridegroom takes the head which he gives to some bachelor of his own party. The man who is given the head in this manner may not marry a widow but must marry a girl not previously married and there is now great unwillingness to receive the offering on this account because of the greater expense connected with the latter kind of marriage. In consequence the young unmarried men will hide themselves and the father of the man who receives the head will be very angry and will threaten to have his revenge on the morrow.

As soon as the pig's head has been given, the bag containing the property for the girl is put down in the midst of the people with a yam beside it and the bridegroom walks round the bag four times, puts his hand on the bag and takes up the yam to give to his sister who takes it away. Some of the bridegroom's party then come to take the bag and the people present, all of whom are armed and bear shields, begin to fight. The girl is then taken to her future home, the people of her own village standing in the way and resisting her removal, and anyone who offers much resistance is appeased with a present. When the husband's village is reached the girl, still wrapped up, is put down in the open space of the village and the husband presents pigs to the girl's father, at least four being given, and if the husband is an important man more than this number, and mats are also given at the same time. The sister of the husband who had wrapped up the girl now unwraps her and the bride then opens the bag containing her property and after giving some to her father's sister, hands the rest to her husband. Then the father of the girl distributes the four pigs he had received to his brothers and to his sister's son, the latter being regarded as the chief recipient, and the girl then takes up her abode in the house of her husband's mother.

Property 209

When she first goes to live in the house of her husband's mother, and if she is young it may be for several years afterwards, the girl will not speak to her husband who does not live in the house but continues to spend all his time in the club-house. The man may sometimes come into the house to give food to his mother but even this would be exceptional. When the girl is old enough the husband will tempt his wife to speak to him and may offer her food but so long as she does not speak there are no sexual relations between them. This state of things may last for years and it was said that a wife would sometimes die without having spoken to her husband in which case marital relations between the pair would never have taken place. When she speaks, even if only a single word, the man will tell his mother that his wife has spoken and the pair will live together as man and wife.

Property,

It was said that formerly all property belonged to the verana, viz. the subdivision of the moiety the exact nature of which is uncertain (see p. 190), and it was said that even if property belonged to individual persons it was not inherited by individuals but by the verana as a whole. Canoes are in no case individual property even now, but always belong to the verana as a whole.

At the present time it is clear that property in general is owned by individuals and should properly be transmitted to the sister's son, though a change is going on whereby the property of a man is being inherited more and more by his own children. Thus, formerly the trees of Hugo (see pedigree on p. 191) would have gone entirely to Allan, Catharine, Mercy, Anika, Muriel and Vuivira, Allan getting by far the largest share, but now the children of Hugo would have the larger portion and the sister's children very few. The fact that Allan gets the largest share is due to his sex, not to the nearness of his relationship, for it was said that if Muriel had been a boy, she would have received the same share as Allan. It seemed quite certain that the sister's son should properly inherit the house. At the present time if a man expresses the wish that his children shall remain in the house, the sister's son always allows them to stay, but if he lives in the house with them he is the more important

2IO Melanesian Society

person in it and if he quarrelled with the children, it would be they who would have to leave and build elsewhere while the sister's son would remain in possession of the house.

John Pantutun of Mota who had spent a long time in Pentecost was emphatic that inheritance by the sister's son was much more habitual than in his own island.

Identity with Animals.

The people of Pentecost were said by John Pantutun to associate themselves with animals which were supposed by him to correspond to the tamaniu of his own island (see p. 154). He told the following story of a man and his son whom he knew well while he was living on the island. The father, whose animal was a shark, was angry with his son who had refused to roast a yam for him. When soon after this the son went out in search of flying fish, he saw a large shark in the water which swam round and round the canoe making the sea rough and then put its nose over the outrigger of the canoe and upset it. The son cried out with fear and expected to be eaten but was allowed to swim away while the father, in the form of the shark, took the canoe ashore and hung the paddle, fishing line and the fish which his son had caught over the door of his house. The son was picked up by some friends and on reaching home was telling the story to his father when he looked up and saw the things over the door and was told by his father that he had been the shark and had given him a lesson not to disobey in future.

The Loll.

I obtained very little information about the Loli which corresponds to the Banks Sukwe. The name of seven ranks were given in order of importance, viz. Vira, Dali, Livusi, Virei, Osisi, Bwaranga, Moli and Tari, but apparently only Vira, Moli and Tari are now left. These words are prefixed to the personal names of those who possess the rank and the names Viramata, Molbwoitu, Tarimona, etc. in the pedigree on p. 191 are examples of such prefixes.

The names of women have similar prefixes which denote differences of rank but it was said that the rank is acquired altogether by the killing of pigs and is not connected with any

Kava 2 1 1

organisation resembling the Loliox Sukwe. The highest rank of the women is Motari corresponding to the Vira of the men; Maitalai corresponding to Dali, Metari to Osisi, Vevlo to Bwaranga, MesaH to Moli and Muei to Tari. Only the three highest ranks are permitted to wear certain ornaments. It will be noticed that several of these prefixes occur in the names of women in the pedigree and they were present in other pedigrees which were collected, so that it would seem as if women were preserving these distinctions more completely than the men.

In Pentecost as in the Banks all differences of rank disappear when at sea.

Kava-drinking.

The following account of the method of making and drinking kava on Pentecost, where it is called maloku, was said to hold good also of Aurora (Maewo) and Lepers' (Omba) Islands. In these islands the root is grated on a long piece of coral into a large wooden bowl. The root is first washed carefully and several pieces are held in the moistened hand and rubbed against the piece of coral. Here, as in the Banks, kava should only be drunk in the men's house, though the people are now beginning to drink it elsewhere.

Two deep cups are used; one into which the juice of the scraped root is squeezed and the other used as the drinkingvessel. The scraped root is put in dampened coconut fibre which acts as a strainer and the mass is squeezed so that the juice falls into one of the two vessels. The scrapings are then put back into the wooden bowl in which the coral has been left and are washed by pouring water over them, and then they are again put into the strainer and the juice squeezed into the cup which will perhaps now be filled with liquid. The scraped root is then thrown on a heap, often several feet high, which has accumulated from previous drinkings. The strainer is then washed with some more water and folded so as to form a kind of bag, sometimes in the shape of a flying fox, sometimes in the shape of the head of a bird. The cup which is full of kava is then passed round the strainer and the contents poured into it so that they fall into the other cup which had so far been used to hold water. As the kava is poured into the strainer this is lifted higher and higher so that the liquid

14 — 2

falls Into the drinking-cup from a height and forms a froth which stands well above the top of the cup when it is filled. Finally, the strainer is squeezed so that all the liquid falls into the cup.

The man who is to drink takes up the cup in both hands with much deliberation, blows off some of the froth and drinks from the edge of the cup, shutting his eyes as he does so, it was said, in order that the froth should not get into them. While he is drinking no one may pass behind him nor may anyone walk about.

Both maker and drinker sit squatting but not cross-legged as in the Banks, and after drinking both men continue to sit in the same position for a minute or two, when the drinker may clean his mouth by chewing a little sugar-cane. Before drinking a man will take off anything tight that he may be wearing and will stand up and stretch himself till his joints crack.

A man must not eat before drinking kava. If an oven is opened when he is about to drink, he will keep his share till afterwards and if a man is seen to put his food into a basket it will be known that he is going to drink kava.

Sometimes a man may make kava for himself but it is more usual for two men to act together, each making it for the other, and if there are many people present, they will form pairs, each of a pair making for the other, though sometimes the contents of the cup will be divided among two or more. When a company fall into pairs in this way it is called a fight, the idea being that each man shall make his companion incapable and unable to eat. It was stated that the kava made in this island is much stronger than in the Banks or Torres Islands.

A rich man after drinking will take sugar-cane in his mouth and after spitting this out, will take a second piece, shoot it out of his mouth and utter a long drawn out cry which is a sign to everyone that he is a rich man.

At the present time people drink kava daily in Pentecost without special ceremony but it is also used at the feasts which take place after death, especially on the more important feasts on the tenth and hundredth days after death.

A story is told in Pentecost that a man once saw a rat nibbling a root of kava and watched it fall down and presently come to life again. He saw this happen several times and then tried the root himself, this being the beginning of kavadrinking.

Brother and Sister Avoidance 213

Lepers' Island.

This island, the native name of which is Oba or Omba, is the only part of southern Melanesia where, so far as is known, there exists definite avoidance between brother and sister. My information on this head comes from John Pantutun, and on such a matter his account can be trusted, for the custom had excited his interest greatly and he had attended closely to what had happened during his visits to that island.

In Lepers' Island a brother and sister never see one another after the girl is tattooed about the time of puberty. It is the girl who leaves the house of her parents and goes to live with her mother's brother. She will only do so, however, if she has a brother, and if there are only girls in the family they continue to live with their parents. If after the separation brother and sister meet on a path the girl will get out of the way and both will look down so as to avoid seeing one another. They will never say each other's names nor will they speak of one another. When a woman has married and had children it may happen that her brother will want to go to her house to see his nephew and in such a case the boy's mother will leave the house before her brother enters it. The avoidance even continues after death, for if one dies the other will not enter the house where the corpse is lying but will mourn outside.

It seemed as if there was a trace of a similar avoidance between brothers or at any rate of the respect which usually accompanies such avoidance. If a man is one of a crowd, all of whom are laughing, and the brother of the man comes on the scene the man at once leaves off laughing and becomes quiet.

John said that he had asked the people why brother and sister avoided one another but without result. He was inclined himself to think that its object was to avoid the possi- bility of sexual relations, being no doubt influenced by the undoubted relation between avoidance and sexual relations in other parts of Melanesia including his own island.

In Lepers' Island the relations between a man and his sister's son are of the same kind as in the Banks Islands.

I was told on Mota that in Lepers' Island it was the custom for a woman who wished to have illicit intercourse with a man first to approach the paternal aunt of the man and obtain her consent and it is probable that this practice

is also followed elsewhere in this part of Melanesia; that the father's sister not only chooses a wife for her nephew but also arranges more transient relations.

Anaiteum.

In the southern islands of the New Hebrides I obtained only one system, that of Anaiteum. This system was obtained from three boys at Vila, and is almost certainly correct in its main features, though I cannot answer for the correctness of the words phonetically. I found the sounds very difficult and to my ear quite different from those of Pentecost or the Banks Islands with which I had become fairly familiar. The terms for several of the relationships have been given by Inglis^ and the two sets of terms agree though our ways of spelling the words often differ considerably. I have thought it best to give the words as I heard them, but I have added in brackets the forms as given by Inglis.

The terms were obtained by means of incomplete pedigrees, and are given with the first person possessive.

Etmak. Father and father's brother. Probably also mother's sister's husband.

Resik. Mother, mother's sister, father's sister, mother's brother's wife, wife's mother and husband's mother.

Ngalo-unjak (inhalunjak). Child, son or daughter; brother's son, both man and woman speaking, and sister's son, woman speaking ; also son's wife and daughter's husband, woman speaking.

Matak. Mother's brother, father's sister's husband, wife's father and husband's father.

Ngawani-unjak. Sister's son and wife's brother's child ; also daughter's husband and son's wife, in both cases man speaking.

Etpok. All kinds of grandparent.

Mapok. All kinds of grandchild.

Natamng-unjak. Husband, husband's brother and sister's husband (w.s.) ; also mother's brother's son and father's sister's son (w.s.).

Engak. Wife, wife's sister and brother's wife (m.s.); also mother's brother's daughter and father's sister's daughter (m.s.).

• Dictionary of Aneityumese Language, London, 1882.

Nenga-unjak {nehgak). Wife's brother and reciprocally sister's husband (m.s.). Inglis gives natmetpok as an alternative term for this relationship.

Hudhnak (hudnak). Husband's sister and brother's wife (w.s.).

I have left till last the terms for brothers and sisters because I am uncertain about them. I obtained them early without paying much attention to certain irregularities and then became so interested in the other features of the system that I failed to go into these irregularities fully. It seems clear that etuak is a term used between two brothers, but it was also applied by a woman to her elder sister. Otherwise the term used between sisters was said to be natahingerak and this word was also used for his sister by a man. A woman was said to call her brother natamaingerak. If this information is correct, there are two terms in place of the usual reciprocal term for the brother-sister relationship, one of which is also used for the younger sister by a woman. Probably here, as so frequently in Melanesia, the terms for these relationships are in a state of transition.

The special interest of this system is in the terms which are brought into connection with one another by the crosscousin marriage. The mother's brother, the father's sister's husband, the wife's father and the husband's father are all classed together and are denoted by the same term. Next, the father's sister, the mother's brother's wife, the wife's mother and the husband's mother are all also classed together, but with the anomalous feature that they receive the same term as the mother and the mother's sister. Also, a man calls his daughter's husband or his son's wife by the same term as his sister's child and a woman classes her son- or daughterin-law with her brother's child. Lastly, the female cross-cousin has the same name as the wife while the male cross-cousin is classed with the husband. All these correspondences are exactly what is to be expected as a result of the cross-cousin marriage, though according to my informants this form of marriage is not now practised in the island.

Another interesting feature of this system is the difference in form of the first person possessive, which may be either a simple -k, -erak or -unjak. Mr Ray informs me that the first denotes closer possession than the others, a man calling his wife's sister enga-k, "my inseparable enga," while he calls his

2i6 Melanesian Society

wife's brother n-enga-unjaky "my separable enga." It will be noted that the closer possessive is used for the wife and her sisters and for the brother's wives while the less close form is used by a woman of her husband and his brothers. It will also be noted that the less close form is used between brothers and sisters while in the relationship between mother's brother and sister's son, the former takes the less, the latter the more separable form. This system was too hastily recorded to allow much weight to be attached to these differences but they are suggestive of a possible means of studying systems of relationship. The use of these different possessives, if constant, probably has some meaning and does not depend merely on euphony.

CHAPTER IX

THE SANTA CRUZ ISLANDS

In this chapter I give an account of material collected in the islands of Ndeni and Vanikolo and in the smaller islands, usually called the Swallow Group, but known locally as the Reef Islands, a term which I propose to use. I paid brief visits to several of these islands and was able to obtain a little information about them elsewhere, and here again I am indebted to Mr Durrad for a most valuable account from Santa Cruz. Even with this addition, however, the material is very scanty and in some respects contradictory and I only put it on record in order to give some indication of the interest of this region and of how little we yet know of its social organisation.

Santa Cruz Island.

I have three sources for my account of this island : — a few notes collected by myself, an account sent to me by Mr Durrad, and the notes of Wilhelm Joest published after his death by Baessler\

I obtained an unsatisfactory account of the social organisation from two young natives who came from the island of Temotu at the north-west corner of the large island. These boys said that there are four social groups called nau which take their names respectively from the Tnbu, a fish (the goltogoa of Mota) ; the mbua or shark ; the mbembla, a red fish resembling the trumpeter-fish; and the tambao or mammy apple (Papaia). The members of each nau may not eat the animal or plant after which the nau is named and they are also forbidden

• Neue Siidsee Bilder, 1900, p. 386.

2l8

Melanesian Society

:^

/

Social Organisation 219

to eat a number of other animals or plants, but the boys were only able to give me the names of the forbidden objects of their own groups. The mbu people may not eat the octopus {m6)y a sea-snake called vo, the fowl {kio) and a red yam called ningiamb while the mbembla people may not eat the turtle {vu), a sea-crayfish called londoi, the octopus and a big banana called papindo. The mbua people are allowed to eat the flesh of the white part near the tail of the shark but none of the rest of this fish. If a man catches a fish which he may not eat he gives it to a man of one of the other groups.

People of the same nau are not allowed to marry, but a man of one nau may marry a woman of any of the others. I was unable to obtain from these boys even such brief pedigrees as would allow me to obtain satisfactorily the mode of descent, there being evidently some strong inhibitory influence, probably an objection to say the names of their relatives. They said, however, that a man belonged to the same nau as his father.

People of a nau believe in their descent from the animal which gives the group its name and in one case there is believed to be a physical resemblance with the forbidden animal, the mbembla people being recognised by their red eyes.

Mr Durrad's account was obtained from a boy at the Mission School in the Banks Islands. According to this account there are nine social groups which take their names from the following animals: — the taklai, a bird; fishes called mbu [t\& goltogoa of Mota), mbembla (the matawut of Mota), nienda, naver or sea-eel and the blapupulu or fresh-water eel; the nadu, the porpoise or dolphin; the mbona, or pigeon; and the kiondo or bush-fowl.

The evidence seemed quite clear that a man belongs to the group of his mother. The boy from whom the account was obtained belonged to the nienda while his father was of the mbu group.

There is a story that the taklai bird hatched out a brood by a stream which carried out the nestlings to sea. Various fishes took charge of the young birds and began the different groups.

According to Joest there are twelve nau, a man of one nau always marrying a woman of another. These nau are named after animals and plants; several after sea animals

including the mbua or shark, the natu or dolphin, the betila or whale and fishes called nioda, mbu and mbilla; three after land animals, viz. the kuli or dog, the mbo or pigeon, and the kio or fowl; and three others after fruits, the talao or Papaia, the niaka and the kanalapiti. Anyone who eats these fishes or the pigeon will fall to pieces, his teeth dropping out first, while those people who belong to groups named after plants are not only prohibited from eating the plant but also may not dig it. The people of the dog group may not give any of their food to a dog. According to Joest also a man may not utter a word which forms part of the name of the animal or plant belonging to his nau. It is believed that the nau to which a man belongs can be told by inspection of the folds on the palm of the hand.

These three accounts leave no doubt as to the connection of exogamous social groups with animals or plants the use of which as food or otherwise is subject to certain restrictions. On the other hand, there are important discrepancies between the accounts, and certain features of these discrepancies make it almost certain that they depend on local differences of culture within the island.

Of these differences the most important concerns the mode of descent. Joest is quite clear about the fact of male descent and, though my evidence on this point was not satisfactory, I was also told that a man belongs to the group of his father. Mr Durrad's evidence, on the other hand, shows definite matrilineal descent. The divergencies in the lists of animals and plants from which the groups take their names, though probably due in part to defective information, are also such as might be expected to depend on local differences. The only name which occurs in the same form in all three lists is that of the fish called mbu but the mbilla of Joest is evidently the same as the mbembla of the other two lists. A similar change from mb to / also seems to have occurred in the name of the Papaia which I obtained as tambao whilst Joest spells it talao. The fowl occurs in all three lists though in my own it is only an accessory prohibited object of the mbu people. The shark occurs in my list and in that of Joest but is absent from that of Mr Durrad. The nioda of Joest is certainly the nienda of Mr Durrad'.

• See also 0'Ferrall,yi7«r«. Anth. Inst. 1904, xxxiv, 224.

System of Relationship.

Mr Durrad obtained this account by means of the genealogical method from the same boy who gave him the account of the social groups. The terms are as follows : —

Derde. Father, father's brother, mother's sister's husband.

Lainge, vocative form ida. Mother, mother's sister, father's brother's wife.

Malangi. Child, brother's child (m.s.), husband's brother's child, wife's brother's child and sister's child (w.s.). The term is also applied by a woman to her husband's younger brother.

Kalengi. Brother (m.s.), used without distinction of age.

Inwengi. Sister (m.s.).

Malwengi. Brother (w.s.). These three terms are also used for the children of the father's brother and the mother's sister according to sex and age.

Malamalwengi. Brother's son (w.s.), also a descriptive term.

Kambungi. Mother's brother and reciprocally sister's son (m.s.); also used reciprocally between a man and his sister's daughter's child.

Inwer derde. Father's sister, literally, sister of father, i.e. a descriptive term.

Kave. Mother's brother's child and father's sister's child. These relatives are spoken of collectively as kaveknge or kavekunge.

Derdende derde. Father's father.

Laiderde. Father's mother.

Malemalenge. Son's child, both (m.s.) and (w.s.).

Derdende lai. Mother's father.

Laindelai. Mother's mother.

Malainiangi or malaingiani. Daughter's child, both (m.s.) and (w.s.).

All these terms for grandparents and grandchildren are descriptive. Derdende derde is literally "father of the father," laiderde is " mother of the father," and so on.

Kanalangi. A reciprocal term for husband and wife. It is not however used by husband and wife, who address one another by their personal names.

Kandongi. The wife's father and mother and reciprocally

the daughter's husband. The term is used for the husband's elder brother and reciprocally for the younger brother's wife (m.s.); also for the wife's elder sister and reciprocally for the younger sister's husband (w.s.). Further, a man applies this term to the wife of his mother's brother.

Imbungi. The husband's father and reciprocally the son's wife (m.s.); also the sister's son's wife.

Lambungi. The husband's mother and reciprocally the son's wife (w.s.); also the husband's sister whether elder or younger and reciprocally the brother's wife (w.s.).

Tangi. Elder brother's wife (m.s.). The word was said also to mean mother, and this would agree with the reciprocal term for the husband's younger brother which is malangi.

Lomaingi. The elder sister's husband (w.s.) and younger sister's husband (m.s.). The latter is also called malemnge. Lomaingi is also applied to the father's sister's husband.

Iniengi. Wife's younger sister, the reciprocal terms being lomaingi and malemnge.

Malaki. The father of the son's wife and reciprocally the father of the daughter's husband.

Songe. The mother of the son's wife and of the daughter's husband.

Functions of Relatives.

The father's sister takes a newly born child away from its mother on the third day, bringing it back to the mother to be suckled. She gives a pig for the feast which takes place when the child's ears are bored to which the father's brothers are the chief contributors, but at the feast made when the child is first clothed she is herself the chief contributor. The father's sister chooses a wife for her nephew and gives a piece of feather-money to the mother of the girl she chooses.

A man and his mother's brother's wife may not see one another. Should this happen by accident the man will break several of his arrows and the woman will smash one or two of her water-bottles.

When a man dies his widow may not marry his elder brother, because she calls him ka^idongi and she may neither speak to him nor see him. She may, however, marry one of his younger brothers whom she calls malangi or child.

Similarly, a woman can speak freely to her elder sister's

Plate XII

Fig. I. 'illaye on Icniolu.

Fig. 2. Kouml house at Tcmotu.

Functions of Relatives ii'^

husband whom she calls lomaingi but she calls her younger sister's husband kandongi and may neither see him nor speak to him\

The system of relationship has several features of interest. In the first place there is the use of descriptive terms for several relationships, viz. the father's sister, the four kinds of grandparent and the corresponding kinds of grandchild. Next, there are the various features whereby a distinction is made between the elder and younger brothers and sisters of a consort, these being definitely associated with the presence of regulations which do not permit the marriage of a woman with her husband's younger brother nor conversation with the husband of her younger sister, and this latter usage almost certainly implies that a man may marry the younger sister of his wife but not her elder sister. A peculiar feature of the former kind of marriage regulation is that the man and woman who may marry address one another by terms which are said to mean child and mother.

It is clear that those relatives of the categories just considered who may not marry call one another kandongi, so that this term may be taken to connote prohibition on marriage. It is therefore interesting that it is also applied both to the wife's mother and to the wife of the mother's brother, though the fact that it is also used for the wife's father, i.e. used between males, makes one doubt how much value can be attached to its use in these cases. An unusual feature of the system is the existence of two terms for the relationship between the parents of a married couple.

A large village which I visited on Temotu (see PI. XII) contained both round and rectangular houses. The round houses occupied by the married people formed the greater part of the village and were grouped very close to one another, leaving only sufficient space to pass between them as is well shown in Mr Beattie's photograph (PI. XII, Fig. i). The much larger rectangular houses situated on the outer edge of the village are the mandwai or houses where the men eat and sleep. There were several mandwai but I could not ascertain with certainty whether all the men of one nau use the same building. Probably they did so at one time but not at present. So far as I could discover kava is not used in this island.

• For similar customs with other relationships see O'Ferrall, /o«r». Anth. Inst. 1904, XXXIV, p. 223.

Vanikolo.

In this island, often called Vanikoro, a brief account of the social organisation was obtained. There are ten exogamous groups, taking their names from the following objects: —

The mere, a fish called susu at Rowa in the Banks Islands ; the vesenamaka, a hermit-crab ; the vera or stingray ; the nomerue, a fish (the iga tnatur or sleeping fish of Mota, so called because it is only caught at night and is supposed to sleep during the day) ; the ive (the kwonglau of Mota), a sea-lion (?) ; the wanue (the gaiiase of Mota), a mullet ; ambumi, grass ; wire, water ; nepie, fire ; and tegniete, a bowl.

A person belongs to the group of his mother ; thus, my informant, Aulu, was of the wire group, his mother being a wire woman while his father belonged to the tegmete people.

In all the cases in which a group takes the name of a fish, this fish may not be eaten, while there are restrictions of a different kind for the people of other groups except the nepie people, in whose case I could not discover that there was any prohibition. The wire or water people may not drink the water of a certain bubbling pool ; the members of the tegmete division may not eat food prepared in a bowl and the ambumi people may not walk on grass.

In all cases the people trace their descent from the object from which they take their name. The nepie people believe that they are descended from a fire which, it is said, can still be seen ; the tegmete people trace their descent from a child which floated to Vanikolo from some other place in a bowl ; the ambumi people believe that grass gave birth to a female child ; the wire people are descended from water, and the remainder from their respective fishes.

The following terms were obtained from Reuben, a native of Rowa in the Banks Islands, who had been for some time a teacher of the Melanesian Mission in Vanikolo.

Aia, aianga. Father, father's brother and mother's sister's husband.

Papanga (^ pwapwanga). Mother, mother's sister, father's brother's wife, and also father's sister and mother's brother's wife.

Nisili. Child. According to Reuben the mother's

brother's son is also nisili, which, if correct, would be in agreement with the Banks Island practice, but on this point the evidence of a Banks Islander must be received with caution. It is, however, in favour of the correctness of the statement that Reuben said there was no term for the father's sister's child, so that in this case it is evident that he was not confusing the practice of his own island with that of Vanikolo.

Aluanga. Brother (m.s.) and sister (w.s.). No distinction between elder and younger.

Menenge {menenge ?). Term for the brother-sister relationship.

Nggeanga. Mother's brother.

Lamokanwani, Husband.

Venimingani. Wife.

Mulianga. Wife's brother and its reciprocal, sister's husband (m.s.) ; also husband's brother and its reciprocal, brother's wife (w.s.).

The rules about relationship by marriage are the same in Vanikolo as in the Banks. Those who offend have to pay turtle-shell and pigs. Reuben spoke of turtle-shell as the money of Vanikolo. Flat pieces of it are chiefly used in this way but pieces worked into rings are also used and have a higher value. There is no difference between Vanikolo and the Banks in the relations between a man and his mother's brother or nggeanga who helps his nephew to buy his wife, while uncle and nephew help each other with their gardens. If either kills a man, the other contributes towards the compensation-money, and they help each other in fighting, canoe- making and fishing. A man always accompanies his mother's brother when the latter goes in his canoe, and the nephew will obey his uncle more strictly than he will obey his father. The nephew can take anything belonging to his uncle.

A man greatly respects his father's sister whom he must obey. She does not, however, either choose his wife nor has she the power of vetoing his choice but she helps him to pay the wife's people. Reuben contrasted Vanikolo with the Torres Islands, saying that in Vanikolo the wife is bought and not in the Torres. A man can take anything belonging to his father's sister, and she may take from him but not to the same extent.

When asked about the relations with the grandparents in Vanikolo Reuben said that "you mould your life on that of

your mother's parents, you do all the things that they did, sacrificing and planting gardens where they did." In Vanikolo the grandparents will not take any share of the property of their son as happens in Rowa.

When a man's father and mother die the people cut down nearly all the coconut trees either possessed. A man will destroy the bows and arrows of his dead parents and also his own and will make new ones. When a child is born, on the other hand, the parents plant a great many coconut and Canarium nut trees.

The moon is the leno'i of the Vanikolans, their vui which they "worships"; it created them and the whole world. The moon is " worshipped " in a special house about a hundred yards from the village. Only men and boys gather together for this purpose, women being excluded. This house is much larger and higher than those in which the people live. The door of the house is on the side towards the mountainous interior of the island and within the house on the side nearest the village they build a platform about five feet high, about three feet broad and six feet long. On this they put four straight pieces of wood painted red, and over these four longer pieces are placed lengthwise to which coconut leaves are tied. At the four corners of the platform and extending above the platform as high as the roof of the house four upright posts, painted red and with pointed tops, are set up. The idea of having the door of the house towards the mountain is that the dead go to the mountain and the direction of the door gives the " worshippers " the right direction to their future home.

The house is built by one of the chiefs, who are now four in number, and he is the first to "worship" and then other people enter. The chief prays standing in front of the platform, underneath which there are three round stones. In front of these stones are two rows of skulls, ten in each row, and offerings of cold food are placed in front of each skull. It is when this food has been put out that the people begin to pray. The chiefs pray one after another as follows :

" May the heads of the dead, may the moon and the mountain guard us from our enemies that they may not hurt us."

" May the heads of the dead, the moon and the mountain

• I.e. Reuben regarded the Icnoi as the equivalent of the Banks vui and used a word for the attitude of the people which means " worship " in the Mota language.

Kava ^.TTi

make our gardens flourish, and give us plenty of taro, yams, tomago, breadfruit, nuts, coconuts, pigs, turtle-shell."

Then the first prayer is repeated.

The chiefs then dole out the food which has been offered to the skulls and when the people have eaten, the ceremony is at an end. The people stand during the prayer and only sit down when they begin to eat.

The preparations for the ceremony take about four days and the " worship " goes on for five days. It is done every year when the taro is ready for digging. When the ceremony is over, the skulls are taken away and put in bowls in the houses.

After the feast they drink kava. When each man drinks, he pours out the dregs and prays to the heads of the dead, the moon and the mountain that they may have a plentiful crop of kava.

In this island kava is chewed by four or five youths who put the chewed mass in a large bowl which is called wumeto. This bowl has no legs but is put on a stand so that it does not touch the ground. The boys who chew sit round the bowl and when they have filled it with their lumps of chewed root, one pours in water and stirs with a stick. When he has well stirred he leaves the kava to settle for a time, but there is no straining. Then he takes out the liquid in a coconut cup and hands it round to the chiefs first and then to the ordinary people. Kava is never drunk in Vanikolo except on ceremonial occasions. It is used in Utupua in the same manner as in Vanikolo.

It must be remembered that these notes were obtained from a native of another island but he was a most intelligent observer and I have little doubt of the essential accuracy of his account.

Reef Islands.

I visited two of these islands, Matema and Nukap or Nukapu, and obtained a little information from natives seen elsewhere.

On both these islands the language is predominantly, if not altogether, Polynesian, while physically the people have the appearance of being mixed, though in general inclining, so far as I could observe, much more to the Melanesian than to

15—2

the Polynesian type. I received the impression that the women had a more Polynesian appearance than the men, while some of them seemed to me to have a definitely Mongoloid appearance suggesting Micronesian influence.

In these islands there are definite exogamous social groups called mata. I heard of eight altogether in the islands, but only some of these are represented in any one island. Thus, in Nukapu there are four, the Pelembo, Pependal, Penvel and Pelengam, while on Pileni there are five, of which two, the Pelembo and Pependal, correspond with those with the same names in Nukapu, while the other three are the Pelewe, Pekuli and Pepali. The mata of Fenualoa are probably the Pelembo, Penepe, Pelewe, and Pependal or Vevendal. In Nukapu, and probably in the other islands, descent is in the female line, a man belonging to the mata of his mother.

The members of each of these groups are prohibited from eating certain animals and there are probably others which are either forbidden to the people as a whole or to certain combinations of the mata. The accounts I obtained were incomplete and in some cases conflicting, and the following account of the forbidden animals of each group must only be taken as an approximation to the actual conditions, for my informants were not certain about the animals of groups other than their own.

On Nukapu the Pelembo may not eat the turtle (/on); the ala/a, a fish called oa in Santa Cruz ; a small white fish called mat ; and the eel, both that of the land (navanga) and that of the sea [tuna). The Pependal may not eat the fishes called kos and onu, the latter being the one of Mota, while they are also forbidden to eat the alala. The Penvel may not eat fishes called aput (the valangansel of Moid), paniu and sinopil (}), and also the alala and the mat. The forbidden fishes of the Pelengam are the al and nef, the latter being the takwale of Mota.

On Pileni the Pelewe may not eat the flying fox (peke), the sting-ray (fae) nor a fish called awau said to resemble a Scorpaena. My informant was of this mata and was uncertain about the animals of the other groups but he knew that the Pelembo might not eat the fishes called fangafuea and ngadala, the latter being the same as the nef of Nukapu. These people do not eat the sea-eel while some eat the landeel {tuna) and others not, it being a matter of dispute whether

Reef Islands 229

it is the same animal as the sea-eel or not. At the present time all the people of Pileni eat turtle except in times of sickness when it is forbidden to all the mata.

The accounts suggest that the restrictions are breaking down to some extent, and this is certainly the case with the restrictions on marriage within the mata, for my Pileni informant knew of two recent cases of infringement, in one of which a Pependal man had married a Pependal woman, while in the other case two people of the Pelewe were concerned. Formerly both couples would have been killed.

The houses where the men eat and sleep are called afalau. A Nukapu man said that each mata had its own afalau, but that at present the Pelengam and Pependal have very few people and they use the houses of the Penvel and Pelembo. On Pileni the people were said to use the afalau without regard to the social groupings, but if there were a dispute between two mata, such as the Pependal and Pelewe, the Pelewe of one afalau would drive out its Pependal members who would join the Pependal of another afalau. Recently a Pependal man committed adultery with a Pelewe woman but the two groups separated without fighting, the Pependal giving money as compensation.

The only account of relationship which I obtained was from a young native of Pileni and I had no means of checking it. The following were said to be the terms of this island : —

Opa. Father, father's brother, husband of mother's sister.

Itei. Mother, mother's sister, wife of father's brother.

Ateliki. Child when the father is speaking.

Taku tama. Child, mother speaking.

Toga or toka (togaku). Term used between brothers irrespective of age.

Tukutungani. Brother (w.s.).

Tukutuahine. Sister (m.s.).

These three terms are used also for cousins of all kinds, i.e. including children of mother's brother and father's sister.

Ingoa. Mother's brother.

Ilamotu. Sister's son (m.s.).

Tupuna, apu and pu, the latter terms being those used in address. All four grandparents and also the father's sister. Grandparents collectively are called ngoku tupuna.

Makupuku. All kinds of grandchild and also the brother's child (w.s.).

Fungoku. Reciprocal term used between a man and his wife's parents. The father of a consort may also be called taniana fungovai.

Ngane. Wife's brother and reciprocally sister's husband (m.s.).

People are spoken of as the parents of their children, thus, the father and mother of Leposia would be called tamo Leposia and sino Leposia respectively. If the wife of the mother's brother had a child, she would always be addressed and spoken of in this way, and except by their own children parents are generally spoken of in the same way.

The exceptional feature of the Pileni system is that the term for grandparent is also used for the father's sister who in turn addresses her brother's child in the same way as a grandchild.

My Pileni informant compared the mother's brother to a schoolmaster. He scolds his sister's son if he does wrong, and uncle and nephew share each other's possessions. The father's sister is highly respected. She can forbid a marriage of which she disapproves and may take a part in arranging the marriage, but it is doubtful whether her power is greater than that of other relatives such as the grandparents who were said to be respected equally with the father's sister.

The most important relative in Pileni is the ngane, including wife's brother and the sister's husband. Those who call one another by this name both honour and avoid one another. One will not say the name of the other, and each will go aside if they have to pass. They only speak to one another from a distance, but each has to do the bidding of the other. They share each other's possessions.

Joking or chaffing is called vakada, this corresponding to the poroporo of the Ba:nks Islands. A man jokes those he calls toga, including the children of his mother's brother, except his own brother. A woman is never chaffed and if a man and woman were heard to joke one another it would be suspected that sexual relations had taken place between them.

A man may not marry the daughter of his mother's brother. She is called tukutuahine and is regarded as a sister.

In a Reef Island village there is a house called fale atua (see PI. XIII, Fig. i) which contains wooden stocks (PI. XIII, Fig. 2) called atua. Each of these stocks has a special name,

Plate XIII

Fig. 1. \fale atiia of the Reef Islands.

Fii

Interior o'i falc alita showing the stocks.

Ahta 231

the biggest one seen in a house on Pileni being called Tangteala {tang, cry, te ala, the path, i.e. crying in the pathway). A stock is made and painted and then borne on the shoulders of a man along the beach to the village where it is to be kept, the bearer running with it into the fale atua and putting it in its place. A drum is then beaten and feasting and dancing follow. Before the food is distributed, some of it is laid before the stock which is addressed as follows : — " Here is your food! Do you show your affection for us and keep sickness from touching us."

Anyone who goes on a journey by canoe and meets a storm will call on one of these atua by name, saying " Have pity on us ! Give us a good wind ! " and at the same time food is thrown on the top of the house of the canoe. (The canoes of the Reef Islands are large enough to have houses on them.) If the wind changes, thanks are offered to the atua and a feast held when the traveller returns home.

The term atua is applied to certain other beings and objects. It is not given to all the animals forbidden as food to the mata or clans, but the flying fox and the nakatc, probably a large sting-ray, are both called atua. This name is also given to ghosts of the dead and to certain stones situated in places which are not visited by anyone. One of these places had been seen by my informant when in search of a lost arrow. The stone was about two feet high surrounded by a low fence of stones. Formerly a ceremony was performed by a chief at this stone in which he buried some food left by an enemy, asking the stone to make the enemy die.

So far as I could discover kava is not drunk in the Reef Islands.

CHAPTER X

THE SOLOMON ISLANDS

There are several distinct cultural regions in these islands, with great differences in the mode of social organisation and in the way of counting relationship. In the larger islands of the Solomons there are two distinct populations, the people inhabiting the coast and those of the interior, who may be spoken of as the coast and bush people respectively. These people are hostile to one another and at the present time we know absolutely nothing of the social organisation of the bush people so that this chapter will deal exclusively with the people of the coast.

Beginning at the south-eastern end one cultural region may be said to comprise San Cristoval, Ulawa, Malaita (or Mala) and possibly the south-eastern end of Guadalcanar. We know at present very little about this region and probably considerable differences of culture will be found within it, such differences being indicated by the variations in the systems of relationship. Another region comprises the islands of Florida or Nggela, Ysabel and Guadalcanar (probably only its north-western half). The island of Savo may be included in this district though differing from the others in language and some other respects. A third region which may be spoken of as the Western British Solomons comprises the islands or groups of islands called New Georgia including Ruviana; Kulambangara or Nduke ; Eddystone Island also called Simbo, Narovo and Mandegusu ; Ganongga (usually hitherto called Ranongo) and Vella Lavella. The Shortland Islands, Treasury Island and the islands of the Bougainville Straits form another region which will not be included in this survey since it will be independently dealt with by Mr G. C. Wheeler. There remains in the British part of the group the island of Choiseul

"o

V

■ST/ a

-;^:::>

^ «

<3

^

I

^ fl

\

234 Melanesimi Society

about which we know at present too little to enable us to assign it to any of the cultures which have just been mentioned, but it is probably most nearly allied to the islands of Bougainville Straits. The island of Bougainville^ is also included in the Solomon Islands. Though I did not myself visit it, I am able to give the system of relationship of one district (see p. 258).

The district of the Western British Solomons will be fully dealt with by Mr Hocart and myself in another work and I give here only a brief outline of its systems of relationship. I was also able during brief visits to obtain several systems of relationship and make a rough study of the social organisation of the two eastern regions.

The Malaita and San Cristoval Region.

Little is known of the social organisation of this region. There is little doubt that in parts patrilineal succession and inheritance exist with no organisation in clans or other similar groups, but the Rev. F. H. Drew tells me that there are probably clans which take their names from, or are otherwise connected with, birds, possibly with matrilineal descent, in some parts of San Cristoval. The coast people consist of many different tribes, speaking different dialects and probably having many differences in culture.

The most satisfactory system of relationship was obtained in the island of Ulawa near the southern end of Malaita, its correctness being kindly confirmed by the Rev. W. G. Ivens of the Melanesian Mission. This island has probably a social organisation similar to that of the southern end of Malaita. The following are the terms of the system : —

Ainaku (with the first person possessive), in direct address mamau. Father, father's brother, father's sister's husband and mother's sister's husband, in all cases in the usual classificatory sense.

Nikeku, in address teitei. Mother, mother's sister, father's sister and the wives of the father's brother and the mother's brother.

1 I have suggested elsewhere {Geograph. Jottrn. 1912, xxxix, 458) that this island might more suitably be grouped with the Bismarck Archipelago in spite of its geographical connection with the Solomons. It is not only one with them politically, but its general culture falls more into line with that of the Bismarck Archipelago than of the Solomons.

Kaleku. Child, used as the reciprocal of amaku and nikeku.

Ulaku. The term given by a man to a brother and by a woman to a sister and used independently of respective age. It is applied in the same sense to all four kinds of cousin, the children of both brother and sister of father or mother.

Iniaku. The term for the brother-sister relationship. It is used also between cousins of all four kinds when these are of different sex.

Uweliku. A reciprocal term for the mother's brother and the sister's child.

Wauwaku. The reciprocal term used between the grandfather and grandchild, being applied to the father of both father and mother and reciprocally to the child of either son or daughter when a man is speaking.

Pwapwaaku. The corresponding reciprocal term for grandmother and grandchild.

Hungaoku. A reciprocal term for the father and mother of husband or wife on the one hand and the corresponding four kinds of son- or daughter-in-law on the other hand.

Iheku. Used for the eight different kinds of brother- and sister-in-law.

The special interest of this system is its exceptional simplicity. All relatives of the generation above, other than those related by marriage, are a^naku or nikeku except the mother's brother and corresponding relatives in the classificatory sense, and all of the same generation call one another either ulaku or iniaku according to sex. Similarly, all the relatives of the husband or wife of his or her generation are iheku while all of the preceding generation are hungoku. The feature of reciprocity is very decided, applying both to the relationship of parent- and child-in-law and to that of grandparent and grandchild, and only in the case of the grandparent is there a distinction according to sex.

I am indebted to Mr Ivens for the information that the system in use at Saa on the west coast of the southern end of Malaita is the same except that the distinction of brother and sister according to the sex of the speaker is not present, both brothers and sisters being called asiku, independently of age. There is also here a term given by the parents of a man to the parents of the man's wife and reciprocated, viz. akaro, but this is also a general term for relatives by marriage, the term for a marriage feast being aharota.

In Ulawa relatives are spoken of collectively by adding the prefix rohai (ro, two and Aat, the reciprocal prefix), thus, rohainikeina means mother and child, and a man speaking of his wife and child will say ''rohainikeina inau!' Similarly rohaihungauna is used of those who stand in the relation of parent- and child-in-law and rokaiihana of those of the same generation who are related by marriage.

Marriage is regulated by kinship ; a man will not marry one whom he calls iniaku in Ulawa or asiku in Saa, i.e. he will not marry the daughter of the brother or sister of either father or mother, using these terms in the classificatory sense.

In the island of Malaita I obtained two incomplete systems from the districts of Lau and Fiu, and these have since been confirmed and supplemented by the kind help of the Rev. A. J. Hopkins of the Melanesian Mission. The following are the Lau terms : —

Maa nau [nau being the first person possessive). Father, father's brother and mother's sister's husband in the classificatory sense.

Tei nau. Mother, mother's sister, and father's brother's wife.

Mwela nau. Child ; mela being the reciprocal of maa and tei.

Auwa nau. The term for the elder brother when a man is speaking, used also in the same sense for son of the father's brother or mother's sister when older than the speaker.

Sasiku or ai imburi. Terms for younger brother (m.s.) and for a younger sister (w.s.), imburi meaning " after." Sasi is also used in a wider sense by both men and women for those of the same sex, whether brothers (m.s.) or sisters (w.s.), cousins or friends.

Ai nao. Elder sister (w.s.), inao meaning "before."

Mwaina nau. The term for the brother-sister relation and for the corresponding relationship between the children of two brothers or of two sisters.

Aiya nau. A reciprocal term applied to the father's sister and used by her of and to her brother's child.

Ko nau. The mother's brother, the wife of the mother's brother, the husband of the father's sister* and also the four kinds of grandparent, the term being in all these cases reciprocated, so that it is applied to all kinds of grandchild as

• According to one informant this relative is called maga.

well as to the sister's son, the husband's sister's child and the wife's brother's child.

Di nau. The child of the mother's brother and the father's sister, i.e. a term for cross-cousin.

Arai nau. Husband.

Afe nau. Wife.

Fungo nau. A reciprocal term for parent- and child-inlaw, including father or mother of husband or wife and the husband or wife of son or daughter. Also used as a general term for relatives by marriage.

Luma aa nau. The sister's son's wife and sister's daughter's husband (m.s.) and probably other relatives by marriage.

Bara nau. I was given this term for all kinds of brotherand sister-in-law, irrespective of sex of speaker, but Mr Hopkins heard that it is a general term for relatives by blood, bara being the fence which marks off the group {J. village).

Sata nau. A word used generally for friend and applied to one another by the husbands of two sisters, and by the parents of a man and wife.

Saula nau. The wives of two brothers.

This system stands in a most interesting relation to those of Ulawa and Saa. It resembles them in having few terms for relatives by marriage but differs in having a term for the father's sister as well as for the mother's brother and in distinguishing cross-cousins from the children of two brothers and two sisters. In connection with the latter distinction it is noteworthy that the husband of the father's sister has the same name as the mother's brother. An exceptional feature of the system is that the grandparents, or rather the grandparent-grandchild relationships, are denoted by the same term as is used for mother's brother and his sister's child and other relationships of persons only one generation apart.

In the Lau district it was said that a man must marry outside his village^ and no case could be discovered in two or three pedigrees where a marriage within a village had taken place. This is the only indication of village-exogamy which came under my notice in Melanesia.

The following are the terms from Fiu, a place on the west coast of Malaita.

Maa nau, maaka nau or maaku. Father, father's brother, and husband of the mother's sister.

1 I do not know the exact nature of the institution here spoken of as a village.

Tea nau. Mother, mother's sister and father's brother's wife. The mother's sister is also called tea fo/o nau, fofo meaning "upon."

Mwela nau. Child.

Ai naoku. The elder brother (m.s.), or elder sister (w.s.).

Aiimburiku or sasiku. The younger brother of a man; the former is also the younger sister of a woman.

Mwaunzvane nau. The sister of a man and brother of a woman. These terms are used for the children of the father's brother and mother's sister according to the usual rules.

A'ai nau. Father's sister and wife of the mother's brother.

Mwae nau. Mother's brother and father's sister's husband.

Di nau. Cross-cousin.

Kdo nau. A reciprocal term for the grandparent-grandchild relationship.

Arai nau. Husband.

Afe nau. Wife.

Hunga nau or funga nau. A reciprocal term for the parent- and child-in-law relationship.

Luma nau. Wife's brother and husband's brother, and reciprocally sister's husband and brother's wife when a man is speaking. The sister's son's wife and sister's daughter's husband are distinguished as lu7na aa nau. Luma is the word for "house."

Sai laku. Husband's sister and its reciprocal, the brother's wife of a woman ; also the wife's sister's husband and husband's brother's wife.

There is here a close resemblance with the Lau system, not only in form, but in the actual terms used. It therefore becomes a matter of importance that kdo nau, evidently corresponding to the /^^ nau of Lau, is at Fiu the term for the grandparent-grandchild relationship only, which makes it probable that in the Lau system this word is primarily the term for grandparent and has been extended to the mother's brother and other relatives of that generation. This argument is strengthened by the fact that the word for grandparent in Florida and Guadalcanar (see pp. 242 and 244) is kukua, to which the kdo of Fiu is evidently related, the place of k being taken by a break.

Another important point is that the father's sister and the wife of the mother's brother are denoted by the same term

in Fiu and that there is only one term for mother's brother and father's sister's husband in both systems. Aai nau is evidently the same word as the aiya nau of Lau, the consonant being represented by a break.

There can be little doubt that we have in these examples two variants of one system in which the mother's brother and the husband of the father's sister were denoted by one term and the father's sister and the wife of the mother's brother by another while the term for grandparent was originally different from either.

A striking feature of these two systems of Malaita is the occurrence of three kinds of possessive. That which occurs most frequently is nau as a separate word following its noun, but for several relationships we have the suffix -ku (first person) which in one form or another is so widely spread throughout Melanesia. In the case of certain relatives by marriage in Fiu the possessive is laku and Mr Hopkins thinks that sai with which this form occurs is related to sata, friend, which takes the possessive nau, the form in laku showing a closer connection. It may be noted that the term sasi which bears the suffix -ku is evidently the same word as the asiku of Saa.

From the island of San Cristoval I am able to give two systems from the districts of Heuru and Rafurafu respectively. Both are incomplete but certain leading characters come out clearly.

In the Heuru system, of which I have also received an account from Mr Drew, amagu is the term for the father, the father's brother and the mother's sister's husband. Inagu is mother and mother's sister. Garegu is the term for child. Men call their brothers, and women their sisters, doora, irrespective of age, while the term for the brother-sister relationship is asigu. The mother's brother is mauagu. It is doubtful whether there is a special term for cross-cousins. The term for the grandfather-grandchild relationship is uwaiagu and this term may also be used for the father's brother. The grandmother-grandchild relationship is waeagu (?). The reciprocal term for parents-in-law and children-inlaw is hungogu. Several of the terms are the samie as in Ulawa, the Heuru system thus appearing to be a more complex form of the Ulawa system.

The Rafurafu system, for which I am indebted to Mr Drew,

is very different and has a character which I did not find elsewhere in Melanesia, viz. the presence of two prefixes, one for males and the other for females ; thus, the elder brother is waogaku and the elder sister kaogakw, the wife's father is fongoku while the wife's mother is kafongoku. The wife's brother is waforoku and the wife's sister kaforoku.

The following are the terms of the system : —

Wama. Father, father's brother.

Keinaku {katitd). Mother, mother's sister and father's sister.

Kareku. Child, the reciprocal of the preceding terms.

Waogaku. Elder brother.

Kaogaku. Elder sister, both man and woman speaking.

Wasiku. Younger brother.

These terms are also used for the children of the mother's brother according to sex and age.

Mamau. Mother's brother and, reciprocally, sister's son.

Wauwa. Father's father and mother's father and reciprocally grandson. A grandson may also be called wasiku kare, apparently a descriptive term, son of my younger brother.

Kakaku. Father's mother and mother's mother.

Fifaneaku. Husband.

Mwaneaku, Wife.

Fongoku. Wife's father, husband's father, wife of son or husband of daughter, and also, father's sister's husband.

Kafongoku, Wife's mother, husband's mother (.'*), and also wife of mother's brother.

Waforoku. Wife's brother, sister's husband both (m.s.) and (w.s.), and husband's brother.

Kaforoku. Wife's sister, brother's wife (m.s.), husband's sister and its reciprocal, brother's wife (w.s.). In idiCi, foroku is a general term for all the eight kinds of brother- and sisterin-law, males and females being distinguished by the appropriate prefix.

The terms for the grandparents are particularly interesting in relation to those of Ulawa. The Rafurafu term for each kind of grandfather is wauwa, the same term as is used in Ulawa, so that it is probable that the initial wa of the latter is a surviving male prefix. The term for the two kinds of grandmother is kaka or kakaku which is to be compared with xh^ pwapwaaku of Ulawa, in which there appears to have been an interchange between k and pw. It seems highly probable

Florida 241

that in its terms for grandparents the Ulawa system retains prefixes denoting sex corresponding with those still found in the Rafurafu system of San Cristoval.

Florida, Guadalcanar, Ysabel, Savo Region.

The culture of this region differs in many important respects from that of the islands which have just been considered. In Florida, Guadalcanar, Ysabel and Savo there are exogamous clans with matrilineal descent. The number of these clans differs in different parts of the region, varying from six in Guadalcanar to two only in one part of Ysabel. With each clan there are associated a number of objects ; animals, plants, material objects or human beings, which may be regarded as sacred, the name for these objects in most of the islands being some variant of the word tindalo. The systems of relationship of three of the islands are very similar to one another, though there are important differences in detail. But that of the fourth, Savo, differs greatly, especially from the linguistic point of view.

The social organisation and systems of relationship of the four islands will be now considered in detail. The accounts to be given were obtained during brief visits to the islands and are certainly imperfect and fragmentary.

Florida or Nggela,

In this island there are at the present time, so far as I could discover, only four clans or kema; the Honggokama or Manukama, Honggokiki, Nggaombata and Kakau, two others recorded by Codrington, the Himbo and Lahi, having apparently disappeared. I did not obtain in this island any account of the sacred objects or tindalo of the different kema, but these have been recorded by Dr Codrington\

The following are the terms of the Florida system of relationship : —

Tamanggu, mama (vocative form). Father, father's brother, father's sister's husband and mother's sister's husband. One man also used this term for the mother's brother.

Tinanggu or ino (vocative form). Mother, mother's sister, father's brother's wife and father's sister.

1 M., 31.

R. M. S. I<5

Dalenggu. Child, the reciprocal of the two preceding terms in all their senses.

Tiiganggu or hoganggu. This is applied by a man to his elder brother and by a woman to her elder sister and is used in the usual classificatory sense for all members of the clan of the same generation when they are of the same sex and older than the speaker. It is also applied in the same way to the children of the father's brothers, although these may be of different clans.

Tahinggu. The corresponding term for the younger brother of a man and younger sister of a woman ; it is the reciprocal of tuganggu and is used in the same wide sense as this term.

Vavinenggu. The reciprocal term for the brother-sister relationship, used in the same wide sense as the terms applied to one another by persons of the same sex.

Tumbunggu. The term for mother's brother which is used reciprocally for the sister's son of a man. It is also used for the husband of the father's sister who was otherwise called by the same term as the father. According to Dr Codrington* the brother of the father's mother is also tumbunggu. I was not able to confirm this and it is possible that this usage has now disappeared or is in course of disappearance.

Mavunggu. The term for cross-cousins, the children of mother's brother or father's sister, but these relatives are also sometimes called tuganggu, tahinggu or vavinenggu according to age and sex, i.e. cross-cousins are often not distinguished from cousins of other kinds.

Kukuanggu. A reciprocal term for all four kinds of grandparent and four corresponding kinds of grandchild.

Vunonggu. A reciprocal term for the four kinds of parentin-law and the corresponding four kinds of child-in-law. It is also used for the wife of the mother's brother who in her turn uses it for the sister's son of her husband.

Ivanggu. The term for all the eight kinds of brother- and sister-in-law.

Husband and wife call one another tauunggu.

Guadalcanar.

There are probably considerable differences in the details of the social organisation of different parts of this island and

' M., 41.

Guadalcanar 243

the following account, obtained only during a hasty visit to the north-western end of the island, even if approximately correct for one district, will certainly not hold good for the whole island.

The exogamous groups are in general^ called kema, and their sacred objects tindao. The kema are six in number, viz. the Lakwili, Kindapalei, Haumbata, Kakau, Kiki and Simbo.

The Lakwili have as tindao certain men who were said to have been the first men of the clan, certain images and two animals, the mauvo or eel and the kohe, a. small fish, neither of these fish being eaten by the members of the group. The Kindapalei have as tindao their first man, together with a snake called choholisi, the moon and sun spoken of together in one word as vulamanaso, and a sacred fire called lake tambu. The Haumbata have also their first men, a shark called baheanapombo and the naroka, a pigeon, neither shark nor pigeon being eaten. About the other groups the information was less definite, but it seemed that neither the Kakau nor Kiki could eat the shark, while the Simbo people were forbidden the monitor lizard.

In Guadalcanar the tindao are much reverenced, and my informant, a Christian, said that they were worshipped. If a Haumbata man wishes to kill an enemy on the land he goes to a special place belonging to the naroha bird and calls on this bird for mana (power) and susuliha (strength). He offers food in the form of a pudding as well as fish, pork and tobacco, and the bird gives him the mana which enables him to kill his enemy. If, on the other hand, the Haumbata man wishes to kill his enemy on the sea, he makes offerings to baheanapombo, and this shark will break the canoe and eat the enemy.

Though the restrictions on their use as food show that baheanapombo and naroha are used of species of shark and pigeon respectively, these animals were often spoken of as if they were individuals. This is certainly so in the case of the sacred snake (choholisi) of the Kindapalei which is a very big creature living on a rock by a place called Koli. The place is forbidden to all except the Kindapalei, and the people of that division only go there to "worship" him. If others

1 According to Mr Woodford {A Naturalist among the Head-hunters, London, 1890, p. 40) they are called kua at Veisali.

16 — 2

244 Melmtesian Society

wish to pass the place, they must go in a canoe or walk far out on the reef. The Kindapalei people take puddings and other offerings to the snake and obtain 7nana from it just as the Haumbata get mana from their bird naroha. The Kindapalei also obtain mana from the sun and moon, and from the sacred fire called lake tambu. The latter is a fire which springs out of the rock at a certain place, and when the people take offerings these are burnt in the fire. If the people kill a man they take the tongue and lips as an offering to lake tambu. There is a special spot sacred to the moon and sun, vulamanaso, where similar offerings are made.

If the animal tindao are eaten, the result is illness; thus a man who eats the naroha bird falls ill, and blood comes from his mouth and nose. To make him well, offerings are made to the bird and the man may recover. If the baheanapombo is eaten sores break out on the body of the offender.

The terms of relationship were obtained from a native of the north-western end of the island and probably they only hold good for that district. They were : —

Tamanggu, in address mama. Father, father's brother and mother's sister's husband.

Tinanggu, in address koni. Mother, mother's sister and father's brother's wife.

Dalenggu. Child.

Tasinggu. Used for both elder and younger brother by a man. A sister is also tasinggu or tasinggu kakavi. These terms are also used for the children of the father's brother and the mother's sister and all members of the same clan or kema of the same generation.

Nianggu. Mother's brother and reciprocally the sister's son of a man ; also the wife's father and probably also the husband's father. It is possible that it is also used for the husband of the father's sister.

Tarunganggu. Father's sister, the wife of the mother's brother and the wife's mother ; probably also the husband's mother.

Kukuanggu. A reciprocal term for grandparents and grandchildren.

Ivanggu. A term including not only brothers- and sister-in-law as in Florida and Ysabel but also the crosscousins.

Ysabel 245

Ysabel.

Here again there are certainly different social groupings in different parts of the island. In general there seem to be three exogamous groups which are called vinahuhu, the sacred objects of which are called tindadho, but I am indebted to the late Dr Welchman of the Melanesian Mission for the information that in one part of the island there are only two groups and that at Kia at the north-western end of the island each of the three main groups is divided into a number of smaller sections.

The three vinahuhu are called Vihuvunagi, Posomogo and Dhonggokama. The chief tindadho of the Vihuvunagi is the manuhutu, that of the Dhonggokama the 7nbelama or frigate-bird, and that of the Posomogo, the higara, a parakeet. In each case the tindadho may not be eaten. It is probable that each of the vinahuhu has other tindadho, but I could only learn about those of the Vihuvunagi who were said to possess the shark {ele), crocodile {vua), snake {poli), eel {oloi) and thunder (rete), the four animals being forbidden as food. It is possible, however, that some or all of these are sacred to the whole community and are not the exclusive tindadho of any one clan. I was given an account of the origin of the Vihuvunagi. An eagle {manuhutii) called Vihuvunagi laid eggs in a nest on a big tree and when the eggs broke there came out of one a woman who lived in the tree for a time and then, coming down, was made a chief. She had many children and sent them all over the island and it is from them that the Vihuvunagi are descended.

At Kia at the extreme north-western end of the island each of the three primary clans is divided into a number of groups, each of which takes its name from a natural object. The following account was sent to me by Dr Welchman : —

The Vihuvunagi have six subdivisions, taking their names from the mbahei, a tree like the banyan ; the mbaumbahulu, a blue pigeon ; the rumu or dugong ; the hombarae, a tree with edible leaves ; the mamara or paper-mulberry, and the etingi, a large banana. The Dhonggokama had nine divisions, nggahili, a white cockatoo ; the taunu or sun ; the gogosulu or porpoise ; the nggenggefe or flying fox ; the rurugu, the bivalve Unio ; the paike or cuscus ; the kopi or boatbill heron ; the memeha sondu or toucan, and the vavalu, a fish.

The last two divisions are now extinct, one having disappeared during the last five years and the other within fifteen years.

The Posomogo have twelve divisions, the kusa, (?) kingfish ; the tengge or turtle ; the posa, a black banana, wild and inedible ; the ronggisi, a shell-fish, Pteroceras lambis ; the kikituru, a night-bird, resembling a picture of the gerfalcon ; the mbulau or Conus genei^alis ; the etieti or large areca nut ; the sesehu, a grass ; the fafaleke, a clam ; the koilo or bright yellow coconut ; the taraoa, a yellow land-snail [Carocolld), and the pakehana, (?) a small sword-fish.

I give two systems from Ysabel, from the districts of Bugotu and Nggao. The former was collected by myself but has been confirmed in its general features and supplemented by the help of Mr Edmond Bourne of the Melanesian Mission. For the Nggao terms I am wholly indebted to Mr Bourne,

In the Bugotu district the terms are as follows : —

Tamanggu, vocative form mama. Father, father's brother, father's sister's husband, and mother's sister's husband. Mr Bourne tells me that it is also applied to the husband's father while the vocative form, mama, is used in addressing the wife's father.

Idonggu^. Mother, mother's sister (who is also called varua idonggu), mother's brother's wife, father's sister and father's brother's wife. It is probably used also for the husband's mother.

Dadkenggu. The term for child, reciprocal to the two preceding.

Toganggu, tahinggu and vavinenggu or vaivinenggu. Used in exactly the same sense as the corresponding terms in Florida. In familiar address the term kaka is used for elder brother (m.s.) and elder sister (w.s.), while iku is similarly used in addressing a younger brother or sister.

Tumbunggu, tu77ibuni. Mother's brother and sister's son of a man. It is also used reciprocally between sister's son's child and the mother's brother of the father, who may also, however, call one another dadhenggu and tamanggu.

^ This word was given to me in the form indonggu, perhaps owing to the fact that my informant, though belonging to Ysabel, had lived for most of his life in Savo and Guadalcanar. Mr Bourne tells me that the proper Ysabel sound in this word is that of the simple d, although in general the d is nasalised in this as in the other islands of the region. According to Dr Codrington (Af.Z., 546—7) the use old ox «</ characterises different "families" or sets of people of the island.

Relationship 2^'j

Pajanggu^. Cross-cousin. When a man speaks of this relative he uses the form tamapajani.

Kuenggu (voc. kue). Male grandparent, both father's father and mother's father.

Kavenggu (voc. kave). Female grandparent.

Kukuanggu. Grandchild, reciprocal to both kuenggu and kavenggu ; used also for the sister's daughter's child who may also be called dadhenggu.

Tauunggu. A reciprocal term for husband and wife, but it is not used now, the term kulanggu, meaning "my friend," taking its place.

Vungaonggu or vunonggu. A reciprocal term for parentsand children-in-law. It is also applied to the sister's son's wife and the sister's daughter's husband and is used reciprocally by them for the mother's brother of the consort.

Ivanggu. The term for all kinds of brother- and sisterin-law.

The following are terms of the Nggao district of Ysabel: —

Manggu [mage). Father.

Donggu (voc. doge). Mother.

Tunggu. Child.

Tiganggu (tiga). Elder brother (m.s).

Tahinggu \kuge). Younger brother (m.s.).

Ka-rudunggu. Brother-sister relation.

Nimbunggu. The reciprocal term for mother's brother and sister's child.

Fakdrainggu. Cross-cousin.

Kuenggu (kue). Mother's sister's husband.

Garanggu. The reciprocal to the last for the wife's sister's child.

Ndnonggu. Sister's son's wife.

Kavenggu {kave). Father's mother.

Ketonggu and kerague. Terms applied to the husband and (?) also to the wife.

'Nogu. Wife's mother.

Notifaie (vave). Wife's sister.

Savo.

The social organisation of the island of Savo resembles in its general features that of the other three islands of this

• This word was given to me in the ioxva panjanggu.

region, the resemblance being especially close with Guadalcanar, as might be expected from the geographical position of the island. I was told of five matrilineal clans, here called ravu, VIZ. the Lakwili, Gaumbata, Dhonggo, Kakauga and Kikiga,

Each group possesses objects corresponding to the tindao of Guadalcanar, here called manjali. The Lakwili do not eat a small monitor lizard called sangavulu ; the Gaumbata reverence an image and a ghost-woman and do not eat a monitor lizard called vava ; the Dhonggo have as manjali a ghost-man and the bird called tambu koso (sacred bird) which is the same as the manuhutu of Ysabel ; the Kikiga have three animals forbidden as food, the shark, a large flat-fish called limanibarava and a pigeon called kurau. The animal manjali seemed to be less sacred than in Guadalcanar, for my informant, who had told me that the tindao are " worshipped " in Guadalcanar, was emphatic that this did not take place in Savo but that the animals were merely tambu and not eaten.

Although the social organisation has thus the same general characters as in other islands of this region, the people of the island speak a different language, Savo being one of those places in which languages are found widely divergent from the general Melanesian type. This difference has been already instanced in the names for "clan " and for the objects associated with it and is still more apparent in the system of relationship which differs widely from the other systems of this region, not only in vocabulary, but still more strikingly in the nature of its possessive pronouns. As we have seen, in most other parts of the Solomons these take the form of suffixes so intimately blended with the actual terms of relationship that the two are never heard separately. In Savo, on the other hand, the possessives are separable prefixes which vary with the sex of the person addressed.

The following are the terms, the possessive being given as usual in the first person singular : —

Mau or ai mau. Father, father's brother and mother's sister's husband.

Mama or ai mama. Mother, mother's sister, father's sister, father's brother's wife and mother's brother's wife.

Aiva nyumba. Son, when a man is speaking, i.e. the reciprocal of mau.

Savo 249

Aiva zumba. Son, when a woman is speaking, i.e. the reciprocal of mama.

Daughters are aima nyumba and aima zumba respectively.

Aiva toni. Elder brother when a man is speaking.

Aiva tot. Younger brother (m.s.).

Aima toka. Elder sister whether man or woman is speaking.

Aima totoa. Younger sister, whether man or woman is speaking.

Aiva toka. Elder brother (w.s.).

Aiva totoa. Younger brother (w.s.).

It will be noticed that in each case the choice of the prefix depends on the sex of the person addressed or spoken of, while the term of relationship itself varies with the age and sex of the speaker. These terms are used also for cousins of all kinds, whether children of father's brother, father's sister, mother's brother or mother's sister, the same distinctions being observed according to sex and age as in the case of actual brothers and sisters.

Aiva kulaga. Mother's brother. He addresses his sister's son as aiva kulaga and his sister's daughter as aima kulaga.

Aiva kukua. Grandfather and grandson.

Aima kukua. Grandmother and granddaughter.

Aiva vungau. Wife's father, husband's father, and daughter's husband.

Aima vungau. Wife's mother, husband's mother and son's wife.

Ai manggu. Wife's brother and reciprocally sister's husband (m.s.).

Aima manggu. Wife's sister, husband's sister and brother's wife (w.s.).

Aiva manggu. Sister's husband (w.s.).

Aiva mbasa. Husband's brother.

Aima mbasa. Brother's wife (m.s.).

As this system is so exceptional it may be useful to record a full list of the possessives. They are as follows : —

Ai mama, my mother ; no mama, thy mother ; lo mama, his mother ; ko mama, her mother ; mai mama, our mother (inclusive) ; avei mama, our mother (exclusive) ; pei mama, your mother ; zei or dhei mama, their mother.

Elder and younger brothers of the father may be distinguished as aiva nduzi 7nau, my great father, and aiva nyart mau, my little father.

The three systems of Florida, Guadalcanar and the Bugotu district of Ysabel are evidently closely related to one another. One of their interests arises from the evidence of change now going on in them. Though there is a definite term in Florida for the mother's brother, viz. tumbu, it is clear that this relative is now often called ta77ia, otherwise used for the father. Further, it is probable that the use of tumbu for the brother of the father's mother which was recorded by Dr Codrington thirty years ago no longer exists, though this usage is still present in the Bugotu district of Ysabel. Again, cross-cousins who are usually mavu to one another sometimes address one another as if they were brothers or sisters. There can be little doubt that the process which is going on here is one of simplification, of the loss of terms which once existed rather than the addition of new terms accompanying the formation of new concepts of relationship. The systems of this region are already very simple, vastly simpler than those of the Banks or Torres Islands, and if as I suppose they are undergoing still further simplification, they would seem to be approaching in character such extremely simple systems as those of Ulawa and Saa.

The Savo system is also simple, its only exceptional feature, so far as structure is concerned, being in the nomenclature for brothers and sisters, the words toka and totoa used between a brother and sister being also used between two sisters, while two brothers use different terms toni and toi. It is when we turn to its linguistic side, and especially to its possessive pronouns, that we discover its very exceptional character, so exceptional that to one like myself, who had been spending several months collecting Melanesian and Polynesian systems, it produced the effect of a most startling novelty. There was not only the use of prefixes instead of the more usual suffixed pronouns of Melanesia, but there was the very striking fact that these varied with the sex of the person addressed. Usually in Melanesia when terms vary with sex, the use depends on the sex of the speaker ; a person will be addressed differently according as the speaker is a man or woman, but here it was the sex of the person addressed which made the difference, both men and women

Social Organisation 251

using one pronoun when addressing a man and another when addressing a woman.

It will have been noticed that the four islands of this region show a general agreement in the names of their clans, and even when the names differ, there is still a correspondence in function ; thus, not only may the Dhonggokama of Ysabel, the Honggokama of Florida and the Dhonggo of Savo not intermarry, but there is also associated with them the Kindapalei of Guadalcanar. A Guadalcanar man belonging to the Kindapalei would not be allowed to marry a Dhonggokama woman if he went to live on Ysabel, but would have to choose from one of the other two vinahuhu.

The clans of Florida, Guadalcanar and Savo correspond so closely that the regulation of marriage between inhabitants of the three islands presents no difficulty, but the correspondence of the three of Ysabel with the others is less simple. The Dhonggokama of this island correspond to both the Honggokama and Honggokiki of Florida ; the Vihuvunagi to the Lakwili of Guadalcanar or Savo ; and the Posomogo to the Kakau of the other islands.

Throughout the matrilineal region it is believed that the clan to which a man belongs can be told by inspection of the lines on the palm of the hand. I spent some time in the endeavour to ascertain the arrangement of the lines believed to be characteristic of each kema, but was not able to reach any definite conclusion. The only point about which I was able to satisfy myself was that the lines on the two hands might have patterns assigned to different ke^na and, when I pointed this out, it was at once said with every appearance of confidence that the right hand showed the kema of the mother while the left hand showed that of the father. I cannot, however, avoid the suspicion that this proposition was put forward as a means of escaping from the difificulty raised by my observation of the differences between the two hands.

The Western British Solomons.

In this group of islands there is a fairly uniform culture. There are no social groups corresponding to clans, and marriage is regulated entirely by kinship, no one being allowed to marry one with whom he can trace genealogical relationship. The systems of relationship in the different

252 Melmiesian Society

islands of the group show a general resemblance in form, being all of a simple kind, but the system of Vella Lavella differs greatly from the rest, not only in the nature of the terms, but also in its possessives which take the form of prefixes, Vella Lavella being, like Savo, a place where the language departs from the usual Melanesian type.

The following is the system of Eddystone Island or Mandegusu, also called Simbo and Narovo, these being properly the names of districts of the islands

Tammiggu. Father, father's brother, father's sister's husband, mother's brother, mother's sister's husband, father's father and mother's father.

Tinanggu. Mother, mother's sister, mother's brother's wife, father's sister, father's brother's wife, father's mother, mother's mother and, lastly, the wife of the elder brother certainly, and perhaps also the wife of the younger brother, in both cases man speaking.

Tunggu. Child, the reciprocal throughout of tama and tina, and thus applied to the children of both brothers and sisters by both men and women ; to all grandchildren, and, lastly, to the younger brother of the husband and possibly to all brothers of the husband.

Tuganggu. The elder brother of a man and the elder sister of a woman.

Tasinggu. The younger brother of a man and the younger sister of a woman, often used in place of lulunggu.

Lulunggu. The brother of a woman and the sister of a man.

These three terms are used also for cousins of all kinds according to sex and age.

Gamburunggu. This term was used occasionally for the sister's son (m.s.) though this relative was more generally called tunggu. According to some it might be used for children of all kinds, i.e. as the equivalent of tunggu, but whenever it was given spontaneously, it was used for the sister's son and there is no doubt that it should be applied properly to this relative and is probably an old term which is now usually replaced by the general word for child. It would seem to mark an old distinction which has now almost disappeared.

Roanggu. Parents-in-law and children-in-law, and also ' See Geographical J ourn., 191 2, xxxix, 459.

wife's sister and sister's husband (w.s.) ; also used for the elder brother of the husband and the wife of the younger brother.

Ivanggu. Wife's brother and sister's husband (m.s.) and husband's sister and brother's wife (w.s.).

A man calls his wife manggotanggu and a woman her husband marenenggu.

A doubtful point is the term for the brother's wife. According to the best opinion the wife of the elder brother is called tina, the word applied to the mother, and corresponding to this she calls the younger brothers of her husband tu. Many of the natives now call the wives of all brothers, older or younger, tinay but it is probable that this is an extension of meaning.

In this island there is a special reciprocal form in which groups of persons who stand in a given relation to one another are denoted by the terms given above preceded by tama. Thus, tamatasi is a term for brothers ; a man will say that two or more persons are tamatasi. Similarly, tamaluluna is used for a brother and sister or for a group of persons who stand in this relation to one another, though the term tamatasi is also used for this purpose. Again, tamaivana is a collective term for persons who are iva to one another and another name for this group is tamatasi roroto, meaning brothers or sisters by marriage. Other terms of this kind are tatamana used for father and children ; tamatina for mother and children ; and tamaniana for husband and wife.

The Eddystone system is remarkable for its extraordinary simplicity and for the poverty of its nomenclature. Thus, one word is used not only for all male relatives, and another word for all female relatives, of the preceding generation, but these terms are also used for those of the generation before that, i.e. the terms for father, mother, uncles and aunts are also used for the grandparents.

The system of Vella Lavella is as follows, the possessive (not given in general) being the prefix a which in some cases alters the following consonant.

Mama. Father, father's brother, father's sister's husband, and mother's sister's husband.

Niania. Mother, mother's sister, mother's brother's wife, father's sister and father's brother's wife ; also the wife of the brother, perhaps only of the elder, and the wife's sister.

Menggora. The reciprocal of the two preceding terms, used for the children of the speaker and also for the children of brothers by both men and women and for the children of the sister of a woman ; also for the husband's brother or younger brother and for the sister's husband of a woman.

Kaka (with first person possessive anggakd). Usually used for the elder brother and sister by both men and women.

Visi. Younger brother and sister, both men and women speaking.

Sanggi. A term used in some districts for the brothersister relationship ; probably the correct term for it, though those of different sex usually now address one another as anggaka or avisi according to age.

These terms are used for the children of both brothers and sisters of father and mother, i.e. for all kinds of cousin.

Papa (with first person possessive ambapa). Mother's brother.

Pakora. The sister's son of a man.

Taite (with possessive andaite). All four kinds of grandparent.

Mambuzhu. Grandchild.

Ravaja. A reciprocal term for parents- and children-inlaw.

Mani. Wife's brother, sister's husband (m.s.); husband's sister and brother's wife (w.s.).

Here, as in Eddystone, there was some doubt about the proper terms for the husband's brother and the wife's sister and their reciprocals. It is probable that the wives of both elder and younger brother are called niania while a woman calls both elder and younger brother of her husband menggora, but it may be that the use of niania should be limited to the wife of the elder brother and menggora to the younger brother of the husband. Similarly, there was some question whether the elder sister of the wife should not be niania and her younger sister menggora, but it was almost certain that a man called all the sisters of his wife niania and that reciprocally a woman called the husbands of all her sisters menggora whether they were older or younger than herself.

In structure the system of Vella Lavella is like that of Eddystone except that there are quite definite terms for the mother's brother and sister's son, so that all relatives of the preceding generation have the same name as the father or

Functions of Relatives 255

mother except the mother's brother. The Vella Lavella system also possesses terms for grandparents and grandchildren.

While the system of Vella Lavella thus resembles that of Eddystone in its general structure, it differs from it absolutely in the nature of the words used and in its possessive pronouns. Instead of the usual possessive suffix, we have a prefix so closely blended with the term of relationship that the initial consonant of the latter is often altered. This prefix is simply a and does not, as in Savo, vary with the sex of the person addressed but it is tempting to see in this prefix a simplification of the different prefixes of Savo. While in the latter island the system has become in general simple in structure but has preserved its complexity in its prefixes and in the terms for the relationship between brothers and sisters, the simplification of structure in Vella Lavella would seem to have been accompanied by a corresponding simplification of the possessive prefixes.

Functions of Relatives.

In the Malaita and San Cristoval region I was unable to obtain any account of social functions associated with different relationships but according to Dr Codrington^ the mother's brother has a particularly close relation to his nephew. It is probable that there are special duties or restrictions connected with relationships in some parts of the region, but I do not think they can be numerous or important, at any rate in Ulawa.

In Guadalcanar a man may not say the name of his mother's brother, of his father's sister or of his brother's wife. The father's sister also may not be touched but there is no similar avoidance of the other two relatives. In this island there is definite evidence of avoidance between brother and sister ; they may not say each other's names and if one is in a house, the other may not enter. If a man wishes to give anything to his sister, he must put it down and go away and the sister will come to take it. In Savo there is no prohibition on the name of the wife's mother but brother and sister may not touch one another.

1 yj/., 50, note 2.

In Ysabel a man must not say the name of his wife's mother or a woman that of her husband's father. These relatives may be in each other's presence and may talk, but must not touch one another.

I did not inquire into the matter in Florida but Dr Codrington states' that there is no difficulty about meeting, or mentioning the name of, the father- or mother-in-law or any of the wife's kindred.

In the Western British Solomons we could discover no restrictions or duties especially connected with ties of relationship, except that in Eddystone Island a man must not go too close to, and must not "swear at," his wife's mother.

Marriage.

In Ulawa and in parts of Malaita and San Cristoval marriage is regulated, so far as I could discover, by kinship. In the Lau district of Malaita there appeared to be local exogamy, two persons of the same village not being allowed to marry. I was not, however, able to collect a sufficient number of pedigrees to put the matter to a complete test.

In the matrilineal region there were or had been certain other regulations in addition to those dependent on clanexogamy. In Guadalcanar it seemed almost certain that the cross-cousin marriage is still practised, but the matter could not be tested genealogically.

In Ysabel there was some difference of opinion as to whether the cross-cousin marriage is still practised, but it was clear that there was a tradition of its presence. The information I had obtained was confirmed by the late Dr Welchman who had inquired into the point particularly and had been told that this form of marriage was allowed though cases did not often now occur. Another marriage regulation of this island is that a man may not marry a woman from whom he has received food.

In Florida I could discover neither the actual presence nor a tradition of the cross-cousin marriage, but my information was much less satisfactory than in the other two islands

M. 43.

Marriage 257

of this region. In Savo it was stated definitely that the cross-cousin marriage does not take place.

In the Western Solomons (Eddystone, Ruviana, Vella Lavella) marriage is regulated entirely by kinship, two persons between whom any genealogical relationship can be traced not being allowed to marry.

It is the existence of the cross-cousin marriage in the matrilineal part of the Solomons which gives the systems of relationship of that region their special character, for we find in them in varying degree certain correspondences which are the natural result of this form of marriage. Thus, in Guadalcanar where almost certainly the marriage is still practised, there is a common term for the mother's brother and the wife's father, another common term for the father's sister, the wife of the mother's brother and the wife's mother, and still another term which includes in its connotation both cross-cousin and brother- or sister-in-law. All these correspondences are the natural result of the cross-cousin marriage. In Ysabel where it is clear that the cross-cousin marriage has existed in the past, cross-cousins and brothers- and sisters-inlaw are distinguished in nomenclature, but it was also clear that the Bugotu term for cross-cousins,/^'^, was not in universal use, and it would seem probable that it is an innovation^ and possible even that its introduction may have been associated with the disappearance of the cross-cousin marriage. In Florida also the term for cross-cousin, mavu, would appear to be relatively modern^ and never firmly established, crosscousins being now often classed with brothers and sisters.

In Florida it may be noted that the term vunonggu is used both for parents-in-law and for the wife of the mother's brother, a feature which would again be the natural result of the cross-cousin marriage.

The social organisation of Florida, Ysabel and Guadalcanar presents such uniformity that when we find a form of marriage still present in one island, and a tradition of its presence in another, we may safely conclude that it was also once present in the third, and we thus have an excellent instance showing that terms of relationship may be survivals of a form of marriage which has existed in the past.

The systems of Fiu and Lau in Malaita also present the

1 This was Dr Welchman's opinion.

2 See Codrington, M.^ 41, note.

correspondence between mother's brother and husband of father's sister, and that of Fiu between father's sister and mother's brother's wife which would follow from the crosscousin marriage.

One of the systems of San Cristoval, that of Rafurafu, stands in a still closer relation to this form of marriage in the possession of a common term for the wife's father and the husband of the father's sister and of another common term for the wife's mother and the wife of the mother's brother.

The Bum System of Bougainville.

In the theoretical portion of this book I shall make large use of a system of relationship from the district of Buin^ in the island of Bougainville which has been recorded by Dr Thurnwald" and therefore I give an account of its main features here. 1 am greatly indebted to Dr Thurnwald for some details which were not recorded in his original paper.

M6mo, my father, son speaking ; vocative form dna.

Mdka, my mother, vocative dna. M6ka is also used for the father when the daughter is speaking.

Both mdmo and moka are used for the father's brother, and d?ta (?also m6ka) for the mother's sister. Mdmo is also used by a woman for her brother's son.

Rum. My son.

RUro. My daughter. These terms are also used for the brother's children (m.s.) and the sister's children (w.s.).

lo. Brother (in general).

Tditanu or daidanu. Elder brother (m.s.) ; also father's father, husband's elder brother, and elder sister's husband (w.s.).

Roromoru. Younger brother (m.s.); also son's son (m.s,), husband's younger brother and younger sister's husband (w.s.).

Mdmau Elder sister (w.s.) ; also father's mother, wife's elder sister and elder brother's wife (m.s.).

Rdrokei. Younger sister (w.s.) ; also wife's younger sister and younger brother's wife (m.s.).

' This district is also called Telei ; see P. W. Schmidt, Globus, 1909, xcv, 206; also Wheeler, Zcitsch. f. Kolomahprachen, igii, i, 290, where many of the terms are given in slightly different form.

2 Zeitsch.f. vergleich. Rechtswiss. 1910, XXHI, 330.

Buin Relationship 259

Nonoi. Reciprocal term for the brother-sister relationship; also father's father (w.s.) and its reciprocal the son's daughter (m.s.).

The above terms are used according to the usual classificatory rules between the children of two brothers or of two sisters. Thus, a man would call his father's brother's son if

older than himself tdita ; if younger roromoru. A man and his father's brother's daughter will call one another nonoi

and a woman will call her father's brother's daughter mamai or rdrokei according to age.

Agu. Father's sister ; also wife's mother, and husband's mother.

Mipo. Brother's son (w.s,). He is also called mdmo.

Papa. Mother's brother ; also wife's father and husband's father.

Rdguru. Sister's son (m.s).

Rdgura. Brother's daughter (w.s.) and sister's daughter (m.s.).

Bbboi. Mother's brother's son and father's sister's son (m.s.). Also wife's brother and (?) sister's husband (m.s.).

Abure. Father's sister's daughter (m.s.).

Gdmuroi. Mother's brother's son (w.s.).

Mats. Father's sister's daughter (w.s.) and reciprocally mother's brother's daughter. Also husband's sister and brother's wife (w.s.).

Ndna. Mother's father.

TSte. General term for grandfather ; also mother's mother.

Ruge. Son's child (w.s.) and daughter's child (both man and woman speaking).

Iru or moru. Husband.

Ina. Wife.

This system has several remarkable features. The nomenclature for the mother's brother and the father's sister is clearly such as would arise through the cross-cousin marriage, and this form of marriage will also explain the application of the terms for cross-cousins to brothers- and sisters-in-law. So far as these terms are concerned the exceptional feature of the system lies in its richness of nomenclature. There are no less than four terms for cross-cousin ; boboi used between two

17 — 2

26o Melanesian Society

males, mats used between two females and abure and g^muroi used between those of different sex. In Dr Thurnwald's notes it is only stated that dbure is used of the father's sister's daughter (m.s.) and g^muroi of the mother's brother's son (w.s.) and no terms are given for the father's sister's son (w.s.) and the mother's brother's daughter (m.s.). It is probable, however, that they would be gdmuroi and dbure respectively, in which case g^muroi would be a term for a male crosscousin used by a woman and dbure a term for a female cross- cousin used by a man. If this is so, the terms would fall into line with the customary mode of nomenclature for brothersand sisters-in-law. There would be one term used between two men and another between two women while those of different sex employ two terms. In this system the terms used between cross-cousins of the same sex are also used between relatives by marriage of the same sex.

Another exceptional feature of the Buin system is the presence of special terms connoting age used between sisters and both these and the other terms for brother or sister are also used of the parents of the father. A man calls his father's father tditanu, the term he uses otherwise for his elder brother, and a woman calls her father's father nonoi, used also for her brother. The father's mother, on the other hand, is called indjnai both by males and females, this term being otherwise used by a woman of her elder sister, so that in this case both men and women address a grandparent by a word which is used only by a woman when applied to a sister.

Corresponding to these usages, a man calls his son's son by the term he otherwise uses for his younger brother and his son's daughter by the term he uses otherwise for his sister, but a woman does not, as might be expected, class her son's children with brothers or younger sisters, but calls them rtlge, the term for other categories of grandchild.

The terms tditanu, roro77toru, mdmai and rdrokei used primarily between brothers or between sisters are also used for a number of relatives by marriage of different sex, a special feature of interest being that these terms, which are only addressed to those of the same sex when used between brothers or between sisters, are used between those of different sex when they apply to relatives by marriage. Thus, tditanu which is primarily a term applied by a man to his elder

Buin Relationship 261

brother is used by a woman of her husband's elder brother and her elder sister's husband while mamai which is primarily a term applied by a woman to her elder sister is used by males when it denotes the wife's sister or the brother's wife. There seems to be a definite rule that terms used between those of the same sex when they denote brothers or sisters are used between those of different sex when they denote brothers or sisters by marriage.

CHAPTER XI

FIJI

A NUMBER of systems of relationship were obtained from several places both on the coast and in the interior of Viti Levu. The only Fijian systems hitherto recorded^ are of the classificatory kind, comparatively simple in character, their special feature being the clear dependence of many of the terms on the cross-cousin marriage which is known to be practised in the island. Among several tribes in the mountainous region in the north-eastern part of Viti Levu I found systems fundamentally different from those previously recorded, possessing some very unusual features. I found also that the systems of the coastal peoples had many modifications in different parts of the island, and one of the most interesting features of the work to be now recorded is the demonstration of the great variations which may exist in the systems of relationship used by people who live close to one another and differ little in culture and in general racial characters.

Before considering these systems a brief account may be given of the general features of the social organisation. That of the coast and especially of the Mbau and Rewa districts has already been often described. It has reached an advanced stage of development, a process of welding together of many once separate elements having taken place, probably to a great extent in recent times and largely as the direct or indirect result of European influence.

The organisation of the tribes of the interior has been less fully described, and I spent much of my time in an endeavour

' Morgan, Systems of Consanguinity and Affinity of the Human Family y Smithsonian Contributions to Knowledge, Vol. xvii, p. 573; and B. Thomson, The Fijiansy London, 1908, p. 182.

Social Organisation

to discover its features, but with very imperfect success owing to the short time at my disposal, and the following description

must only be taken as a tentative and certainly very incomplete account.

The people of the mountainous region of the interior are

divided into a number of small independent bodies which have the same names as the districts they inhabit\ I obtained the systems of three of these, the Nandrau (also called Navuta), the Navatusila and the Dhawanisa while something was learnt about the organisation of some others.

Each of these bodies of people is divided into a number of groups called matanggali and in some cases these matanggali are again divided into smaller groups called itokatoka^. Each primary body has a chief or chiefs and its members are bound together by the common possession of a more or less sacred object, usually an animal, often a bird, which its members may not eat. These primary bodies are to a large extent endogamous, marriages usually taking place between members of the same group, though marriages with persons of other groups are not excluded and certainly occur occasionally at the present time. Further, each of these bodies seems to have its own dialect, while in the past warfare between them was more or less habitual, and these primary bodies may thus be regarded as tribes and in the remainder of this article they will be so called.

The primary subdivisions called matanggali 2iXt. also largely endogamous at the present time, whatever they may have been in the past. Sometimes a matanggali has a sacred object, usually an animal which its members may not eat, though free to the rest of the tribe. In several cases I obtained lists of the matanggali of the tribes together with any functions or customs peculiar to them, and I give a few examples in the following pages. Though doubtless incorrect in detail, they may illustrate the general nature of the social organisation.

The men of a matanggali slept in the same mbure or men's house and each 7natanggali had its own mbure. Usually there was a belief in the descent of the members from a common ancestor, thus the members of the chief matanggali of the Nandrau, called the Talandrau, trace their descent to a man of light colour of that name who came to Viti Leva from another place. In some cases the matanggali of a tribe

' According to de Marzan {Anthropos, 1907, ii, 400) these tribes often take their names from trees, less frequently from animals. The use of these names for the districts is probably recent.

^ I am doubtful whether the inland tribes possess these subdivisions ; they certainly occur in some of the people of the coast.

had different functions ; there was not only the definite and apparently universal distinction between the matanggali of chiefs and ordinary people, but the matanggali of the ordinary people were sometimes differentiated by the possession of special occupations.

Systems of Relationship.

The systems of relationship were obtained almost entirely by means of the genealogical method, though occasionally, where time was short and I had to do with an intelligent informant, I merely collected the Fijian equivalents for the various relationships I wanted, but I only did this after I had become familiar with the general features of the systems with which I had to deal. I met with no special difficulties in the collection of the pedigrees. The old taboo on the names of brother and sister had either entirely disappeared or had become so little effective as to present no obstacle, and there did not seem to be any source of confusion arising from adoption or from customs connected with names. My chief diffi- culty arose in the collection of the terms themselves owing to the spread of the Mbau dialect of the Fijian language throughout the population. It seemed clear that, at any rate in the interior, each tribe had its own dialect, these, so far as I could judge, differing considerably in vocabulary from one another, but the dialect of Mbau is now coming into wide use, and often I had much difficulty in ascertaining the proper terms of the tribe with which I was dealing. The conditions reminded me very closely of my work with the Todas^ in India, where I had to be continually on the watch to see that I got the real Toda words and not those of the Badaga language. In the case of the Fijian tribes the discovery of the terms proper to the tribe was of essential importance, for the system of Mbau being of a much simpler nature than those of the interior, many of the terms of the latter had no Mbau equivalents, and the use of the Mbau dialect would have obscured, if it had not altogether hidden, many of the special features of the mountain systems. Indeed, in my first attempts to record these systems I completely missed some of the chief features owing to my failure to obtain the terms actually employed by the natives in their natural intercourse

• The Todas, 1906, p. 15.

with one another, i.e. they gave me the terms they were accustomed to use in talking to Europeans.

The pedigrees obtained from the people were fairly full when one considers how they were collected. They were obtained in very few days and I have no reason to suppose that my informants were special adepts in genealogical lore. The specimens collected, of which one example is recorded on p. 274, show that the Fijians, at any rate in the mountains, preserve their pedigrees as fully as other Melanesians. In one feature the pedigrees were exceptional. I was not only given all the descendants of a grandfather or great-grandfather, but in some cases my informants were able to give me the names of their ancestors in the direct line for eight or nine generations, though they did not know the collateral lines so far back. Their pedigrees were thus a combination of the usual Melanesian pedigree which a man knows only for three or four generations but in all collateral lines, and the other kind of pedigree in which descent is traced back directly to some very remote ancestor.

I must also mention that the pedigrees were isolated examples. My only guarantee of their accuracy is that the terms of relationship obtained by their means were consistent.

In the record which follows I give the terms with their possessives in the first person. This is necessary, not merely because the terms are always used in this form, but because there are two sets of possessives used in the island of Viti Levu, one in the form of inseparable suffixes and the other in the form of possessive nouns* which precede the term of relationship, and we shall see later that the kind of possessive may afford evidence as to the nature and provenance of the different terms.

The Coastal Systems.

The following were given to me as the terms of the Mbau system : —

Tamanggu. Father, father's brother and husband of mother's sister, used also for more distant relatives of these classes according to the ordinary rules of the classificatory

• By " possessive noun " I mean a word such as nonggu, i.e. no-nggu, in which the true possessive pronoun is suffixed to a particle which is now generally held to represent a noun.

system. An elder brother of the father is tamanggu levu or great father, and a younger tamanggu lailai or little father.

Tinanggu and nau. Mother, mother's sister and wife of father's brother, nau being used only as a term of address.

Luvenggu. Son and daughter, and generally as the reciprocal of tama and tina.

Tuakanggu. Elder brother when a man is speaking, and elder sister when a woman is speaking.

Tadhinggu. Younger brother of a man and younger sister of a woman.

Nganenggu. Brother (w.s.), and sister (m.s.).

The three preceding terms are also used for the children of all those who are called tama and tina, according to relative age and sex.

Ngandinanggu, ngandi, momo and vungonggu. All these were given as terms for the mother's brother and others whom the mother would call ngane, and also for the husband of the father's sister. Momo is only used in direct address and vungonggu only when speaking of these relatives. Ngandinanggu and its contraction ngandi are also used for the father of husband or wife while vungonggu, an alternative term for these relatives, is used reciprocally between a man and the consort of his daughter or son.

Nganeitamanggu or nganei. Father's sister and wife of mother's brother ; also the mother of husband or wife. These relatives may also be called vungonggtc and this term is used reciprocally between a woman and her brother's child, her husband's sister's child and the consort of her daughter or son.

Koya na ngonia\ Sister's son, man speaking. The sister's son is also called vasu and is spoken of in conversation as nonggu vasu, but is always addressed either as koya na ngoma or by name.

Tukanggu. Father's father and mother's father.

Mbunggu. Father's mother and mother's mother.

Makumbunggu. Grandson and granddaughter, whether child of son or daughter.

Tavalen^gu. Son of mother's brother or father's sister, when a man is speaking ; also wife's brother and sister's husband (m.s.).

Ndavolanggu. Child of father's sister or mother's brother

^ Koya is probably the pronoun "he." If so, I know of no reason why it should have been given with this term of relationship.

whose sex is different from that of the speaker ; also the wife's sister and brother's wife (m.s.); the husband's brother and the sister's husband (w.s.).

Ndauvenggu. Daughter of father's sister or mother's brother (w.s.); also husband's sister and brother's wife (w.s.).

Wahnggu. Husband or wife. There was some question whether the relatives who call one another ndavola might not also address and speak of one another as wati.

Vuarumbi. Great-grandfather.

It is a Mbauan custom to speak of a man as the father of his son, thus Sailosi, the father of Siva, may be called tamai Siva. It did not appear that this practice was habitual, but was followed by anyone who wished to pay a compliment to the father.

Groups of persons standing in a given relation to one another are denoted by the term of relationship with the reciprocal particle prefixed, as in veinganeni, veindavolani. Thus, the term veivungoni is used to denote the mother's brother, father's sister, mother's brother's wife, father's sister's husband, sister's children of a man and brother's children of a woman, parents-in-law and children-in-law.

Variants of the Mbau system were obtained in other villages of the coast. The system just recorded was obtained in the village of Mata Levu on the Wainimbokasi, a branch of the Rewa. In the village of Tokatoka, not more than two miles distant, the term for the father's father is tumbunggu instead of tukanggu. The mother's father is called tukanggu ; the mother is nggei and it was said that the father is nau, a term applied elsewhere to the mother. Similarly, in the village of Nausori, higher up the Rewa than these villages, there were several modifications. Here the father's father is tu^nbunggu, the father's mother na ndamanggu, the mother's father tukanggu, and the mother's mother mbunggu. The mother's brother is called mo7no and my informant, a chief of the district, was quite confident that ngandinanggu was not used in Nausori for this relative. The sister's son is nonggu vatuvu. A man calls the daughter of his mother's brother or father's sister watinggu, used also for the wife, and this term is reciprocated by the woman. The wife's sister and the husband's brother and their reciprocals are also called watinggu. A woman, on the other hand, calls the daughter

of the mother's brother or father's sister raivanggu^^ this being the equivalent of the Mbauans' ndauvenggu. The husbands of two sisters or the wives of two brothers call one another karuanggu.

Here again people are spoken of as the parents of their children, thus my informant who had a son called Philemon would be called tamai Philemon, while his wife was tinai Philemon.

Though the proper term for the brother-sister relationship is ngane, a woman is often called by a man tuaka or tadhi according to her age.

A distinct variant of the Mbau system was obtained (only in outline and not by the genealogical method) from a member of the Nokanoka tribe which lives on the north-east coast of the island east of the Tavua river. The terms used in this system are characterised by elision of the letter /, in some but not in all cases ; thus, tama becomes ama, tina is ina and wati, wdi.

The chief features in which the system differs from that of Mbau are as follows : — the mother's brother and husband of father's sister are called ngandinanggu but not vunonggu, and similarly the father's sister and wife of the mother's brother are only nganeitamanggu^ or nganei ; the sister's son (m.s.) is natunggu^ and his wife vunonggu. A sister would be spoken of in conversation as nganenggu but would be addressed as tuakanggu or tadhinggu'^ according to age. All four grandparents are unibunggu and grandchildren vakumbunggu. The child of mother's brother or father's sister is either karua or kdamanggu, a corruption of katamanggu, and the term avalenggu, used in Mbau for the cross-cousin, is limited to relatives by marriage, but is used for the sister of the wife as well as for her brother. The husbands of two sisters are also avalenggu while the wives of two brothers are ndauvenggu.

Still another variant was obtained from a member of the Narambula tribe which lives on the north coast of the island. Here the term for father is manggu ; the mother is nau, the mother's brother ngandi and the father's sister nganei. A man calls the children of his sister vunonggu and applies the same term to their husbands or wives. The grandparents are tumbunggu and the grandchildren makumbunggu. The sons of

• Ra-ivanggu.

^ In these words the / is retained.

the mother's brother or father's sister are karuanggu and their daughters ivatinggu. The term tavalenggu is Hmited, as among the Nokanoka, to relatives by marriage, and is applied by a man to the brother's wife, the sister's husband and the wife's brother, while the wife's sister is watinggu. The parents of husband or wife are vunonggu, a term also applied reciprocally to the son's wife and daughter's husband. The husbands of two sisters are karuanggu.

The Mbau system I have given agrees closely with that recorded by Thomson \ The differences are that he gives vungo as the term for sister's son as well as for son-in-law and gives only ngandina for the mother's brother. He also gives tu7nbu as a term for great-grandparent. My Nausori system agrees even more closely with the system drawn up by Fison in 1869 and recorded by Morgan^ The only important difference is that whereas I give wati as the term for the wife's sister and husband's brother and their reciprocals and also for cross-cousins of different sex, Fison gives^ for the latter the same term as in the Mbau system and nonggu ndaku for the relatives by marriage, but adds that some natives gave ivatinggu. It is a striking fact that this should be the only difference between two accounts obtained at an interval of nearly forty years. The agreement gives a fair assurance that the inland systems which I recorded with still more care are also correct, at any rate in their main features. Before passing on to these inland systems the general features of those found on the coast may be briefly considered.

The Mbau and Nausori systems are clearly based on the cross-cousin marriage. In the Mbau system the classing together of the mother's brother and the father of the consort, of the father's sister and the mother of the consort, of the cross-cousin and the wife's sister and husband's brother and their reciprocals is so definite as to afford one of the best existing examples of the absolute dependence of terms of relationship on forms of marriage. The Nausori system goes one step further and gives the cross-cousin of different sex the same designation as is used for the husband or wife.

The Mbau system is exceptional in having three separate terms for three categories of cross-cousin and brother-in-law

• Loc. cit.

• Loc. cit.

3 Op. cit. p. 581.

and sister-in-law; one used between men, another between women and the third between those of different sex. One interest of the varying systems is the difference in the terms for the wife's sister and husband's brother. We have a transition from the Mbau system in which these relatives are classed with the cross-cousin to the Nausori system in which they receive the same designation as the actual wife.

A further feature of great interest is the multiplicity of terms for mother's brother in the Mbau system. One of these, momo, we shall meet again among the inland tribes and it is probably an ancient term which still remains in use in direct address. Another term, vungo, is clearly related to the position of the mother's brother as father-in-law through the cross-cousin marriage. A third term, ngandi, is an abbreviation of ngandinanggu which is clearly a descriptive term, being the ngane of the tina.

A term even more clearly descriptive is that for the father's sister, viz. nganeitanianggu, the ngane of the tama. It is unusual to find descriptive terms in Melanesia. They are well established in Samoa (see Chap. XII I) and considering the old-standing relations between Samoa and Fiji, it is very tempting to regard these terms as due to Samoan influence and therefore of relatively recent origin, but this is a question which I shall deal with more fully later. One point may be mentioned here to which I shall also return later. While nearly all the terms of relationship suffix their possessives, there are certain exceptions, viz. vasu or vatuvu and ndaku^. We shall see that both of these terms are found among the inland tribes.

The Inland Tribes.

I obtained three systems from tribes of the mountainous interior, viz, those of Nandrau^ and of the Navatusila and Dhawanisa, all of which possess the same general characters though differing in details of structure and in many of the terms.

The system obtained most fully was that of the Nandrau people, occupying the two villages of Nandrau and Ndumbua.

^ This word also means "back" and then suffixes its possessive, as in ndakunggUy my back.

2 Properly Nandrau is the name of the chief village occupied by a tribe called the Navuta, but my informants nearly always spoke of the Nandrau people and I follow their example.

272 Me lane Stan Society

This tribe has a large number of matanggali which are of three different ranks; the matanggali turanga or those of the chiefs; the ^natanggali dhauravou or those of the ordinary people, the name dhauravou meaning " young people " ; and one matanggali vanua supplying the 7nata ni vanua, the heralds and intermediaries between the chiefs and the ordinary people.

There are four 7natanggali turanga called respectively Talandrau, Naroko, Nandrongo and Vuanisalusalu, these being the names of four brothers from whom the members trace their descent. These were men of light colour who came to the island from some other place, landing in Viti Leva Bay.

The matanggali of dhauravou rank are seven in number: the Natatea, Naurodha, Nakoro, Nabarimai, Naindundu, Wailevu, and Nasei or Vangganasei. Several of these had special functions; thus, the Naurodha made gates in time of war; the Nakoro, Naindundu and Nasei planted food; the Wailevu also planted food and had in addition the special privilege of eating the pig's head, usually the perquisite of the warriors; the Nabarimai furnished the priests. The ov\y matanggali va7uia is called Nakorowaiwai and, in addition to supplying the assistants of the chiefs, its members planted food.

A group of people associated with Nandrau and said by some to be another matanggali of that tribe is formed by the Nasalia^ which again had two divisions ; one called Nambouvesi, and the other Dhauravou, the servants of the Nambouvesi. Still another people associated with the Nandrau are the Nambilia, including the Toma ni ivi and Kaivudhi 7natanggali. The Nasalia and Nambilia are probably distinct tribes which for some reason have become more or less closely fused with the Nandrau.

I did not obtain the names of any itokatoka of these matanggali and am not sure that any such subdivisions of the tribe exist.

The tambu or sacred animal of the whole of the Nandrau tribe is the nggiliyango, called ngginggi in the Mbau dialect, a small black bird with a long beak which was said to feed on chillies. The Nandrau people believe that they are descended from the nggiliyafigo bird who told them not to "steal women"

• This place is shown in the map at a considerable distance from Nandrau.

The Nandrau 2']'7,

on pain of death, i.e. it prohibited adultery. Some of the matanggali may not eat other animals free to the rest of the tribe ; thus, the Wailevu may eat neither the dog nor a fish called ndabea and the Kaivundi, a division of the Nambilia, may not eat the snake. The Dhauravou division of the Nasalia were not allowed to eat human flesh but, when they killed a man, had to take the body to the Nambouvesi. Since the latter was the division of the chiefs, it is possible that this is a restriction of a different kind; that the Dhauravou were merely forbidden to eat man because his flesh was the perquisite of the chiefs.

In addition, several divisions of Nandrau were subject to restrictions on the use of certain plants as food. Thus, the Nasalia people might not eat the via or giant caladium, nor could they eat the soanga or wild banana^ The Nandrau people also as a whole might not eat yams during the two months called Uluvatu and Vunangumu corresponding approximately to January and February, i.e. beginning with the new moon In January, but it was said that this practice used to be common to all the Fijian people.

In this and other tribes of the interior a man respects the animal of his mother as well as that of his father. The way in which the matter was put to me was that a man will not eat the tambu animal of a tribe to which he is vasu (see p. 292).

I was told an interesting instance of the evil effects believed to follow non-observance of these restrictions. A man whose people had always eaten snakes married a woman to whom these animals were tambu. He did not give snakes to his children but their food was cooked in pots which had been used to cook snakes and the constant illness of his children was ascribed to this cause and, in order to escape from continual trouble, he left his wife and took another woman.

The system of Nandrau was investigated by means of four fairly large pedigrees obtained from independent witnesses and from some of these the terms of relationship were worked out very completely. The fullest of these is given on the next page and is used in what follows to illustrate some of the more complex ways of using the Nandrau terms. An important feature of this as of the other inland systems is that with some terms possessive pronouns are used in the same way as among

^ Also called the Mountain banana. It has great fruit bunches and seeded fruits (see Guppy, Observations of a Naturalist in the Pacific, Vol. 11, p. 414).

R. M. S. l8

Melanesian Society

rt

^

E

^

E

'5

■5

II

<

Q

II-

< -J

O

— H

<

<

iz;

f^

^

g II-

& K

4. Q

i ^

O 2;

rt

2; XI

rt

c/0

J O

S II -

<

II — z < u

o ^ o

<

<n

i-W "

tL,

<

S',

■— >

-<

<

M

H

Ol

<

m

_i

cu

/S

w

c/}

■s

w

1— >

-J

;n

w

Vh

w

-S c

3 3

o w

■J

1 -

I! — >

c

^

<

0 .

II-

[x]

bd

c«

<

%

" o

■>TJ

H

W

•A

c _ c

l-J

Relationship in Nandraii 275

the coastal people while other terms take possessive nouns. In the following list those terms in which the possessive is suffixed are given in the form which has been used in the coastal systems, viz. with -nggu, the first person. When no suffix is given, it means that a possessive noun is used, that of the first person being nonggu. The following are the terms with the senses in which they are used.

Kam {kamu or ikamu when used in direct address). This is primarily a term for father and it is also used according to the ordinary principles of the classificatory system for the father's brother and for others of his generation whom the father would have called brother. Ordinarily, however, this term when used for the elder brother of the father is qualified by adding levu, meaning great, while for the younger brother there is a different term, tadkita, to be referred to again below. The husband of the mother's sister is also called kamu. In addition to the ordinary extension of the term for father, the word kam is also applied to the grandfather (father's father) of the father and apparently again it would be applied to the father's father of this man if his name had been preserved. Thus, in the pedigree Simoni called Waimboidha kam, and Josephata, his son, would give this name to Tilatila. Ancestors of alternate generations were thus given the same appellation as the father. The term kam has a point of interest in that it is unlike an ordinary Fijian word and does not end in a vowel. I was clearly told that the ordinary term for the father is kam. and that the form ikamu was especially used in address.

Tadkita. This has already been mentioned as a term for the younger brothers of the father and is equivalent to the iamanggu lailai of the Mbau system. The word would seem to be a compound of tadhi, younger brother, and ta, the first syllable of tama.

Nau. This is primarily the term for mother and, as is usual in the classificatory system, it is also used for the sister of the mother and for the wife of the father's brother. It is also occasionally applied to the wife of the mother's brother. In addition to these applications general in the classificatory system, it is also used in certain exceptional ways. The wives of the son and of the brother's son (m.s.) are called nau and this term is also applied to the wife of the sister's son (m.s.). As an example, Ndelana called Kelera nau and

18—2

Simoni gave this name to Eta and Latileta, the wives of Ovini and Apele (Abel) and to Vadhisevu, the wife of Avisalomi (Absalom), Thus, women of the generation younger than the speaker are called by the same name as that applied to the mother, another case of the same term being applied to persons of alternate generations.

Luve7iggu is applied to the children independently of sex, and also to the child of a brother by both men and women. It is also applied by a woman and her husband to the child of the woman's sister and also apparently by a woman to the son of the husband's sister, these, with the exception of the last, being according to the customary usage of the classificatory system. It would probably also be applied to the greatgrandson, but naturally opportunities for this would not often occur.

Ttiiua, tuakanggu and tadhinggu. Tutua is probably the proper Nandrau term for elder brother but the Mbau term tuakanggu is often used. Tadhinggu is the only term I heard for younger brother. These terms are also applied by brother to brother, by sister to sister, and by brother to sister and vice versa. When these terms are applied by a man to a woman the prefix ra is sometimes used, thus, an elder sister is called ratuakanggu and a younger sister ratadhinggu.

Nganenggu is not used in direct address but is apparently used when a man speaks of his sister or a woman of her brother. The distinction in the use of terms for brother and sister according to the sex of the speaker thus appears to be absent when the terms are used in address, and I received the impression that it was in process of disappearance altogether.

The various terms for brother and sister are applied to the children of the father's brother and of the mother's sister according to the orthodox usage of the classificatory system. In addition, however, they, and especially tadhinggu, are employed in a very exceptional manner which falls into line with the peculiar use of kam and nau already described. A man calls his son's son tadhinggu and the father's father is sometimes called tutua or tuakanggu, though much more commonly this relative is tumbunggu. In the pedigree Ndelana called Josephata tadhinggu and Simoni gave this name to Eseroma, his brother's grandson. According to one informant the son of the sister's son would be called tadhinggu,

Relationship in Nandrau 277

Simoni giving this name to Vanivasa^, and reciprocally the father's mother's brother is tutua. It was also said, however, that the son of the sister's son would be tavale. It is possible that there are different usages in some points among the Nandrau people for there is little doubt that this group has had a complex origin. The use of tadhinggu for the son's son is constant and was confirmed over and over again, and often the reciprocal term, tutua, is applied to the grandfather. We have thus another example of a special relation between members of alternate generations, grandfather and grandson applying to one another the same terms as those used between brothers.

Momo. Mother's brother, not used otherwise than is customary in the classificatory system.

Saivai. Wife of the mother's brother ; it seemed that this term is often replaced by 7iau.

Tukai. Father's sister and, like momo, only used as is customary in the classificatory system for those whom the father would call sister.

Vasu. The sister's son of a man, with the customary extension, Savenadha being the vastt of Simoni as well as Ovini and Apele. As already mentioned, the wife of a vasu is called nau. A man also calls his sister's daughter vasu but she may be distinguished as alewa vasu.

Tumbunggu. The term now usually applied to the father's father, though it is probable that this relative should properly be called tutua, a term which was, as a matter of fact, often employed. These terms are also used for the brothers of the father's father.

Tatai. Father's mother, used in general for the wife of one called tumbu.

Taitai. The mother's father and others whom the mother would call tama.

Mbunggu. The mother's mother and in general the wife of a taitai.

Ndiva. Used by a woman for her son's son; she would call him nonggu ndiva.

Tangi. Applied by both man and woman to the child of the daughter; the child would be called nonggu itangi. While tadhinggu is thus the term reciprocal to tutua or tumbunggu, ndiva is the reciprocal of tatai, and tangi is reciprocal to both

1 This word is the Fijian form of Barnabas.

taitai and bunggu, although once or twice makumbunggu was used as the reciprocal of bunggu.

Tavale\ This term is applied to one another by the children of brother and sister. Thus, it is used by both men and women for the children of the mother's brother and of the father's sister. It is also used for various relationships by marriage otherwise called veidhakavi and veilavi. No distinction is made according to sex, the term thus mcludmg those who would be distinguished as tavale, ndavola and ndauve in the Mbau system. According to one informant the child of a vasu would also be tavale.

Mani. A woman would call her husband nonggu mani or sometimes naitaukenggu, "my master."

Mangua. A wife is called ''nonggu alewa mangua'^y It seemed clear that 77iangua is not used for the sister of the wife.

Vunonggu. A term used by a man of his wife's father and mother, and reciprocally they apply this term to him. The son's wife, on the other hand, is called nau. A woman and her husband's parents are also vuno to one another. The husband of the father's sister was said to be called vunonggu, but I am not confident of the correctness of this.

Ndaku. A term used with the possessive noun by a man for his brother's wife and by a woman for her husband's brother. It is thus a reciprocal term used between men and women.

Veidhakavi, veilavi and tavale. These are terms for marriage relationships, the exact meaning of which I failed to discover. Veidhakavi is applied by a man to his wife's brother and reciprocally to the sister's husband, but it was not used in other cases when, so far as I could see, the relationship was the same. Again, the term veilavi was also used for this relationship, but it was used as well for a relationship between man and woman, i.e. for the wife's sister and for the husband of the sister (w.s.). In one case veilavi was used as a term reciprocal to veidhakavi, thus, Simoni called Saranggia veilavi and was called by him in return veidhakavi. In this and other cases the proper application of the terms was said to depend on the places to which the people belonged or rather the places to which their mothers had belonged, thus, Saranggia was said to have been called veilavi by Simoni because his mother had come from another place so that he

' I am not certain of the form of possessive taken by this word. ^ Mangua = t)\di.

Relationship in Nandrau Trjc)

was not vasu to Nandrau \ I regret very much that it was impossible to clear up the exact meaning of these terms for, if I had been able to do so, I believe that it would have contributed much to the understanding of the marriage regulations of the Nandrau and other mountain people which are at present veiled in much obscurity. Whatever the meaning of these terms, it seemed that they were being gradually replaced by the term tavale.

Raivanggu. Applied to one another by women, being used for a brother's wife and reciprocally for a husband's sister. It has the same meaning as the Mbau ndaiivenggu and, so far as marriage relationship is concerned, is used in the same sense.

Kawa. A term meaning descendant, used by a man for anyone descended from himself though I was only given the term when the separation was great, as in the case of Ndelana and Vanivasa.

Vakuvu. Whenever a man is vasu to a matanggali, he calls all the members of that matanggali vakuvu.

Karu. Applied to one another by the two wives of a man, thus, Vunimbeli and Nawati would call one another karu.

The designation given to a son or daughter is altered if he or she become a father or mother, thus, a son who marries and has a child is called luverau or " child of us two " by both his father and mother, this being evidently connected with the custom, found among the Nandrau people, as generally in Fiji, of speaking of people as the parents of their children. Here, however, a man seems to be called the son of his own child, Simoni being called luvena ko Eta or luvena ko Josephata. Similarly, a daughter after having a child is called by her father nana ko nonggu tangi while the mother would call her either this name or nana ko na ntakumbunggu.

I investigated less completely, but in each case by the genealogical method, the systems of relationship of two tribes or communities of a kind resembling tribes, the Navatusila and the Dhawanisa, both inhabiting tracts of country in the neighbourhood of Nandrau. Their systems had the same general characters as that of Nandrau, though in many cases the actual terms differed, these being merely part of the differences in dialect of the three peoples.

^ Since the majority of marriages take place within the tribe, a Nandrau man is usually also vasu to Nandrau.

28o Melanesimi Society

The Navatusila.

This tribe occupies a district not far from Nandrau
The missionary Baker was killed by these people who were among those who practised the Nanga cult. My information was obtained in only one interview from an old man who had just been discharged from the prison at Nandarivatu and there are many omissions and doubtless more inaccuracies. I should not have ventured to publish information so gained if it had not been that the system shows a very close correspondence with that of Nandrau in its general structure. I need hardly say that it would have been quite impossible to obtain even this information without the use of the genealogical method.

My account of the matanggali is very fragmentary. The sacred animal of the whole Navatusila people is the nganivatu, a fish-hawk, which may be eaten by none of the people of the tribe. The chief matanggali is called Tuivandrasinga, the members of which may not eat the dhongge, a pigeon. Other matanggali are the Ivisi, Nanoko and Yasawa"^ who may not eat dog; the Narembawho may not eat the bird called remba; the Namarama who may not eat the fowl ; and the Naviaraki and Tambaivunanggumu who have no forbidden animal except that common to the whole tribe (one account gave also the Talenakulu and the Nalotawa as other matanggali). The Navatusila people as a whole believe that they are descended from the nganivatu and each of those matanggali which have a sacred animal also trace their descent from that animal.

Here, as in the case of the Nandrau, I could not find that there are any subdivisions of the matanggali corresponding to the itokatoka of some tribes.

The Navatusila term for father, father's brother, etc., is momo, the term used in Nandrau and by other peoples of Viti Levu for the maternal uncle. This transference of term from one relationship to another seems very strange but is explained by the fact that mo7no is a common word for chief and is applied by chiefs to their fathers and thus seems to have come into general use in this sense among

' In the map people of this name are given in two situations. I do not know how closely they are associated with one another. ''■ This is the name of a group of islands.

The Navahisila 281

some tribes. The word for mother is nene and it is used in the same peculiar senses as the nau of Nandrau. The terms for child, brother and sister are the same as among the Nandrau people, but perhaps tutua was used more constantly for elder brother than among the latter. Tadhinggu is used for the son's son as well as for the younger brother exactly as in Nandrau. The term for mother's brother is koko and that for the mother's brother's wife tamdani, though as in Nandrau ne7te was sometimes used. The father's sister is ngwanita, an abbreviation of ngwaneitamanggu, and the sister's children are vasu but I did not ascertain the terms for the wife or son of the vasu.

The father's father is called tutua, the same word as that applied to the elder brother, and as already mentioned the son's son is correspondingly tadhinggu. The mother of the father and also the mother of the mother are called tatai and each calls her grandson viangonggu. The sons of the father's sister and of the mother's brother are both called tavale as in Nandrau and the same term is applied to the wife's brother and sister and to the sister's husband (m.s.). The proper term, however, for the reciprocal relationship of wife's brother and sister's husband (m.s.) is vikila, this corresponding to the veidhakavi or veilavi of Nandrau or to both of these terms. The husband's brother and the brother's wife (m.s.) call one another nonggu vitambui, this corresponding to the nonggu ndaku of Nandrau. The wife's father and mother are called vunonggu and this term is also applied by a woman to the son of her brother.

The Dhawanisa.

I obtained the names of the following matanggali of this tribe' : — the Leiekumba, Nakorosovivi, Votualevu, Nambetongondrongondro, Nambelanga, Nandala, Numbu and Keti^ The whole tribe traces its descent from a small aquatic animal of some kind called the ndravindravi which they may not eat but I could not discover that the different matanggali had any animals peculiar to themselves. I was also told that none of the Dhawanisa would eat the soanga or wild banana

1 I do not know the exact situation of this people, but they probably occupy a district north-west of Nandrau.

2 These names were usually given the prefix kai., as Kainumbu, and Kaiketi, kai meaning apparently " people of"

when it was not fully ripe but it was very doubtful whether this was on account of any sacred character.

The terms of relationship were obtained fairly completely by means of two extensive pedigrees but it was not possible to brinof out the full connotation of several of the terms.

The father is called momo by the chiefs but vava by the ordinary people, this usage thus helping to explain the peculiar transference of the term for the uncle to the father which is found among the Navatusila. The younger brother of the father is tadha corresponding to the tadhita of Nandrau. The term for mother is nene or nenei. The names for child, brother and sister are as among the Nandrau and Navatusila peoples, and as among the latter tutua is evidently the proper term for the elder relationship. Tadhin^gu is used for the son's son and also for the son of a vasu as in Nandrau. Momo is used for mother's brother as well as for the father of a chief, the former being sometimes distinguished as momo ngwandi while the wife of the mother's brother is nenei. The father's sister is ngwanita, an abbreviation of ngwaneitamanggu, and her husband is dhawai. The sister's child is vasu and, as already mentioned, the son of a vasu is tadhinggu. The wife of a male vasu and the husband of a female vasu are both dhawaiy the same term as is applied to the husband of the father's sister so that we have here another example of a term being applied to the members of alternate generations.

About the terms for the different grandparents there was some difference of opinion between my informants, which was said to be due to one of them giving the words of the closely allied Nggaliyalatini dialect instead of that of the Dhawanisa. The father's father is either tumbu7tggu or tutua ; the father's mother either nggu tai^ or tatai ; the mother's father also nggu tai or tatai\ and the mother's mother tatai according to one and mbunggu according to another of my informants. Except in the case of a man speaking of his son's son, a grandson is called viangonggu. It thus seemed as if there were present a tendency to call grandparents tatai (or the corresponding term nggu tai) and grandchildren viangonggu, except in the relationship between a man and his son's son, and it would seem as if we had here a case of the disappearance of distinctions which are perfectly definite in the other two mountain systems.

• Nggu = my.

The Dhawanisa 283

The child of the mother's brother and of the father's sister is called tavale and this term is also used for the reciprocal relationship of sister's husband (m.s.) and wife's brother and it is also applied to the wife's sister. The brother's wife (m.s.) and husband's brother call one another /anggu vitambui, as among the Navatusila, but with another form of possessive noun. A man calls his wife langgu alewa and it was said that the husband is called nggu yatangane or "my man," though there was some doubt whether this word did not belong to the Nggaliyalatini dialect.

One of my Dhawanisa informants was able to carry back his pedigree in the direct line for five generations and the terms used for these ancestors differed somewhat from those of Nandrau, The great-grandfather was called momo, the same term as was applied to the father (my informant was a chief) but with the addition of the word vakarua, meaning second, and this momo vakarua would have called his greatgrandson luvenggu vakarua. So far this is in accordance with Nandrau usage, but all ancestors beyond this were called tumbunggu, the great-great-grandfather being tumbunggu vakarua and his father tumbunggu vakatolu (third grandfather).

The three systems of the Nandrau, Navatusila and Dhawanisa peoples, which are essentially on the same plan, present many features of great interest. My data from the Navatusila and Dhawanisa are scanty and if they stood alone I should hesitate to lay much stress on their special features, but their agreement in all essentials with the more fully recorded Nandrau system can leave no doubt about the general accuracy of the information, though further inquiry would certainly show the necessity of some revision in detail. The most novel and important feature of these systems is the use of the same terms for the members of alternate generations. Among each of the three tribes a man gives the same name to his son's son as to his younger brother, and conversely the grandson gives to his grandfather the same term as that he uses for his elder brother. Among the Nandrau people the son's wife is called nau, the same term as is applied to the mother, and the son's wife and the mother are probably also classed together in the other two systems. In another relationship among the Nandrau people, there is a corresponding

usage, the son of a vasu (i.e. sister's son's son) being classed in nomenclature with a brother. A third character which places members of alternate generations on the same footing among the Nandrau and Dhawanisa people (and probably also among the Navatusila) is that the grandfather of the father receives the same name as the father while the grandfather of the grandfather and the grandfather himself are also similarly designated. Lastly, among the Dhawanisa the husband of the father's sister and the husband of the sister's daughter^ receive the same name, which again involves bringing under the same head members of alternate generations. These four features all point to the former existence of some social regulation which brought members of alternate generations into the same category and put them on a level for some purpose, but what this purpose may have been must be left for later consideration.

Another unusual feature of the Nandrau and its allied systems is the wealth of terms for grandparents and grandchildren. In many forms of the classificatory system, at any rate in Oceania, the parents of the father receive the same designation as the parents of the mother and often no distinction corresponding to sex exists, so that there is only one term for all kinds of grandparent. In the Nandrau system, on the other hand, there are four distinct terms for the four grandparents and also four terms for the reciprocals of these, i.e. each of the four kinds of grandparent has his or her appropriate term for the child of son or daughter. The distinction is not as with us between a grandson and a granddaughter but between the child of a son or of a daughter irrespective of sex, and each male grandparent uses a term different from that employed by his wife.

The NaTnbombudho.

One other system of the interior of the island was obtained fairly fully though only from one informant ; that of the people called Nambombudho obtained from Malakanauatu, the mbuli of Nasonggo. This village is on the chief road to Suva, and whether for this or some other reason it falls into line in its chief features with the Mbau system though showing definite evidence of community with the mountain systems.

' The term is also applied to the wife of a sister's son.

The N ambo7nbtidho 285

The father is tamanggu but the husband of the mother's sister, who usually has the same name, is called nonggu ikau^. Mother and child are called iinanggu and luvenggu. The mother's brother was said to be either momo, vunonggit or ngandinanggu. The father's sister is nganeitamanggu and both her husband and the wife of the mother's brother are vunganggii. The sister's son is nonggu vatuvu.

The names for grandparents and grandchildren are difficult to understand and I do not like to vouch for their accuracy. The father's father is called tumbunggu and the father's mother ndamanggu, evidently an abbreviation of the na ndamanggu of Nausori (see p. 268). The mother's father is makumbunggu, a word generally used for grandson; this use was known to my informant but several examples of its application to the mother's father were obtained from the pedigree. A man calls his son's child and a woman her daughter's child makmnbunggu but a woman calls her son's child nonggu ndiva and a man his daughter's child nonggu itangi. The last two words are used in the same sense as in Nandrau and it seems likely that some of the other terms have been borrowed from the Mbau dialect, but were not used by my informant correctly. It is at least noteworthy that the two terms which undoubtedly occur in the systems of the mountain tribes should be used in the same sense as by those people.

A man calls the son of his mother's brother or of his father's sister tavalenggu and he addresses their daughter as nonggu ivola but I omitted to obtain the term used by two women. The term tavalenggu is also used for the sister's husband (m.s.) and for the wife's brother while the husband's brother and the brother's wife (m.s.) are called ndaku, the term used for these relationships by the other mountain tribes. The wife's father and mother are vunonggu but nganeitamanggu was also used once for the latter relationship. Here as elsewhere people are spoken of as the parents of their children and a man whose daughter has had a child will call this daughter nanairau, or our mother. There can be little doubt that this system must have been at one time of the same kind as that in use among the Nandrau people but is undergoing modification so that it now shows some of the features of both inland and coastal systems.

1 Mr Hocart tells me that this word is also used for a step-father.

An account was obtained of the tambu objects of several niata7iggali of Nasonggo. The Nodhavondu beHeved in their descent from the owl {Juki) which they might not eat. The Kainasonggondina traced their descent from, and might not eat, the snake, and if they did so, the arms, legs and neck of the offenders swelled. The Kaisalandina might not eat the nda77tani, a yam with purple flesh. Other matanggali, the Kaisavu and the Naisausau, had special stones called uluvatu^ which seemed to be included in the same category as the tambu animals or plants. Only the priests were allowed to touch these stones which were believed to protect the people of this maianggali.

The People of Tavua.

A fairly full account was obtained of the Tavua people on the northern coast of Viti Levu. There are four mata7tgga/i : the Kainavauvau, Kainambuna, Kaivanuakula and Kaimbila. The last was said to be different from the rest ; its people had lived elsewhere, it was said in the Yasawa Islands, and after having been burnt out of their homes in warfare with a confederation of tribes, they had come to join the Tavua people. Their language differs from that of the rest of the Tavua people, though it is gradually becoming more like it. Each of the four matanggali has its own chief, but in the time of the last great war they combined to elect a supreme chief who was the grandfather of my informant, Ratu Veniasi.

Each of the four matanggali is divided into itokatoka, called also mata ni mbure or "face of the mbure,'' each itokatoka having its own men's sleeping house.

The following are the itokatoka of the Kainavauvau matanggali; the Nambulou, Kainggavindi, Kaindingi, Kainaimburu, Wandai and Yavusambalavu, the last word meaning the great house. The Wandai furnished the mata ni vanua or messengers of the chief.

The Kainambuna have two itokatoka called Nduira and Nduiata, both of which were said to be a ikove ni tambua^, and they supplied officers to the chiefs of the Kainavauvau. This matanggali would therefore appear to correspond to the 7natanggali vanua of the Nandrau.

• Mr Hocart tells me that uluvatu is also the name of small spirits corresponding to the Ittve ni ivai of other parts of Fiji. 2 The snatcher of the whale's tooth.

The Tavua 287

The itokatoka of the Kaivanuakula are the Vindrandra or itokatoka of the chiefs, the Kainasonini, Rara, Mbamba, Nggaudhia and Namara.

The Kaimbila have as itokatoka the Nandula or group of the chiefs, the Vinangonasau, Vinangoiratu, Lova, Tavasi, Naiavusambalavu, meaning the long group, and the last again had six subdivisions called Oimua, Nasariti, Natokoraki, Vunikura, Tavasi and Mondelau. Each of these itoka- toka as well as each subdivision of the Naiavusambalavu had its own mbure^.

All the four groups of the Tavua people have snakes as their tambii animals. It is doubtful how far these snakes are real. Some of their characteristics must be more or less mythical but it was said that the animals could be seen and handled, and the rule that they may not be eaten also seems to imply that some, at any rate, are real.

The Kainavauvau and Kainambuna possess red snakes descended from a certain mother snake called Tunandal The snakes of the Kainavanuakula have broken tails and are descended from a snake with a broken tail called Mundu. The snake of the Kainambila lives on a certain kind of banana called nakukoto or matadhawaka and is descended from a female ancestor called Tunavuni.

The itokatoka have no animals peculiar to themselves except one subdivision of the Kainambila, the Oimua, the members of which might not eat the nggarau or crab when they were in their original home and the people believed in their descent from this animal.

The snakes of the Tavua people are believed to be friendly to the people of the divisions connected with them. Thus, if one of the Kainambila saw one of his snakes on the banana they frequent, he would say "Is that you, Vakawali ? Come along ! " and the snake would crawl along the man's arm, climb on his shoulder and twine itself round his neck.

1 I did not learn of any special differentiation of function among these jnatanggali and itokatoka with one exception which I owe to Mr Joske. Sometimes certain families or other social groups have a special reputation as jokers. An example from Mbau is that a certain people told to thatch a house will beat it with small clubs instead, the point of the joke being that the word for "thatch" is the same as for these small clubs. It was said that the Mondelau division of the Naiavusambalavu itokatoka had this kind of reputation.

2 It may be noted that the common origin of the snakes of these two divisions confirms the other evidence that their relation to one another is closer than with the other matanggali.

The following are the terms of relationship of the Tavua people : —

The father, the father's brother and the husbands of the mother's sisters are called ta ; the mother, her sisters and the wives of the father's brothers are nau, while a child is luvenggu, these three terms being used as in the Mbau system. The mother's brother, his wife and the father's sister are called ngivandi while the son of the sister of a man and the child of the brother of a woman are both ngguva^. The term ngwandi is also applied to the parents of the husband or wife though these may also be called vunganggu, and correspondingly the son's wife and the daughter's husband are called ngguva though they also may be called vunganggu. The most exceptional feature of the system is in the nomenclature of the children of the mother's brother and father's sister, these being called tuakanggu or tadhinggu according to age, these being as in other systems also terms for brother and sister. Thus, cross-cousins who in all the other recorded systems of Fiji have special designations are here classed with brothers and sisters. The terms tuakanggu and tadhinggu are used in address between men and women, i.e. they are in use not only between brothers and between sisters but also between those of different sex, and the term nganenggu is not used in address but only when a man speaks of his sister or a woman of her brother. The term ngguva is applied to the child of a woman who is called tuakanggu or tadhinggu even when this denotes the cross-cousin relationship ; thus, a man calls the child of the daughter of his mother's brother ngguva. The term tavale, generally used in Fiji for the cross-cousin relationship, is here used for the brother's wife (m.s.) or the sister's husband (w.s.) and it is also used for the relationship between the men who have married two sisters. It is remarkable that ngwandi, apparently an abbreviation of ngwaneitina, should be used for the father's sister, the wife of the mother's brother and the wife's mother as well as for the mother's brother and the wife's father. The wives of two brothers are called ivanggu, this being evidently the raivanggu of Nausori and Nasonggo without the honorific prefix.

Grandparents of all kinds are called tai and grandchildren

' The first part of this word is possibly the possessive and va an abbreviation

of vasu.

Inland and Coastal Systems 289

viangonggu, the terms thus resembHng those of the hill tribes while the method in which they are used is rather that of the Mbau system. The great-grandfather is tai vakarua and his father again tai vakatolu. People are spoken of here, as elsewhere in the island, as the fathers and mothers of their children, but the term applied to a son or daughter is not changed after either has had a child as among the mountain tribes.

In this system the use of the terms tuaka and tadhi for cousins depends on the respective ages of the fathers, not on the relative age of those who give these names to one another. Thus, if a younger brother has a son before the elder, the son of the younger brother will be tadhi though actually older. This is probably true throughout Viti Levu.

A few of the more important differences which distinguish the systems of the interior of the island from those of the coast may here be mentioned.

The mountain systems are on the whole much richer in terms of relationships, though certain distinctions are absent in them which are present in some of the coastal systems. The richness of the mountain systems is especially noticeable in the terms for grandparents and grandchildren. The Mbau system of the coast, on the other hand, shows its richness in the nomenclature for the cousins who are either the potential spouses of the speaker or of his brothers or sisters, a feature which is readily intelligible in connection with the crosscousin marriage. Another feature of the inland systems is that they distinguish between the parents-in-law and the mother's brother or father's sister, and in addition there are special terms for the wife of the mother's brother and for the husband of the father's sister, while in the Mbau system these relatives usually receive the same names as the father's sister and the mother's brother respectively. The most important and exceptional feature of the mountain systems, however, is the application of the same term to the members of alternate generations with which I have already dealt. There is no evidence of such correspondences in the coastal systems, though there is one term which may possibly be a survival of such a feature. One of the coastal words for the grandfather is tukanggu, which bears a close resemblance to tuakanggu, the term for elder brother, and this similarity suggests that we

have a relic of a time when a man gave the same term to these two relatives as he still does in the mountains at the present time.

Another difference between the two kinds of system is in the nature of the possessives. Suffixed pronouns are used with nearly all the coastal terms, the exceptions being vasu or vatuvu and itdaku which also occur in the inland systems. In the interior, on the other hand, possessive nouns are quite as frequent, if not more frequent, than the pronouns proper, and here the terms which take the suffixed pronouns are also found in the coastal systems while the terms used with a possessive noun are not only peculiar to the interior (with the two exceptions already mentioned), but these terms are often different in the three systems recorded. Further, it may be noted that in several cases in which the inland tribes use suffixed terms, there are alternatives which may be used in their place. Thus, the tuakanggu of the coastal systems is often used in the interior for the elder brother, but this relative is also frequently called tutua and there was little doubt that this is the term proper to the three systems which were studied. Similarly, there are alternative terms for the relatives by marriage who are usually called tavalenggii. On the other hand, it must be pointed out that no alternative was ever used for tad/iinggu, and yet this term was used in a very special sense to include the son's son and the younger brother, a sense limited to the mountains.

I must be content here to point out these differences between the coastal and inland systems. Their significance must be left for the theoretical discussion in the second volume.

Functions of Relatives.

Fiji is noted for two especially striking cases of special functions pertaining to certain relatives.

The special relation of a sister's son to his uncle has reached its highest known manifestation in the vasu institution of Fiji while these islands also furnish a most pro- nounced example of the avoidance between brother and sister. I was not able to collect a sufficient number of concrete examples to allow me to reach any very definite conclusions concerning the relative importance of these functions among

Functions of Relatives 291

the coastal and inland people, but I received the impression that both were somewhat less developed among the inland tribes.

At the present time the avoidance between brother and sister seems to have largely or entirely disappeared. In the old days a man could not utter the name of his sister, but I met with no single case in which the smallest reluctance was shown in giving me the name of a sister when I was collecting pedigrees. Other ancient restrictions were that the brother and sister might not speak to one another and might neither play nor eat together, nor might they wear the same loincloth. If a man wished to communicate with his sister he employed some intermediary. He would never utter the personal name of his sister but would address her as 0 iratou na lewa, i.e. " those women " in the plural, and this method of denoting the sister would also be employed when speaking of her to anyone else. In the Tavua district it was said that a man would communicate with his sister through his mother. If these restrictions were not observed, a man would be regarded as mad and would be killed either with a club or by strangling {kuna).

The special relation between a man and his sister's son, on the other hand, still persists, though shorn of many of its most picturesque features. At the present time among the inland people the mother's brother still takes the chief part in the direction of his nephew's life. He arranges and takes the lead in any ceremonies connected with his nephew and in the old days taught him the art of warfare ; on the other hand, when a man dies the mother's brother does not act as guardian but this duty falls to the father's brothers.

In the old days the sister's son could take any of the possessions of his uncle and could, if he wished, kill any of his pigs. In this respect the old custom is no longer followed, and if the right were exerted, the nephew might be taken to the court where his act would be treated as an ordinary case of larceny^ On the coast it is well known that in the old days the vasu of a chief could go to the town of that chief and take any woman he desired and it seemed that this was once also true of the inland tribes.

I could not discover that the uncle had any special power of stopping his nephew when he was fighting ; if he did so,

^ This item of information was given to me while I was staying with a magistrate and must be accepted with reserve.

19 — 2

it was only looked on as part of his general right to exact obedience. On the other hand, it was said that a man would kill anyone who was fighting with his sister's son.

I inquired especially into a point which the many previous accounts of the vasu institution have left doubtful, viz. whether there is any difference between the duties and privileges of the own sister's son and of the many other persons who would be classed with him through the classificatory principle. It seemed quite clear that the rights would be exacdy the same for the actual sister's son and for the son of a woman called by the same term of relationship as the sister, but in reality the father's brother's daughter. In fact, it seemed that the rights even extended to any one of the uncle's matanggah. It was also said that the husband of the father's sister, classed in nomenclature with the mother's brother, had the same duties and privileges and was liable to have any of his possessions taken by the son of his wife's brother.

The mother's brother takes a leading part in the ceremonial connected with incision \ The feast on this occasion is given by the boy's father but the boy takes one of his uncle's pigs and gives it to his father^ It may be noted that a boy may take his uncle's possessions before incision.

Since marriage now takes place within the matanggali, it often happens that a man is vastc to his own 7natanggali, and it was said that this was regarded as the highest kind of vasu, being called vasu itaukei. The exact meaning of the terms veidhakavi and veilavi in the system of Nandrau, which I could not fully understand, was certainly connected in some way with the distinction between a vasu to his own people and one of the ordinary kind.

Another interesting feature connected with this relationship is that a man will not eat the tambu animal of the tribe to which he is vasu. The result of this is that a man respects the animal of his mother as well as that of his father, but it is

1 I use this term for the operation in which the prepuce is slit longitudinally without the removal of any part. This is the usual mode of performing the rite which is usually known as circumcision in Melanesia and Polynesia, and its relation to circumcision proper is so doubtful that it should be denoted by a special term.

2 Among the inland tribes the boy after incision has to run the gauntlet and can be beaten by anyone on the way back to his home. The dressing is kept on till the wound is fully healed and then buried by the boy in a place which he keeps absolutely secret.

Functions of Relatives 293

noteworthy that the matter was never put in this way but over and over again the prohibition was given as one of the functions of the vasu relationship. This restriction appHes to the tambu plants as well as to the animals.

In the Rewa district the only point I learnt about the vasu which is new, so far as I know, is that a man and his mother's brother always speak to one another in a slow and gentle manner, and that the uncle instructs his nephew in the code of morality.

I made no special inquiries into the functions of the father's sister. I visited Fiji before going to Tonga and nothing was said to give me the slightest suspicion of any important functions pertaining to this relative, but I should not be in any way surprised to find that there exists a relation similar to that found in Tonga (see Chap. XIII).

I could not learn that the wife's brother or the sister's husband had any special duties or privileges.

In the Tavua district a man would not speak to his brother's wife and the restrictions on their intercourse were said to have been the same as those between brother and sister. A man also could not talk to or utter the name of his wife's sister. If he wanted to communicate with her, he had to do so through an intermediary. It was especially said that this differed from the custom of Mbau. There are no restrictions connected with the father or mother of husband or wife.

In the Mbau district veivungoni could not talk to one another \ Since the veivungoni include the mother's brother, there seems here to be a discrepancy with other evidence. What was perhaps meant was that if a man marries the daughter of his mother's brother, he can no longer talk to him. In such a case a man would communicate with his uncle through his father, while if he married the daughter of his father's sister, he would communicate with his aunt through his mother, or in either case his wife might act as intermediary. The personal names of veivungoni are only forbidden in direct address and may be used when speaking of these relatives to others.

^ Mr Hocart tells me that this probably means that they might not talk together for the sake of talking, though they might speak to one another on business. It may be that this is all that is meant in other cases of this form of restriction.

Marriage.

The system of Mbau and those allied to it which are found on other parts of the coast evidently owe their special characteristics to the influence of the cross-cousin marriage. The use of the same term for the mother's brother and the father of the consort on the one hand, and for the father's sister and the mother of the consort on the other hand, and the fact that all these relatives are classed together as veivmtgoni are evidently the direct results of the regulation which makes a union between the children of brother and sister the natural marriage of these people. To this origin is also due the use of the same term to denote the wife's brother and the son of the mother's brother or father's sister. In fact, anyone conversant with systems of relationship could at once diagnose the existence of the cross-cousin marriage in the present or the past from the nature of the systems of the Mbau and Rewa districts.

When we turn to the mountain tribes, the evidence afforded by the terms of relationship for the existence of the cross-cousin marriage is far less definite and indeed the only direct evidence is in the use of the term tavale to denote both the children of brother and sister and certain marriage relationships.

A piece of evidence against the existence of the crosscousin marriage among the mountain tribes is to be found in the distinctive terms for the parents of father and mother. One of the consequences of the marriage of the children of own brother and sister is that a person has only two grandparents and it is therefore natural that in the Mbau system there should be only two terms for the relatives of this generation. In the mountain systems, on the other hand, there are four distinct terms for the four kinds of grandparent, a condition which is incompatible with the fully developed crosscousin marriage, i.e. with the marriage of children of own brother and sister.

The evidence in favour of the existence of the crosscousin marriage derived from the nature of the systems of relationship of the mountain tribes is therefore of very doubtful value, and I failed to obtain any direct evidence of its existence at the present time or in the near past from the pedigrees

The Cross-cousin Marriage 295

collected from the members of three mountain tribes. If the cross-cousin marriage had existed among these people with any frequency, it must have shown itself in the seven or eight fairly extensive pedigrees collected from the three tribes, and yet in none of them was there a single example of a crosscousin marriage. Further, there was an absence of knowledge of, and apparently of interest in, the pedigrees of mother and wife as compared with that of the father, which is difficult to understand if the cross-cousin marriage were extensively in vogue.

When we turn to the system used by the people of Tavua we are met by a state of affairs very difficult to understand. It may be said that the essential feature of the mode of relationship set up by the presence of the cross-cousin marriage is that it divides relatives of the same generation into two very clearly separated classes, one class consisting of persons whose intermarriage when of different sexes is absolutely prohibited and another class in which persons of different sex are natural if not obligatory spouses, and in Mbau and other places where the cross-cousin marriage exists these two classes are definitely distinguished in the nomenclature of the systems of relationship. The unusual feature of the Tavua system is that cross-cousins are not distinguished from other cousins, and yet there is evidence of the existence of the cross-cousin marriage. In this system the term ngwandi is applied to the mother's brother and the father's sister and to father-in-law or mother-in-law, and similarly the sister's son receives the same name as the son-in-law. These features point unmistakeably to the existence of the cousin marriage in the past, and I obtained evidence of its recent occurrence. There was, however, only one example in an extensive pedigree obtained in this district, and when I asked for other cases few could be given, and most of these were not examples of the marriage of the children of own brother and sister but of more distant relatives. The exact condition can only be discovered by more extensive genealogical inquiry, but my impression is that the cross-cousin marriage has almost disappeared in this district, though the system gives clear evidence of its existence in the past. It is even possible that the few recent instances I obtained are due to a reintroduction of the practice owing to influences from other parts of the island.

296 Melanesiait Society

In many parts of Viti Levu the cross-cousin marriage is so essential a part of the social structure that a man may regard the daughter of his mother's brother as his wife by right with no previous negotiation. Thus, in cases of rape brought before the courts in recent times the defence may take the form that the woman, as the daughter of the mother's brother, is actually a wife.

The evidence as it stands thus points to the cross-cousin marriage having formed no part of the social organisation of the mountain tribes ; of its having at one time existed in Tavua where it has almost disappeared, while it would appear to be the essential feature of the marriage regulations of Mbau and other parts of the coast.

My own genealogical material is far too scanty to allow any conclusions as to the actual frequency of the cross-cousin marriage in those parts of the island where it undoubtedly exists. The presence or absence of this institution and its relative prevalence in those parts where it exists can only be satisfactorily studied by means of a far more extended application of the genealogical method. Only in this way can we learn how far cases of the cross-cousin marriage have been between the children of own brother and sister, and how far they are the marriages of more distant relatives who stand to one another in the relation of ndavolana.

I may mention here that in a pedigree obtained from a chief of Nausori there were two consanguineous marriages, in one of which a man had married the daughter of a half-brother of his mother's father, while in the other case the marriage had been with the granddaughter of the mother's half-brother, the man in each case marrying a woman of a generation different from his own. There was no case in this pedigree of a genuine cross-cousin marriage.

Another marriage custom which is practised in Fiji is the levirate ; when a man dies his widow is often married by the brother of the dead man, and I could not discover that there is any special regulation limiting this right to the younger brother.

In concluding this section on Fiji I must again give a warning concerning the provisional nature of the material I have recorded. The whole of it was collected in little more than a dozen working days and any satisfactory control was impossible. It is, however, quite clear that there is very

Variety of Systems 2()']

great variety in the systems of relationship of Viti Levu, to say nothing of other islands of the group, and that we have in these varying systems features of the greatest interest and importance. It is clear that even at this late day it is still possible to learn much about the social institutions of Fiji, greatly affected as these have been by the decay which is destroying the ancient culture of the people.

CHAPTER XII

TIKOPIA

TiKOPiA, sometimes known as Barwell Island, has become well known as the place where Dillon first learnt the probable fate of La Perouse on Vanikolo. It is a tiny volcanic island, not more than seven miles in circumference, situated in latitude 12° 17' S. and in longitude 168° 58' E. The nearest inhabited islands are Anudha or Cherry Island, 60 miles to the north-east, and Vanikolo, 118 miles to the north-west.

According to Dillon Tikopia was first visited in 181 3 by his own vessel, the Hunter, though an ineffectual attempt had been made to get into communication with the natives by another vessel many years earlier, probably by the Barwell in 1798. Since that time it has been visited at intervals, chiefly by whalers, and occasionally labour-vessels have taken away natives of the island, but they have been found such poor workers and so devoid of stamina that vessels now rarely go to the island for this purpose. The island was visited in 1857 by Bishop Selwyn and later again by Bishop Patteson, and about 1905 some teachers from Motlav were placed there by the Melanesian Mission, and the island is now visited twice a year by the missionary steamer, the Southern Cross. It was on one of these occasions that I spent a day on the island. The scene was one which I had no idea might still be witnessed in the Pacific. The evidence of outside influence was of the slightest ; very few of the people wore anything but the native dress, a loin-cloth of tapa stained with turmeric. They swarmed over the ship in the most fearless and free manner, talking vociferously and ready to lay hands on any object which took their fancy, their general appearance and their teeth and lips stained red with betel mixture driving some new members of the crew to hide

John Maresere 299

themselves in alarm. On shore the vociferous and mostly one-sided conversation continued, while the natives were most assiduous in their attentions and demonstrations of affection. We visited one of the chiefs on the other side of the island and were given some excellent food consisting of sago flavoured with coconut, but as we had no knowledge of the language serious inquiry was impossible.

On this visit we took back to the island a man named John Maresere who had been there for many years and had been sent away in a manner shortly to be described and was now about to reside again on the island. When I went on the Southern Cross on her next voyage six months later, I found this man on board, as he had taken the opportunity of the visit of the ship to leave the island. I began to ask him questions and it soon appeared that he had so much knowledge of the manners and customs of the inhabitants and seemed so trustworthy a witness that Mr Durrad and I obtained as full an account as possible of those features of Tikopian culture of which he had knowledge. The account which follows is thus given by one witness only who was not a native of the island, and as the value of the account turns largely on his veracity and trustworthiness I propose to preface the account by a record of his history.

John Maresere was a native of Uvea or Wallis Island, where he had been to a Christian school, and that this school was Roman Catholic was made probable by a medal given to him there which he still wore. His canoe had been driven out of its course on the way to Samoa and had drifted to Tikopia\ He had stayed in the island for twenty years, where he seemed to have acquired a position of some power, partly, at any rate, owing to his medical knowledge. At the end of the twenty years he had been sent away from the island in a canoe, in accordance with one of the customs of the island when inflicting punishment, and had made his way to Vanikolo. There he had been found by the Southern Cross and taken to the Banks Islands, where he had settled on Motlav. On the occasion of our first visit in 1908 he was returning to Tikopia, which he left again on the second voyage of the Southern Cross in that year.

^ It is remarkable that at the time of Dumont d'Urville's visit there was on Tikopia a native of Wallis Island (Howvea) who had been driven to the island in a similar manner.

There is still another of his fellow castaways, John Patita or Virikeka of Wallis Island, on Tikopia, and Mr Durrad obtained from him an account of their journey. The castaways were John Maresere, John Patita and Moses Tongare of Wallis Island and Pongi, Zephaniah, Muri and Antonio of Tonga. The three men last named had killed a man in Tonga and had fled to Wallis Island. They heard that people were looking for them, so they stole the cutter of a trader for whom they were working and taking the others with them tried first to get to Samoa. They took with them yams, taro, coconuts and tomagoes (a special kind of yam). They could not get to Samoa, the wind being against them, so they tried successively to make Tokelau, Vaitupu, Tonga, Fiji, Futuna, Rotuma, and Panapa (? Paanopa)\ and finally reached Tikopia, where the cutter was wrecked in the surf. They were first received by a man named Paefakofe (see pedigree X), who adopted Muri, Antonio and John Patita, while John Maresere was adopted by Paerangifure (see pedigree IX), Zephaniah by the chief of the Tafua, Pongi by the chief of the Fangarere, and Moses Tongare by the chief of the Kafika. Six or seven years ago a ketch belonging to a planter of Epi came to the island and took away Zephaniah, Muri and Antonio, it being believed by the men themselves and by those left on the island that they were to be taken to Tonga on account of the murder committed many years before. Either the captain of the ketch had no idea that he was supposed to be carrying the men to justice or he made use of their guilty consciences to obtain three recruits.

Mr Durrad was able to confirm John Maresere's estimate that he had been in Tikopia for a long time. John Patita was able to point out men who were little children when he landed and a large coconut tree which he had planted soon after his arrival.

The fact that the information we obtained came exclusively from one man made us very attentive to any indications by means of which we might gauge his trustworthiness and I can only say that we became more and more impressed with his merits as a faithful observer as we proceeded. There

1 This list would indicate a very irregular course ; if carried to the northward, it would be natural to try to reach the Tokelau group and if they found they were then moving westward, Vaitupu I. would also be a natural destination, but it is improbable that they would ever have approached Tonga unless they had been carried southwards from the beginning.

Value of Evidence 301

were certain departments of knowledge on which to our disappointment he could tell us nothing ; thus, he had paid no attention to any folk-tales for lack of interest in, or even contempt for, them, and he knew none of the prayers or other formulas because these are always uttered in too low a voice to be heard by those present. In such cases, however, he was always ready to acknowledge his ignorance and could give a satisfactory reason for it. In his prolonged account we detected very few inconsistencies or contradictions and though there are doubtless mistakes, it is probable that these do not bulk largely in the total record.

Since my visit Mr Durrad has spent several weeks in the island and has not only been able to confirm the general correctness of John's narrative, but he has sent me further accounts of several subjects together with many illustrations and pedigrees, which he kindly allows me to put on record here. I was at first inclined to incorporate this information in the account given by John Maresere, but on second thoughts it has seemed a better plan to keep it separate, and John's account is recorded essentially as it had been written before Mr Durrad's account was received. The only feature which has been altered is the spelling. John's account was given in the Mota language, and at the time we were in many cases uncertain whether Tikopian words were not being pronounced as they would be in this language. In general John pronounced f 2iSV, k as g, and r as /, and it is possible that the last feature comes from his native language of Uvea.

John's record has an interest in itself apart from that of the information given in it. It is the account of one belonging to an alien though allied culture, who had spent many years on the island. It shows how far such alien visitors are permitted to enter into the inner lives of the people among whom they settle, and it illustrates the powers of observation and inference possessed by a man who was probably an average Poly- nesian. Further, it has still another interest. As time goes on, anthropology will have to rely more and more on such narratives as those of John Maresere. There are many parts of the world where even to-day the only accounts to be obtained of native modes of life are those given by people such as John Maresere, people who have seen the native life more or less from the outside, sometimes because, like John,

302 Melanesiait Society

they are aliens, sometimes because, as Christian converts, they have not been admitted to full participation in much of the life they are describing. One great task of the future anthropologist will be the valuation of evidence, and it is partly in the hope that John's account may be useful in this respect that I give it untouched, for it is probable that we may yet obtain a full and accurate account of Tikopian culture which will enable the value of his story to be fully gauged.

I should like to point out that in John's account we have in several respects conditions resembling those of Mariner'. In both cases we have aliens settled for a time in Polynesia, and in both cases the accounts were obtained from them after they had left their respective islands. In some ways John Maresere had certain advantages over Mariner, for the latter was living among people wholly alien to himself in mode of life and thought, while John was a Polynesian among Polynesians. Further, Mr Durrad and I who took down his record had, for purposes of comparison, John's knowledge of Banks custom acquired during his later stay on Motlav, and some of the points of greatest interest were brought out by an inquiry into the comparison of Motlav and Tikopian procedure. While his knowledge of Melanesian and other Polynesian custom was thus a source of help, it must be recognised that it was also a source of danger, for it raised the possibility that our informant might confuse the three cultures with which his life had brought him into touch. He recognised himself how greatly the culture of the Banks Islands differed from that of Tikopia, and confusion between these islands can safely be excluded. On the possibility that he may sometimes have been giving us the customs of Uvea in place of those of Tikopia there must be more doubt. Here again, however, he was able in some cases to compare the customs of the two places, especially in medical practice, and though this source of confusion cannot be excluded with certainty, it is improbable that it has had any influence on the value of the Tikopian record.

Before I pass on to consider this record a few words may be said about the physical characters of the people which show

^ Account of the Natives of the Tonga Islands, compiled and arranged from the extensive communications of Mr William Mariner, by John Martin, London, 1817.

Plate XIV

Fig. I. Tikopian man.

?1

-la i ti

Fig. 2. Tikopian men, showing methods of wearing hair.

Plate XV

mM A^

./fiO'Vf-ft

Fig. I. A village in Tikopia.

Fig. 2. A Tikopian hut.

Social Organisation 303

that they must be classed with the Polynesians. They have straight hair which the men wear long, either in flowing tresses, or bound up so as to keep it from covering their faces (see PL XIV, Fig. 2). The features on the whole resemble those of Polynesians though we saw some persons whose appearance strongly suggested an admixture of Melanesian blood, and Mr Durrad's photographs well illustrate the diversity of physical type. The people appear to be exceptionally tall and are probably considerably above the usual height of either Polynesian or Melanesian. There was much variety in the colour of the skin in various shades of brown, again suggesting some Melanesian mixture, but the general colour was nearer that of the average Polynesian than of the average Melanesian.

The Account of John Maresere.

The people of the island are divided into four divisions, called respectively the Kafika, the Taumako, the Tafua and the Fangarere. Each of these divisions has a chief who is called by the name of his division, te ariki Kafika, te ariki Taumako, etc. Each division has its own district of the island and, so far as could be judged, forms an independent section of the community. There is no direct relation between the divisions and the regulation of marriage which depends wholly on kinship. Each of the divisions has associated with it certain animals or plants called atua, a word also used for ancestors, and while most of these atua belong to special divisions, some are common to the whole community. The Kafika have as atua the feke or octopus which they may not eat, a prohibition which they share with the whole population of the island, though the animal is regarded as especially sacred to the Kafika. This division has another atua, the kape, a plant resembling taro, and this may not be eaten by the members of the division while it is free to the rest of the people of the island. The Taumako may not eat the toke or sea-eel and a pigeon called rupe, both prohibitions being limited to this division. The Tafua may eat neither the. tmta (fresh-water eel), the flying fox [peka) nor the turtle [fo7tti). The first restriction is limited to the Tafua, but the flying fox is prohibited to all though regarded as especially sacred to the Tafua. It seemed clear that the turtle once occupied

304 Melanesiait Society

the same position and was eaten by none of the people, but recently the common people of divisions other than the Tafua have begun to eat it, though the chiefs still abstain. It seemed that this departure from the old custom may have been due to the influence of John who had eaten turtle himself without harm and had thereby induced the ordinary people to follow his example. The Fangarere do not eat a fish called one or omi and a small black bird called moko, these prohibitions being limited to this division. No one may eat the sting-ray and it did not appear to be sacred to any one division. A person who may not eat an animal may also not kill it ; if for instance one of the Fangarere caught an one fish he would usually throw it back, but he might sometimes give it to a member of another division. On the other hand, we were told that if a man of one division killed the sacred animal of another he would fall sick and would send for a man of the division to which the animal belonged who would make him well by calling on his atua. This apparent contradiction is probably due to our informant having in mind in the latter case those animals which are regarded as more or less sacred to the whole community, and in accordance with this, it was said that if anyone kills a flying fox the coconut trees would cease to bear, while if the fresh-water eel sacred to the Tafua be killed, the spring supplying the pool where it lives would become dry.

It was clear that these animals had collectively the same name as ancestors and were regarded as such, but the descent was from men who had turned into animals rather than from the animals themselves. The Kafika and Taumako believe in their descent from men who turned after death into an octopus in the one case and an eel in the other. The Tafua believe in their descent from a man who became a flying fox, but have the tradition that a second man of their division became after death a fresh-water eel. Similarly, two men of the Fangarere changed, one to the one fish and the other to the moko bird.

In addition to the prohibitions connected with these atua, three of the divisions are prohibited from eating certain plants, the yam, the taro and the coconut, and these were also called atua, but the consideration of these prohibitions may be deferred till later, when certain ceremonies connected with them will be described.

Chieftainship 305

The Chiefs.

The distinction between the chiefs or ariki and the ordinary people is very definite, this being perhaps most clearly shown by the marriage regulations. Marriage is not allowed between the two classes and if it is found that a youth of the ordinary people has had an intrigue with the daughter of a chief, the man is hanged (? strangled) or sent adrift in a canoe. No punishment is inflicted on the woman, but if there should be a child it would be killed.

Each of the four divisions has a head chief, but the members of the family of the chief belong to his class and are also called ariki. The head chiefs of the four divisions have not all the same rank, but the chief of the Kafika is regarded as the most important, the chief of the Tafua and Taumako next in order, with the Fangarere last.

The respect felt for the chiefs is shown very definitely in the behaviour of the people. If the chief is in his house a visitor who enters will kneel at the door and will move towards his superior on his hands and knees till he reaches him, when the pair touch noses in the customary manner. He then withdraws a pace or two and sits down cross-legged, no other position being allowed in the presence of a chief. The chief would then call out to his wife for food and the pair would talk while eating. When the talk is finished the visitor will ask permission to go and will leave on his hands and knees as he came without turning his back to his superior. If the chief is out of doors when he is visited the man who approaches him will kneel down at a few paces' distance and move towards him on hands and knees to salute him.

If a man wishes to pass a chief, he will say "(9 mata pa' {mata, eye ; /^, father). The chief will reply ''Elau! saere, opoir "All right! Walk, yes!" and the man will go by on hands and knees as when approaching him.

When in the presence of a chief the Tikopians remove the fillet of bark-cloth which they often wear on their heads. A chief's name is never uttered but he is called and addressed as te ariki Kafika, te ariki Taumako, etc. The sanctity of a chief seems too to be shown by the fact that no woman may go in his canoe which is known as a tapu canoe. The only special privilege of the chief is that he receives certain highly prized fish or the first caught if there are several.

3o6 Melanesian Society

A chief is succeeded by his son or, if he has no son, by his younger brother or by his brother's son. Under no circumstances would he be succeeded by his sister's son.

It seemed that the chiefs are definitely the rulers of the island, and that they have the deciding voice in social disputes, but it was clear that their functions are at least as important, if not more important, in ceremonial, in which, as we shall see, they take the leading part. All important decisions concerning social order depend on the will of the chiefs ; thus, it is they who decide whether strangers who come to the island shall be killed or treated as guests, and they settle the nature of the punishment inflicted for any breach of the customs of the island.

For severe offences the penalty is death, occasionally by hanging or possibly strangling, but more commonly by sending the offender out to sea ; if a man, in a canoe ; if a woman, by making her swim from the shore. The chief offence for which men are sent away is that already mentioned in which a man of the ordinary people is discovered in an intrigue with the daughter of a chief, and John could give no other example of an offence which involved this punishment. A man thus sent adrift is put at night in a canoe with some food which is provided by his father. The property of the victim — clubs, paddles, bows and arrows — is put in the canoe with him and he is decorated with armlets and a fillet of bark-cloth, such as is worn in dancing, is bound on his head. All this is done because it is recognised that the man is going to his death and, when the canoe has drifted out of the sight of land, the victim should take his club and break the bottom of the canoe so that he sinks and drowns. As the victim drifts away his relatives wail and they follow the same rules of mourning as if the man had died a natural death on the island. When John was sent adrift he was given a mat-sail and having no wish to follow the orthodox Tikopian custom of dying when he had lost sight of land, he set sail and reached Vanikolo in three days, nourishing himself meantime on twelve coconuts which had been given him for food. Usually only two or three coconuts are given because it is understood that there will be no need for their use. The procedure in John's case differed from that customary in another respect in that he was sent off at midday instead of at night.

Looseness in sexual relations appears to be the chief reason for sending away a woman who is driven to the sea and made to swim away. This punishment is only inflicted on unmarried women and it appears that, whatever her offence, no married woman is ever sent away in this fashion. This was explained by the very great respect paid to a married woman, so great that she would never be treated in this manner.

It sometimes happens that men or women sent away may be rescued by the people of another division who take the intended victim to their own part of the island. People thus rescued become members of the division which has rescued them, and if later their own people should repent and desire them back they would have to give presents in exchange for them. The only other offence recognised is stealing, which does not receive human punishment, but is dealt with by the atua in a manner to be described later (see p. 319).

Relationship.

The system of relationship is of the classificatory kind. The father, his brothers and the mother's sisters' husbands are called pa. The mother, the father's brother's wife and the mother's sister are nau, and the name for child, used in general as the reciprocal of the preceding, is tania. Brothers call one another taina, the elder being distinguished as te rumatua and the younger as te roto ; a brother and sister call one another kave. The terms taina and kave are used by all whom we should call cousins, whether the children of father's brother, father's sister, mother's brother or mother's sister. The mother's brother is called tuatina, and he calls his sister's son iramutu. The father's sister is masikitanga, with tama as its reciprocal.

All four grandparents are called tupuna and all grandchildren makupuna, both terms being used in the classificatory sense.

The husband is matua, the customary Polynesian word for elder, and the wife is nofine.

The father of the wife or husband is called pa, the same word as is used for father, and this term is also used for the daughter's husband by both man and woman.

Similarly, the mother of either husband or wife is nau, a term also used for the son's wife.

20 — 2

3o8 Melanesian Society

Again, a man and his wife's sister are taina to one another and also a woman and her husband's brother. A man calls his wife's brother and his sister's husband tangata and a woman calls her husband's sister and her brother's wife either fine or ma. Thus, men call one another tangata ; women are fi^ie or ma to one another ; while those of different sex call one another taina.

Two relationships of especial interest are those of the masikitanga or father's sister and tuatina or mother's brother. The masikitanga has the deciding voice in connection with her nephew's marriage ; she usually chooses his wife and if he chooses his own and she forbids the match, it would be given up. The masikitanga will give bark -cloth, mats and food to her nephew, but he may neither take them without leave nor ask her for them. Similarly, he will give things to his masikitanga, but she will neither take them nor ask for them. The nephew may go to his aunt's house and both talk and eat with her.

The tuatina or mother's brother has several important functions in relation to his sister's son. It is he who incises^ his nephew and, when the maromaro or perineal cloth is put on for the first time, there is an exchange of presents between the uncle and father of the child. The tuatina of a dead man takes a leading part in digging his grave. A man has to show no special obedience to his tuatina and, though each may take the possessions of the other, there do not appear to be any special regulations connected with the practice.

Another relative of interest is the tangata or wife's brother and reciprocally the sister's husband of a man. These two men may not say each other's names nor, when one is sitting, may the other take anything from above his head. If a man is sitting down his tangata will not go near him but will make a wide sweep round him and, if a man should wish to talk to his tangata, he would do so only from a distance. Similarly, a man and his tangata may not bathe in a stream at the same time, but if a man sees his tangata bathing he will wait till he has finished.

These rules of avoidance also apply to the relationship of wife's sister and the sister's husband of a woman who call one another taina, the term used also for brother and sister.

^ See note on p. 292.

Marriage 309

The wife's mother was spoken of as being very tapu ; nothing may be taken from above her head when she is sitting, and a wide sweep must be made round her when passing. The same rules apply to passing the wife's father but are not regarded as so important. It seemed that the ddtour made in passing him is not so great as in the case of the motherin-law.

Husband and wife address one another by name and the terms matua and nojine are only used in speaking of the relationship between them.

Marriage.

The regulation prohibiting marriage between the ordinary people and the chiefs has already been mentioned. Within each class marriage is regulated purely by kinship. A man does not marry anyone whom he would call kave, whether the daughter of brother or sister of father or mother or a more distant relative through the classificatory system.

A further marriage regulation is that a man may not marry the sister of his wife whom, as we have seen, he calls taina.

The social divisions have no significance in connection with marriage. Provided two persons are unrelated it is a matter of no importance whether they are of the same or of a different division.

The chiefs used to have several wives but in no case at present has a chief more than one. More rarely in the old time the ordinary people might have more than one wife, but this also is no longer practised.

When a man is old enough to marry, a wife is chosen for him by his father's sister, or if he himself chooses, the choice is ratified by her. His relatives then decide on a day to sally out and seize the woman who has been chosen, the bridegroom remaining in the house. The seizure of the woman is resisted, and a fight takes place with clubs and bows and arrows in which no one is ever killed though some are often badly wounded. At the end of the fight the father and maternal uncle of the woman are given presents of mats and cloth, and the woman is taken to the house of her future husband. That night the pair sleep together but the marriage is not consummated. On the next day the husband

makes presents to his tangata, the brother of his wife, and on that and succeeding nights the pair again sleep together without consummation. On the fifth day the wife goes to her father's house and helps to prepare food which she takes back to her new home, and after this the marriage is consummated.

A married woman is treated with great respect and it was said that adultery is very rare. If it occurs and is discovered the husband would at once attempt to kill the offending man, but the woman would only be chastised and not killed. A child suspected to be the fruit of an adulterous union would be killed. If there seems to be any danger of the husband failing in his attempt to kill the adulterer, a general fight may ensue, owing to friends taking the part of each. A severe wound which makes blood flow may bring the fight to an end. "When the injured husband sees the blood, his anger will go." If a man of one division commits adultery with a woman of another it would never lead to a formal fight between the two divisions, or there would be a fight but only "with the lips." Neither for this nor for any other reason do the different divisions fight with one another.

Before marriage cases of sexual laxity are probably not infrequent, though it did not appear that there was any prostitution or habitual freedom of intercourse, A girl found to be a frequent offender in this respect would be killed by being made to swim out to sea till she drowned. John was most emphatic in his statements that a married man would never offend with an unmarried woman. When the illicit intercourse of a youth and girl of the ordinary people results in offspring, the pair usually marry and if they do so there would be no slur on the child. Should the man refuse to marry in such a case, the child would be killed as soon as born. The same holds good for a man and woman belonging to the chiefs' class, but in this case it was said that the man would never refuse to marry the woman. If a man of the chiefs' class has an intrigue with a woman of the ordinary people, there could be no question of marriage. The child would be killed and, as we have already seen, the intercourse of an ordinary man with the daughter of a chief leads to the death of the man and the destruction of the child. Widows should never remarry but they do not always remain chaste. A widow of the ordinary people who is found to have offended is rebuked by her brother, but the widow of a chief is beaten

Childbirth 311

by her son, and the tuatina or mother's brother of the widow would make presents to her son, i.e. to one whom he would call makupuna or grandchild.

A man and a widow who have a prolonged intrigue would never live together unless the woman becomes pregnant when she may go to live with her lover. The union would not be regarded as legitimate but would be called te unoso or feipuke, the terms used for adultery and the act of sexual intercourse. The woman is never allowed to return to her own relatives, and is called taringa motu or "broken ear" in reference to the broken ear which is one of the signs of widowhood (see p. 314). The child of such a union is called tamai te ara or the child of the pathway. He is only allowed to go to the land of his father, and would be driven away if he went to the land of his mother's first husband.

After incision, however, he would be allowed to go to his mother's land, but only because he had been incised by his mother's brother and even then he would have to pay his uncle for any food he took away. This would seem to be the only case which one can regard as that of illegitimacy, for in all other cases of irregularity the child is killed at birth. It may be noted that incision by the mother's brother seems to be regarded as a kind of legitimation.

Birth and Childhood.

Childbirth usually takes place in the house, but the woman may go to the sea and be delivered in a shallow pool by the sea, the latter procedure being especially followed if there be any delay in the birth of the child. In either case the woman is attended by her mother or sister and if the birth has taken place by the sea, the woman who helps will take the child to the house while the mother bathes. She does this first in the sea and then in fresh-water, after which she puts on clean clothes and goes to the house. Both she and the child are then smeared with turmeric, the child having been previously washed, and this smearing of the child with turmeric continues till it can run about. The woman bathes in this way immediately after the birth of the child even if delivery has taken place in the house. The cord is cut with a bamboo knife. The afterbirth is buried anywhere and there do not appear to be any special beliefs connected with it.

In the case of a first-born child the mother will not work for a month, but after the birth of later children she will return to work in three days. When the cord drops off there is an exchange of presents between the father and the maternal uncle of the child, and the piece of cord is thrown away. The name of the child may be given at any time, usually by the grandfather of the child, his father's father.

On the tenth day the father of the mother takes his daughter with her child to his house where they stop for ten days, at the end of which time the husband fetches back his wife and child.

The next event is the assumption of the maromaro or perineal cloth when there is a feast and an exchange of presents between the father and the maternal uncle of the child. The ears are bored at any time and by anyone.

Incision takes place when a boy is about twelve years of age and it is done in the same fashion as in Tonga, the prepuce being slit but no part of it removed. Before the operation the chief of the division will make an offering of kava to the dead and will pray that the boy may pass through the ordeal successfully. The relatives of the boy assemble in the house but the operation is performed by the maternal uncle of the boy just outside the house. While it is being done the relatives in the house cry, the male relatives on both father's and mother's sides cut themselves on the forehead so that blood flows over the face, and the female relatives tear their cheeks with their nails. As soon as the operation is over the boy goes into the house where his relatives are gashing or scratching themselves and there is a small feast. That night all the relatives stay in the house and the next day there is a big feast at which presents are made to the uncle for performing the operation and to the chief for having made the offering of kava. It does not seem that the boy has to go through any ordeal beyond that of seeing his relatives with the blood streaming down their faces. After the incision the boy is smeared with turmeric mixed with coconut cream\ This is done by his mother or her brother and it is put on the shoulders and breast, the cheeks, chin and neck, but not on the back, forehead, lips or nose. When mixed with coconut the turmeric becomes very red and a boy seen

• The expressed juice of the white of the coconut, heated by means of hot stones.

Death 313

on the occasion of our visit was thought to be covered with a red pigment.

Up to the age of puberty the ceremonies for a girl are the same as for a boy but in her case there is no rite corresponding to incision. John did not know of any custom connected with the first menstruation and it seemed certain that there is no seclusion at this time. The tattooing of a boy is done soon after incision and of a girl soon after the first menstruation.

Infanticide.

A Tikopian family is usually limited to four children, any in excess of this number being killed by burying them alive in the house or just outside it ; occasionally five or six may be kept alive but never more. If the first four children are all girls one or more of these may be killed in the hope that succeeding children may be boys, in which case the lives of the boys would be spared.

In the case of twins, called masanga, one or both are usually killed, the decision as to what shall be done resting with the mother. It was said that there has never been a case in Tikopia where twins have been of different sex. While John was in Motlav such a case occurred and when he returned to Tikopia and told the people about it, they were quite unable to believe it, for they said that such a thing had never been known among themselves.

Death.

The dead are interred, usually in the house in which they have died. Occasionally a chief is buried outside and a shed erected over his grave. The burial takes place on the day after death, the relatives in the meantime crying and gashing themselves with stones and bamboo knives. The grave is dug by the maternal uncle and other relatives of the dead man ; it is made very deep and the body is wrapped in barkcloth and mats and laid on its back fully extended, this also being done by the uncle and other relatives. At the death of a chief another chief comes to make an offering of kava before burial takes place and it is probable that this offering is also made in the case of ordinary people.

The period of mourning is long. All the relatives drink only water and eat only mammy apples (Papaia) for four or five months and they often abstain from coconuts, taro, yams and fish, or one or more of these, for as long as a year. They also cut the hair short. The mourning is especially prolonged for the father of the deceased, who may eat nothing but mammy apples for a year. He must not fish nor will he go out in the daytime, for it would grieve him too much to see other young men walking about. In consequence, he will only bathe at night and will not drink where he and his son have drunk before but will seek water elsewhere. After this has gone on for a long time, some man will take pity on the mourner and cook food and take it to him, saying, " You have fasted for a long time and it is good that you should now eat." The father would be very reluctant to take the food, but after a time he would wail for his son for the last time and eat the food, afterwards making a present to the man who had brought it. Other relatives do not fast for so long and at the end of about eight months would be eating their usual food, and about this time the brothers of the dead man would again begin to fish. All the relatives would abstain from betel but only for about two months.

In addition to fasting, a widow shows her loss by making a number of burns on her face with a brand from the fire. She cuts off her husband's hair and wears it round her head till her death ; she breaks the lobes of her ears and always keeps her hair short. She remains in the house for a year, leaving it only at night, and, if it is necessary to go out in the daytime, she must see that no one is about and must cover herself with a mat. At the end of the year she can go out to work, and as we have already seen, she should never remarry.

A widower cuts his hair and keeps it short for one or two years.

For the first ten days after death offerings of food and kava are made at the grave daily and betel is usually offered too. These offerings of food and occasionally of betel continue indefinitely, probably for as long as the existence of the grave is recognised. The kava on the other hand, at first offered daily, is later poured out at longer intervals till six months after death, from which time the offering is made every six months. As we shall see shortly, the graves are the scenes of offerings of kava on many other occasions.

The Atua 315

Magic.

A most interesting feature of John's account of Tikopian culture is the complete absence of anything which can be called magic. During his stay in Motlav John had become acquainted with the very developed and malignant magic of the Banks Islands (see p. 156) and he said that there was absolutely nothing of the kind in Tikopia. If a man of this island wishes for any reason to kill another he would employ ordinary force. On the beneficent side also there do not appear to be any practices resembling those which are usually described under the heading of magic.

The Atua.

All the sacred beings are called atua and many of them are clearly the ghosts of ancestors. Others are the animals connected with the different divisions or reverenced by the people as a whole which have been already considered in connection with the social organisation. The name atua is also given to certain plants to be considered fully a little later. There were almost certainly atua of other kinds of whose nature we might have learnt something if it had not been for the unfortunate lack of interest of our informant in the folk-tales of the people. He told us, however, of one atua who produced the island of Tikopia from the depths of the sea by letting down a line and drawing it up, and it was this atua who gave the island its name. Whatever other atua there may have been, there is little doubt that so far as ceremonial is concerned those derived from ancestors are the most important.

The animal atua of the social divisions are clearly sacred beings. Atua are said to enter into fishes, suggesting that an animal may be only the embodiment of an atua, but the fish in one such case (see p. 319) was not one of those especially connected with a social division. When the people go for journeys in their canoes they throw food as an offering into the sea continually but it was said that these offerings are made to ancestral atua which enter into fish, especially the shark and the one.

Ancestral atua. Offerings are made to the ancestral atua on many occasions. It has already been mentioned that after

3i6 Melanesian Society

death offerings of food are made daily while kava is offered at first daily and later at intervals of six months. These offerings made twice a year are always the occasion of a feast. At the grave of a chief the kava is always poured by a chief and it seemed that a chief usually officiates at the grave of a commoner on these special occasions, though kava may also be offered by a commoner. Food is first put on the grave and then the kava is made by some man present and handed to the chief who first bends down, then lifts the kava bowl above his head and pours its contents over the grave, uttering a prayer in a low voice as he does so. John had often witnessed such a scene but the prayers or other formulas had always been said so softly that neither he nor other young men present were able to distinguish the words. The kava is poured twice over the grave which is the especial occasion of the celebration and then the man who has made the kava will pour it over any other graves in the same house saying ** la te kava " and the chief then prays again at each grave.

When the kava is finished the chief takes a piece of food and saying " O tokai ngatua "or " food of atua^' he throws it on the grave. He repeats this twice so that it is done three times, and also throws some areca nut without either betel leaves or lime, saying as he does so " Kaura tatua " meaning "areca nut of the atua.'' When the daily offerings of food are made, the people utter the name of their dead father or other ancestor and use the expressions " Tokai ngatua" and ''Kaura tatua.'' It is noteworthy that the word atua should be used in the singular {te atua) when food is offered and in the plural {nga atua) when betel is given.

Women are never present at these ceremonial offerings of kava which are made at the graves of both men and women.

When there is a desire for rain, kava is offered in exactly the same way, the prayers no doubt being especially adapted to this purpose. As in the case of the other offerings, John did not know what was said. As we have seen, a similar offering of kava is made before a boy is incised, and this ceremony was said to act {inamaui) through the kava. Kava is also offered when a chief attempts to cure a sick man and before any Tikopian undertakes a journey. It is also, as we shall see, offered when a new garden is made, at the time when the yam-crop is ready for digging, and on other occasions in connection with agriculture. It is also offered before

The Attta

cutting down the tree which is to be used to make the new canoe of a chief, at the cooking of the first catch of certain kinds of fish, and probably on other occasions.

The kava is made in the same kind of way as in Tonga. The root is chewed by some man other than the chief who is to make the offering and the liquid is strained in the fashion of Tonga or Fiji. The bowl in which it is made is called kumete. The plant is extensively grown on the island but is only used for ceremonial purposes and the kava is never drunk.

As an example of the way in which the people look on these offerings I give an account of a recent occurrence. There was a big hurricane at Tikopia not long ago in which so many coconut trees were blown down or damaged that there was a great scarcity of food. Many houses were also destroyed and the inhabitants had to sleep in the open. The people thought that the atua had sent the storm, being angry because they had not been giving enough kava, and in their turn the people were angry with the atua. To make matters worse there followed a drought ; the sun shone continuously and dried up the earth and starvation was staring the people in the face. Offerings of kava were made daily to bring rain but without result. At last the brother of the chief of the Taumako asked John if he could do anything and John prayed for rain which fell three days later so that the taro and yams grew again.

The apparent success of John led to a division among the people, some believing that the rain was due to his intervention while others believed that it had been sent by the atua to whom at last the long continued offerings of kava had become acceptable.

Plant atua. In connection with the sacred animals of the different divisions it was mentioned that these groups had associated with them certain plants called atua.

All yams are sacred to the Kafika who eat these vegetables though they do not like to see anyone cutting them with a knife. It is the top of the yam which is especially regarded as an atua. When a yam is being prepared for eatingjthe skin is always removed with a shell. Similarly, the taro is sacred to the Taumako who regard the eye of the tuber as especially the atua, and they also have an objection to the use of a knife in preparing the plant for food. The

3i8 Melanesian Society

Tafua have the coconut as their atua and again will not have it opened with a knife but use a stone. The Fangarere have no corresponding sacred object though they take part with the Kafika in ceremonial connected with the yam.

Whenever it is decided to make a new yam-garden, an offering of kava is made at the grave of an ancestor, and the people of the Kafika and Fangarere divisions join together to cut the bush and burn the undergrowth, this being done at night. Then the chiefs of the two divisions plant the first yams in a special patch which is called puke tapu. The chief of the Kafika plants the first yam, the chief of the Fangarere the second and the chief of the Kafika again the third, after which the planting is done by the rest of the people.

After the yams have been planted the yam-stick or koso which has been used to dig the ground is put in the house of the chief, to be used for the same purpose in the following year. When the yams are ready for digging, the chief of the Kafika offers kava at the puke tapu and the crop is then taken from the ground. A pudding is made of the yams first planted and after the chiefs of the Kafika and Fangarere have offered kava and prayed, they eat part of the pudding and are followed by the rest of the people.

The first taro is planted by the chief of the Taumako in a part of a garden called 7nata pupuro, having some reference to the eye of the taro. The stick used is called the koso tapu ; it is not taken away as in the case of the yams but is stuck upright in the ground close to the place where the first taro has been planted and remains there till covered by weeds. Later, these are removed when the chief of the Taumako offers kava and puts the stick in his house for use in the following year. When the taro is ready the crop is dug and taken to the chief's house where the taro first planted is cooked with others to make a pudding. The chief is the first to eat this pudding, after kava has been offered with prayer as in the case of the first yam.

The Tafua have no corresponding annual ceremonies in connection with their atua, the coconut, but the chief of this division has the power of putting a tapu on this kind of food. The sign of the tapu is the same as that ordinarily used (see p. 319), and is put by the side of six or eight trees belonging to the Tafua and by two trees of each of the other divisions. When putting up the sign the chief of the Tafua prays and

Tapu 3 1 9

offers kava. While the tapu is in existence, a period which may be extended to several months, no one may eat coconuts. When the sign is removed the chief of the Tafua distributes food and coconuts to the other chiefs, a hundred coconuts to each, the food being distributed in the form of puddings. A tapu is only imposed when the trees are not bearing well and is usually maintained till the nuts which have dropped from the trees have begun to sprout and are ready to be transplanted.

The general tapu imposed by the chief of the Tafua on coconuts is but one example of a widespread practice among the Tikopians. Any chief may tapu a special place in order that the trees may grow to a proper size before the fruit is taken and anyone can initiate such a tapu which is later confirmed by a chief. The sign of a tapu is a bough of any kind stuck in the ground near the tree. On the bough is placed the leaf of the longlong or cycas to which it is bound by a leaf of coconut. Some words are probably said when the tapu is imposed, though John did not know them, but an ancestral atua is certainly invoked, either implicitly or explicitly.

Anyone who infringes such a tapu will fall sick, perhaps with widespread pain, perhaps with swelling of the body. The sick man will get his brother to go to the man who has imposed the tapu and the latter will offer food on the grave of an ancestor to whom he will pray. When the sick man gets well, he will make presents to the man to whom he ascribes his recovery.

While a man who infringes a tapu thus receives punishment by sickness, one who steals goods not protected by a tapu is also punished by the direct action of the atua. If a man loses his coconuts and thinks that some one has stolen them, he spreads the news of his loss among the people and it is proposed to go out that night in pursuit of a fish called aku which has a long and pointed snout. Before starting the man whose property has been taken invokes his ancestral atua to enter into the fish and, if the guilty man is of the fishing party, it is believed that the aku will jump into the canoe during the attempts to catch it and will pierce with its snout the leg of the culprit. If this happens the other people in the canoe at once call out ''Moana ifo toto" " Blood down in the sea," and the whole fleet of

canoes will at once make for the shore. The injured man will cry out with the pain and this together with the general shouts of the people will tell those on shore what has happened. When the canoes are beached the injured man is carried into his house, outside which the people stand, the men beating their foreheads and the women tearing their cheeks till the blood flows. At night all cry over the man and then there is a feast in which the fish which has inflicted the wound is eaten, but neither the injured man nor his relatives nor the man whose coconuts have been taken may eat the fish and the injured man may not eat this kind of fish till he is well. The prohibition of the fish to the man whose coconuts have been taken is often a matter of great interest for it sometimes happens that instead of spreading the news of his loss a man will invoke his atua secretly so that it is only when some one is wounded that the people know what has happened. On such an occasion therefore all watch carefully during the feast to see who abstains from eating the fish\

We were told that the attitude of the people toward the injured man is not one of anger but of pity. The man who had lost his coconuts would be angry when invoking his atua, but as soon as he sees the wound his anger goes^ and the rest of the people sympathise heartily with the sufferer on account of the pain he is enduring. The resentment for a wrong done to a member of the community seems to be wholly swallowed up in the pity felt for one who is suffering.

There are many sacred places on Tikopia to which women may not go. These are called ngange and are all places where there are the graves of ancestors. There are other places where none of the common people may go, whether male or female, though they may be visited by the chiefs, and there are still other places with bones and skulls lying on the ground where not even chiefs may go. No ceremony of any kind is performed at these places at any time and John compared this with the very complicated regulations connected with sacred places which he had observed during his stay in the Banks Islands. As a stranger John had gone to some of the sacred and tabooed places of Tikopia and, as there had been no evil effects, some of the people of the island had followed his example.

^ Doubtless when a man has been wounded in this way there is always someone ready to abstain and thereby get the credit of a successful appeal to his atua. 2 See also p. 310.

Possession 32 1

Possession.

The Tikopians become possessed by the atua or ghosts of their ancestors and, as the rigorous distinction between the two classes of the people is held to apply also to their ancestors, a chief is only possessed by the ghost of a chief and a commoner only by the ghost of a commoner. A man who is possessed shakes and quivers all over, his eyes get red and he begins to shout and, when the shouting is heard by the people, all run to hear what the atua has to say. When Tikopians have gone away from the island and have not returned their friends seek to learn their fate by this means. Thus, some time ago a party of people left the island and later a man in a state of possession said that they had all been killed on an island called Taumako.

The chief of the Tafua is now frequently possessed by the ghost of a man who is believed to have perished in the Reef Islands, either on Pileni or Nukapu. Some time after the disappearance of this man and his companions, the chief of the Taumako became possessed and the people then learnt from the atua that their friends had been slaughtered and that their murderers were at that moment on their way to Tikopia. The ghost said that the strangers were ten in number and that they were to be killed as soon as they arrived. They came as prophesied and were killed. Later, when the chief of the Taumako died, the same ghost came to possess the present chief of the Tafua, the younger brother of a chief who was among those killed, and when the present chief dies it is expected that the son of the murdered man will be entered by his father's ghost. When a man is possessed the people sit down in the house and ask questions which may only be put by those of chiefly rank when the ghost is that of a chief. Thus, in the case just mentioned a chief asked, "Why did people kill you ? " and the reply was, " While I was bathing I felt the arrow in my back and I died."

"Where is the canoe in which you went ?"

" It is now in Vanikolo ; we went first to Vanikolo and while we were there the people of Pileni and Nukapu came and invited us to go to their islands, so we left the canoe in Vanikolo."

" Did they bury you properly ? "

"Yes."

" What did they wrap you in ? "

" In a mat."

This story had a sequel which may be recorded here. While John was on the island a party of people came from the Reef Islands, some from Pileni and others from Matema. When they were seen the people met, and at first it was agreed unanimously that they should be killed. John then went to the Motlav teachers on the island and asked them to intervene. They did so and said that the people had come like birds of the air ; that they also were from distant islands and that the new teaching said that there should be no murder. The head chiefs were all silent but the brother of the chief of the Taumako said, " Why do you remain silent ? This is a true and a good word which these men speak." The chief of the Tafua whose brother had been killed then said that he wished to avenge his brother. The chief of the Taumako replied that the brother had already been avenged and proposed that the people should be spared. The chief of the Tafua then consented and the strangers were received in a friendly manner and remained on the island for a month.

A man who asks a question chews betel and taking some of the chewed mass from his mouth he holds it out to the possessed man saying, " Eat," and it is eaten by the possessed man who is then ready to answer his questioner.

This possession is a very frequent occurrence and a ghost may transfer himself from one person to another as in the case of the chiefs of the Taumako and Tafua. Occasionally a man may be possessed by several a^ua, one coming at one time and another at another.

Often, and especially in case of sickness, the a^ua are invoked in order to discover the future course of events. Thus, people will assemble in the presence of a sick man and will decide to call on some a^ua, it may be the father or some other dead relative or ancestor of the dead man. The words which would be said on such an occasion are as follows :

£ laui ke kalanga tato tupuna ke ao ^ ono kiti ngamgai It is good we call our ancestor we .-* see ^ sickness pe laui pe sie siai ; sakiri atu ki tasi kenea. if good if not ? look for thither to another again.

' Mr Ray suggests that this word may mean "assemble," as in Samoa, in which case the translation of the second clause would run " we assemble to see this sickness...."

Possession 323

"It is good that we call on our ancestor ; we see this sickness if it will be good or not ; we will call upon another man."

The meaning of the last clause is probably that if the ancestor says that the sick man will not recover, the people will invoke another atua, for this, as will appear presently, is what they do.

When the atua is invoked it is not known who of those present will be inspired and all wait patiently for this to happen. When the selected person begins to shake, betel is presented and questions are put to find out whether the sick man will recover, etc. The possessed man approaches the patient and asks to have pointed out the position of the pain and he touches the place, his hand being called te rima te atua or "the hand of the ghost." Then the possessed man says that in the night it will be seen how the patient will fare; that if the pain is less, he is not to be called again but that if it is worse he is to be summoned. If the sick man becomes worse and the ghost is again invoked, the possessed man will on this occasion say definitely whether the patient will die or recover and will never come more than twice. If a favourable verdict is given, the man recovers and makes a large present of food to the man whom the ghost has entered.

If an unfavourable answer is given it will not be accepted but another atua will be invoked and if his verdict is equally unfavourable several more may be summoned before the relatives give up hope. If the second ghost gives a favourable answer, it will be at once believed in spite of the contrary opinion of the first and no further advice will be taken.

It must be remembered that this description was given by one who had a great contempt for the methods employed but it is probably a fair account. According to John the people had taken to consulting him when three or four atua had agreed in an adverse prognosis, and he had then employed remedies which he had learnt from his parents, in many cases with success^

In this procedure there is nothing which can be regarded as treatment unless the laying on of "the hand of the ghost" is to be regarded in this light. It is rather a method of divination to discover whether the invalid will recover. The therapeutical measures of the people seem to be of the

• See an account of these remedies in an Appendix to this chapter, p. 356.

21 — 2

simplest character. When a man is ill, his father or other relative goes to the chief who comes to lay the leaves of certain trees i^rauti and tantmi) on the breast of the patient while he offers kava and prays. The chief pays two visits on successive days and, if his efforts have not then been successful, another chief is called in. When the man recovers he pays the chief who is held to have brought about the cure, and the patient bathes in the sea. Here again the people have often consulted John when the efforts of the chiefs have not been successful.

There are a few simple remedies which do not involve the presence of a chief or of special offerings. Cuts are bathed in hot water and leaves of the kavakava shrub, a plant resembling the kava, are bound on the wound. They are changed the next day when the wound is bathed with hot water and leaves are then put on again and the wound is thus dressed daily till it is healed {inafy heal). For a pain in the back a hot stone is applied locally.

Little information could be obtained about the nature of the diseases prevalent on the land. Many people covered with ringworm were seen when we visited the island. According to John there is no venereal disease. After the return of five men who had been away in a labour-vessel, many people had catarrh and shortness of breath and there were many deaths. There are plenty of mosquitos on the island and fever is said to occur.

Greetings.

The Tikopians attach great importance to certain forms to be observed, not only in the intercourse with chiefs, but also in the everyday behaviour of the ordinary people towards one another. John Maresere had been so long on the island that he had fully acquired the idea that proper respect should be paid both to superiors and elders and he often became almost speechless with indignation at the behaviour of the Melanesian boys on the Southe7'n Cross who would approach and sit down without asking permission or would pass by without a word.

In Tikopia a man who wishes to pass another who is sitting down will call out as he approaches "(9 mata',' mata meaning eye, and if the answer comes '' Poi, erau" "yes,

Plate XVI

Fig. I. Showing method of carrying

Fig. 2. Group of children.

Occupations 325

all right," he goes by. If there are two or more people sitting down, he will say ''Oto, mata' and will be answered as before. Several men going together ask leave to pass in the same way. A man walks by upright but a woman who is passing bends down slightly with her hands between her thighs.

If a man went by without the proper greeting it would be supposed that he was angry and he would have to return and sit down near those whom he had passed. If he explained that his behaviour had been due to carelessness, he would be called a fool and his apologies would be accepted. The greeting when passing a chief has already been considered, the man in this case going by on his hands and knees.

When a man goes to visit another at his house he calls out the name of the occupant when at some distance and if the latter is at home the visitor is invited to enter. If there is no answer the visitor will wait for a time or go away and if it is found that only the wife is at home he will go away. Only near relatives may enter a house without invitation. It is perhaps significant that we were told of this refusal to enter the house when the wife only is at home immediately after we had been inquiring into the subject of adultery.

Occupations.

There is a definite division of labour between men and women. The cooking is nearly always done by the women and a man would certainly never cook in the house. The women make the bark-cloth though occasionally men may help, and it is the women who make the very fine Tikopian mats. Most of the garden-work is done by women, the men only performing the more arduous operations, such as cutting down trees when preparing a new garden, and the women carry most of the garden-produce to the home though men may help in this work. It is also women's work to get firewood. These and other objects are carried on the shoulder as shown in PL XVI, Fig. i, and neither men nor women ever carry anything on the head. House-building is the business of men and they catch all the fish except certain small kinds which are caught with hand-nets by women. Certain male occupations are not followed by all the men of the island but belong to certain limited groups. The crafts so limited

are those of canoe-building, turmeric-making and tattooing, the special craftsmen being in each case called tufunga.

Canoe -making. A man who wishes for a new canoe goes to two tufunga who get other men to help them. The tufunga are taken to the tree which the future owner has chosen and if they agree that it is suitable it is hewn down. It is of the kind called fetau which is the same as the paivura of Mota and it is hollowed out, iron adzes being now used for this purpose. While the canoe is being made the tufunga and other workers are fed by the future owner of the canoe who prepares for them three meals a day, catching fish for them daily and giving them coconuts to drink and betel to chew. When the canoe is ready, the makers are paid and then the canoe is launched, the builders going in it first without the owner. They go to catch flying fish and when they return they roast their catch and eat it. The tufunga then take the canoe to their village and polish it with coral and when this has been done they make a feast for the owner who in his turn makes a feast for them. Both feasts or as they should perhaps be more properly called, distributions of food, take place on the same day so that what really happens is an exchange of food. The owner then goes for the first time in his canoe taking the tufunga with him. They catch flying fish and return to the village of the owner where they beach the canoe and the fish which have been caught are divided out by the owner among those who have accompanied him.

When a chief desires a new canoe he chooses two of the best tufunga and twenty men and when they go to the chosen tree all sit round it while the chief makes an offering of kava. After this the proceedings are the same as for the canoe of a commoner.

The canoe-makers may belong to any division and the occupation is hereditary, the art passing usually from father to son, though others may learn the craft. The Tikopians make very good models of canoes and a man who wishes to become a canoe-maker will make such a model. If the people approve this model the beginner will be encouraged to make the real article which he must do alone in the bush, and if it is well made he will probably be asked sooner or later to act as a tufunga and make one professionally. There is no formal instruction and nothing of the nature of apprenticeship.

Plate XVII

Fig. I. Tikopian canoe.

« 9

LA

Fig. 2. Showing method of paddling canoe

Plate XVIII

Fig. I. Canoe and canoe-shed.

Fig. 2. Group ot canoe-shedi

Occupations '^ttj

When a man is making a canoe on trial he has to cut down the tree single-handed and the labour of doing this is so great that it deters all except those who are very much in earnest. The paddles i^foe) of a canoe and the bowls for kava are also made by the canoe-makers.

A canoe is called vaka ; its body is ora, the part of the outrigger connected with the body is the kiato and the float of the outrigger is called ama. This float lies underneath the kiato to which it is connected by pieces of wood called fuluki, well shown in PI. XVIII, Fig. i. The deck at each end is called ptike ; the bow is mata vaka, the stern amnuri, the cavity of the canoe riu. If the canoe is small, a plank called fono is added to the dug-out structure to raise the sides above the level of the water.

Turmeric-making. This occupation like canoe-building is hereditary but may be taken up by others who learn the necessary procedure. The makers, also called tufunga, may be canoe-makers as well or the two occupations may be separate. The dye ango is made from the turmeric root, also called ango, in a special house near fresh water and the dyers live there while engaged in their occupation.

The root is scraped into a number of bowls into which fresh water is poured and left for an hour or two. When the sediment settles the water is poured away and the pigment put into smaller bowls and heated. Especial care is taken to cover the hot stones of the oven with many layers of leaves so that the bowl shall not be burnt and when the oven is opened, the turmeric has become hard and the cake which is turned out is wrapped in bark-cloth. When it is required, a piece of the hard turmeric is powdered in water and heated together with coconut-cream^ and then rubbed on the cloth which is to be dyed.

Whenever the dyers are working they must eat hot food and may not eat mammy apples or ripe bananas. While occupied they may not cohabit with their wives, there being no similar restriction during canoe-making. Sometimes the turmeric when prepared will be found to be watery and then they will inquire whether anyone has cohabited with his wife and when, as usually happens, some one confesses, the turmeric is thrown away and they begin again. Sometimes it will be found that some one has eaten a mammy apple or cold food

• See note on p. 312.

and then again the turmeric will have to be re-made. We were told that there can be no mistake ; "if anyone has done what he should not, the turmeric will turn out badly." If it is not of a good colour, the turmeric is not used as a dye but is eaten and is regarded as a good food and it is also used sometimes as a medicine. As has been already mentioned, the mixture of the turmeric with coconut-cream makes the dye of a much redder colour than the pure substance.

Tattooing. The third class of people who are called tufunga are the tattooers. The occupation may be combined with canoe-making and dyeing or may be practised separately. The substance which is used is made from the fruit called fua. This is gathered when it has dropped on the ground and burnt in a kind of oven consisting of a flat stone, one side of which rests on the ground while the opposite side is supported by two pieces of wood, all except one side being covered with leaves to keep in the smoke. The fruit when it is burnt makes a soot which collects on the stone so that it can be scraped off and wrapped in leaves. The powder thus obtained is called refu and is kept in a hollow bamboo.

A man chooses the parts of the body where he wishes to be tattooed but the pattern is chosen by the tufunga. Some of the powdered refu is put in a coconut cup and mixed with water so as to make a paste, being stirred with a stick called nun. The man to be tattooed lies on a mat and the operation is performed with a pricker called matau made from the bone of one of the gulls called atoko and lofa. The bone may be taken from any part of the bird and is fixed between a split reed as a handle so that the pricker can be held vertically on the skin while the handle is struck with a light hammer made of the midrib of the leaf of the coconut. When the tattooing is finished the tufunga is paid. The same men tattoo both men and women but the patterns are different in the two cases.

The Tikopians never tattoo the thighs and hips. The pattern on the forearm was said always to be the same, consisting of a very conventionalised mango tree with lines supposed to represent the urunga or wooden pillow. John Maresere had on the chest what looked like a very conventionalised pattern surrounded by fish but it turned out that this was an imitation of an iron saw which some one had brought to the island. In the old days there would have been the

Occupations 329

same pattern here as on the forearm. On the backs of his hands John had eight rows of small elliptical patches and on his back a number of vertical lines called lake 0 tetua. Figures of fish were tattooed on the forehead and cheeks and lofa birds at the outer angles of the eyes\

The forearms of women have the same pattern as those of men but the lines on their backs are transverse and there are also simple lines drawn from the ears along the jaws and a median line from the lower lip to the navel.

Cloth- and string -making. Other occupations are those of cloth- and string-making. The bark-cloth used by the Tikopians is of one kind only and is made from the bark of a tree called mdmi and the name of this tree is also used for the cloth. It is made, usually by women, by beating it out with a hammer after it has been put in the sun to dry for two or three days. The hammer used is called iki and the log on which the bark is laid when being beaten tutunga. After being beaten out to the required thinness it is again dried in the sun and is then ready to be worn, either as it is or after being stained with turmeric. String is made by both men and women. The husks of coconuts are put in salt water for about two months when the outside covering can be stripped off and the fibres separated from the connecting material. These fibres are beaten in the same way as in making bark-cloth ; they are then rolled together on the thighs to make strands which are plaited to make string. The strands are called rino and the verb 'to plait' \s pik. Plaited string consisting of three strands is called kafa, and this is used for house-building, for the lashing of the different parts of a canoe and for making the rope with which sharks are caught. Another kind of string called uka is made by rolling together on the thigh only two strands and this is used for fishing lines.

Women also make a very fine cord from coconut fibre which they use for their girdles.

Fighting and weapons.

No account of any serious fighting was recorded. In the more or less ceremonial fighting which accompanies a marriage,

^ In addition to the pattern already mentioned John had on the chest two stars which had been tattooed in Motlav.

clubs and bows and arrows are used and these weapons are also put in the canoe of a man who is being sent adrift as a punishment.

I obtained a club on the island which is shown in PI. XIX together with two Tongan clubs from the Cambridge Museum^ for comparison.

Currency.

Mats, bowls, hooks, bark-cloth and other objects are exchanged freely but there is no commodity used as a medium of exchange which can be regarded as money.

FisJiing.

The Tikopians have many methods of fishing. To catch flying fish, called 7na/olo, they go out in canoes at night when the moon is not up. Often from 20 to 40 canoes go out together and in this case they form three lines of which that in the middle goes straight forwards while the two lateral lines diverge at first and then converge so as to crowd the fish together. In the bow of each canoe there is a man with a net and in the middle of the canoe another holding a torch while the rest paddle. The fish may be caught when they are seen in the water or as they are flying, and it was said that sometimes a man will catch a fish when it is flying over his head. The fish do not appear when it is light so that as soon as the moon rises all go home. This kind of fishing is very much practised and every night torches are to be seen moving here and there over the water as the people fish. The fish called aku with a long snout is also caught with a net at night. When they see this fish they call out '' te aku, te aku,'' and put the net in the water in front of the fish which goes into it. The nets are at the ends of long bamboos and several men in the canoe carry them so that, if the fish escapes the man in the bow, others behind him may catch it. It is this fish which inflicts the punishment for stealing which has been already described. Another fish, the ika tapu, is caught with a hook baited with pieces of flying fish. In spite of its name this fish is not regarded as in any way sacred.

• I am much indebted to Baron A. von Hiigel for allowing me to represent these clubs.

Plate XIX

c

u

Fishing 331

The shark is caught with a hook baited with flying fish, the whole fish being put on the hook. As soon as a shark takes the bait, a man prepares a loop of a rope, and the animal is drawn to the side of the canoe so as to allow the loop to be put over its head. When in place the loop is drawn tightly and a man then clubs the shark on the head. By this time all are panting with the exercise the shark has given them and they rest for a time and then take the fish into the canoe if it is not too large.

The hammer-headed shark, called matai taringa or " eyes and ears " because it has eyes in the usual position of ears, is also caught with a hook and then two men pull it on the outrigger of the canoe by the two sides of its head before it is killed. They put the kiato or outrigger in the mouth of the shark so that it bites it and then kill the animal by clubbing.

Two other large fish, the varu and the fangomea, are also caught with a hook. John was very emphatic when he told us that these big fish could only be caught because the Tikopian canoes are large and strong. He said it would be quite impossible to catch them with such canoes as those of the Banks Islands. At present iron hooks are used and John could not say what kinds were used before these were introduced. One of the most highly prized fish of the Tikopians is the atu or bonito which is caught with a turtle-shell hook baited with a white feather. They put a bamboo rod carrying line and hook in the stern of the canoe so that the feather floats on the surface of the water. When they come near the shoals of bonito the man in the stern moves the rod so as to draw the feather to and fro on the water. The bonito only visit the island during the first three months of our year.

The people angle for small fish with a hook from the beach or from rocks, and they also use a long net called fangota which is held by two men, one at each end, while other people stand round in a circle and walk towards the net driving the fish before them. Fresh-water fish are caught in a large lake on the island by means of a net called hipenga with a triangular opening. This is set in the water and left there long enough for a man to chew betel. When he returns he finds fish in the net which have usually been killed in their attempts to escape through the meshes.

Prawns called ika viki Idki and small fish are caught by women with little nets called Mti. The women lift stones and

catch the prawns as they dart out. These are cooked after being wrapped in the leaves of a tree called teti.

With the exception of the net used for fresh-water fish they do not use fish-traps nor do they spear fish. Sometimes fish are poisoned with the crushed stem of a creeper called nauari, which is mixed with sand and put in the pools left when the tide recedes and any fish in the pool soon die.-

Certain customs are observed before some fish are eaten. Thus, those who have caught a shark go to sleep as soon as they reach the shore while others cook the fish and food is collected for a big feast. When the feast is ready the sleepers are awakened and usually kava is offered by a chief. The man who has caught the shark then asks the chief to distribute the flesh of the animal and this is done but only to the members of the division of the successful fisherman, the other people having to be content with the rest of the feast.

A man who catches the fish called var'u does not eat it but must present it to the chief of his division receiving a present in return. If two varu are caught, both are given to the chief, but if there are three, two go to the chief and the third is kept by the man who caught it. Bonito are also given to the chiefs. At the first catch of the season the chief offers kava before the fish are cooked and again when the oven is opened and in this case the catch is distributed to the chiefs of the four divisions. After this first occasion the fish may be eaten by those who catch them.

Only unmarried women may go out in the canoes with the men when they are fishing but married women are allowed to go in canoes when a journey is being undertaken.

Fire-making.

Fire is made by the groove method and John said that the method is exactly the same as that practised in Motlav, Uvea (Wallis Island) and other places he had known. Fire is never made by women but only by men who sit on a log called sika in which the groove is made by means of a smaller piece of the same kind of wood which is called kaunatu. Several kinds of wood are used, the best being those called fau and varovaro. A man takes {to) the kaunatu and rubs it to and fro {sika), making the groove {I'uo), in which there collects dust {ef). When the wood begins to smoke

Plate XX

^s. ■■

. i

Plate XXI

Fis^. I. Makiiii^ coconut-cream.

iis(i|ii,m wiinu

Food 333

the movement is made more rapidly i^faka pore) and when the dust begins to glow the people say ''wotu koteafir "the fire has come." The maker then takes the dried husk (puru) of a coconut and drops the lighted dust on it {furi) and blows {iri) into flame. "He blows the fire into flame" is '' Iri koteafi keeka!' This is the only method of making fire which the Tikopians possess and they do not seem to share the desire for matches almost universal in the Pacific, doubtless because any obtained from passing vessels have been so soon exhausted that they have never come to look on them as objects of importance.

Food, etc.

The chief kinds of food of the Tikopians have been already incidentally considered. The staple articles of diet are fish, coconuts, yams, taro and sago, together with banana and mammy apple. There are no pigs on the island and it would seem that the turtle has only been eaten recently. There are fowls {kio) but many Tikopians do not eat them. Turmeric is eaten and is said to be a good food. A favourite addition to food is coconut-cream [sinu) made by grating the white of coconuts and expressing the juice. PI. XXI, Fig. i, shows a man straining the cream by squeezing the grated coconut in coconut fibre.

The coconut shells used as water-vessels and the method of filling them are shown in PI. XX, Fig. i.

The most remarkable feature of the vegetable food of the island is the prominent place of the mammy apple (Papaia) which is the only kind of food allowed in mourning and yet it is a recent importation, having been introduced by the same whaling vessel which brought the first cat to the island.

Betel-mixture is used by the people largely and the areca nut ikatira) and betel \q2S [pita) must be very plentiful. The lime, called kapia, is kept in simple undecorated gourds, and the elderly chief of the Taumako whom I saw on my visit prepared his betel-mixture in a cylindrical vessel with a spatula, exactly in the same way as is done by elderly men in the Solomon Islands. It seemed quite clear that the kava which is used so extensively in ceremonial is never drunk.

Mr Durrad's Account.

The following is an account based on material sent to me by Mr Durrad which includes data derived from an analysis of nine pedigrees.

At the north-east end of the island there is a large lake surrounded by hills from 600 to 1000 feet in height, and the region in the central part of the island south-west of the lake consists of level and swampy ground. The whole island both on the level ground and the hills is more or less cultivated, the fruit and coconut trees numbering thousands, and there is no uncleared forest.

The island is divided into two districts, called respectively Faea and Ravenga, and between the people of these two districts there seems to be an inveterate feud. At Tufenua there is a big rock which marks the division between the two districts. The villages are close to the shore, with the exception of three which stand a little way back, two of these being situated near gaps in the hills surrounding the lake. The names of these villages are given on the sketch-map on p. 335 made by Mr Durrad. On this map are given the names of the social groups which inhabit these villages and it will be observed that there seems to be no obvious connection of these groups with the two districts. The Taumako predominate largely in the Ravenga district but they also inhabit most of the villages on the other side of the island. That there is no connection at present, whatever may have been the case in the past, seems to be indicated by the fact that the late chief of the Tafua lived at Namu, while the present chief has his home at Potimatuang. No case is given of a village inhabited by the Fangarere.

Certain people of the island are known to be of Tongan origin, being descended from men who came to Tikopia from Tonga some generations back.

The atua of the four divisions.

The following information concerning the atua or sacred objects of the four divisions was obtained from Ernest Wirit, a Motlav teacher, living on the island. Mr Durrad gives the atua as those of the chiefs, but most of them also probably belong to all the members of each division.

Plate XXII

View from the hills looking to the south-east.

"^ t«

r

<S)

(!)

<u

!C

^Ti

o

!fl

u

x;

'3 !30

e

V

The atua of the Kafika are : —

Te riringo, a very large black sting-ray, so large that one is said to have been known to catch a man.

Te kuku and te panoko, two small fish.

The crocodile, small specimens sometimes finding their way to Tikopia.

On the arrival of the Southern Cross off Tikopia on April 27th, 1 9 10, a canoe belonging to the chief of the Taumako kept too close to the steamer and was overturned. The people some way off saw the accident and saw at the same time a huge black sting-ray. They shouted " Look ! Kafika is coming here." Then they threw four coconuts into the sea and the sting-ray went away.

The atua of the Tafua are : —

A very large eel of the lake. It is called "TV atua Tafuay The ceremony of installing the present chief of the Tafua division took place in the sacred place of Tikopia beside the lake when the new chief made offerings to the eel.

A large eel of the sea.

Te peka, the flying fox.

Te vai, a sting- ray.

A kind of taro which is eaten by anyone, but the Tafua chief may not look at it. When going past it he will turn aside his head.

The sword-fish (garfish ?). The chief and his eldest son may not eat it, but his younger children may do so.

The distinctive decoration of the chief is the leaf of te katafa, a sort of palm, twisted and worn round the neck.

The following are atua of the Taumako : —

Te peka, the flying fox.

Te riringo, a red sting-ray.

A white eel of the sea.

A black sea eel.

Te pttpumatao, a long fish (the Mota utu).

Te raparapa, a small fish (the Motdi garegare).

Te rupe, pigeon.

Te karai, a rail (the Mota matikd).

On June 8th, 19 10, when the Southern Cross was expected to arrive, the following incident occurred. Ernest Wirit was living as the guest of the Taumako chief, who sent him to get

Oj^erings to Atua 337

some pilaka (a kind of coarse taro), coconuts and taro. He and the men with him brought them in a canoe over the lake from the gardens. They brought five baskets of taro and besides these one man carried five taro roots on his shoulders to the chief's house. The chief said " Look at the sea, how rough it is. I am going to sacrifice (faikava, i.e. to make kava) to the red sting-ray (te riringo) and the black sea-eel that they may take care of the Southern Cross and the Bishop." The taro roots were cooked in the oven and then pounded in a food-bowl and coconut-cream was poured over them. The bowl was then carried into Raniniu, the house of Paeraniniu, the chief's son. The chief sat in his own place near one end of the house, and the bowl of food was put in the middle of the house, the rest of the people present sitting at the other end. Near the bowl was placed some turmeric, scented water, areca nut, betel leaf and lime. The chief chose Paevatere to take the offerings and this man wrapped a small portion of each object in a leaf which he gave to the chief who held it above his head twice and prayed in a low voice so that no one could hear the words. The packet was put at the end of the house beneath a canopy of yellow bark-cloth and behind a fringe-like screen of very young white coconut leaves. This packet was for Riringo, the sting-ray, and another packet was offered with similar procedure to the black sea-eel. The chiefs second son, Paetearikitonga, then made kava which Paevatere took to the chief who held it over his head twice with prayer and then poured it over one packet of food, a second bowl of kava being poured on the other packet. The chief then pounded up betel leaf, areca nut and lime and, after holding the mixture above his head twice with prayer, he placed it with the other offerings. After this every one remained silent for a long time till, on food being brought in to them, the ceremony was at an end.

The atua of the Fangarere are : —

Te fono, the turtle.

Te varovo, a. long fish (the Mota one).

Te tafora, the porpoise (?).

Te punapu, a black fish with a red tail.

Te iofa, a fish with a large head which moves like a flying fox flapping. It is a very important atua. Whenever the chief wishes to invoke its help he lays down a piece of

bark-cloth on the floor of one side of the house, this cloth being spoken of as the property of the atua. It remains there for five days and the chief makes offerings of food to the atua. After the five days are passed the chief hangs the cloth up again. There are several (about ten) pieces of bark-cloth hanging in the house. Sometimes the chief smears one with red turmeric and sometimes he uses only the plain cloth. The offering is made to induce the atua to give back the health of a sick man whose life is supposed to have been taken away. Health and life are synonymous ideas.

The chief power of the Kafika people lies in the possession of a sacrificial stone of great virtue which lies at Potima- tuang (see map). It is buried in the ground and stones are put round and over it to protect it. Offerings are made to it on important occasions when women and children remain within doors during the ceremony. A man is sent to clear the place of weeds which have grown about the stone. Food is cooked in all the houses of the Kafika division throughout the island and when ready this is carried to the stone which is unearthed by two men, put in a basket, taken to the beach and washed and brought back again. All the secondary Kafika chiefs^ bring turmeric and scented water and each takes the stone in turn in his arms and rubs it with the turmeric and the water. Each chief has five pieces of bark-cloth smeared with turmeric and scented water and on these he carries the stone while anointing it. Food and kava are brought which the chief lays upon the stone with prayer. If there be sickness to be cured, he says :

" Wake up ! See, we are all sick. Give back good health to us." If there be a famine, he says, "Wake up! See, a famine has struck the land and there is no food for us to eat. Give back food to us that we may live and not die." If there be a drought, he says, "Wake up! See, the sun is great. Give rain to us." If there be long continued rain, he says, " Wake up ! See the land that the rain is great. Give back to us the sunshine."

When the stone is buried it is wrapped in all the cloths that the chiefs have brought^

• Te patio ariki (see p. 340).

2 Mr Durrad's informant, Ernest Wirit, had not seen the ceremony of offering to the stone. It was described to him in the middle of the night by a man who put his hand over his mouth in case anyone should overhear him.

Plate XXIII

Figr. I. Sacred house of Fans^arere.

Fi.Lj'. 2. SliDolinL; lisli.

Sacred Stones 339

The name of this stone, which is regarded as male, is Te atua takaretoa. It is spoken of and treated with great reverence as a Hving being. There is also a female stone, called Te vatu mai'u, at Ravenga, the district on the other side of the island. The Tikopia legend of creation is that two atua named Tafiki and Metikitiki were fishing from a canoe. The latter sat in the bows of the canoe and paddled and the former let down his line and drew up both the stones together from the sea. The stones gave birth to Tikopia which is regarded as the child of their union. Afterwards the stones were separated and taken to different parts of the island. Now the male stone is at Faea and is the "god" of the sea and the landing place\ for Faea is on the lee of the island. The female stone is the "goddess" of the garden-ground. It is in the possession of Paevatmaru'^ and is buried in the earth beneath the floor of his hut. The method of offering is the same in both cases.

The two atua (the stones) wanted an axe with which to make a canoe. They went to the beach and found four clamshells, two of which they made into axes and the other two they put in a house. With the axes they made two canoes. They prepared food and consecrated the canoes to themselves. The canoes were called te vaka tapu (the sacred canoes). The sacred axes are still in possession of the chiefs who pass them on to each other when they wish to make offerings. The two other shells are said to be in the sacred house of the Kafika chief, but may be taken by another chief if he wants them for ceremonial purposes. The two atua (the stones) also gave birth to a boy and a girl who married and so the people of Tikopia began.

In PI. XXIII, Fig. I is shown a sacred house of the Fangarere situated in the most sacred part of Tikopia. It contains some mounds where chiefs have been buried and a few strips of bark-cloth suspended from the roof The chief of the Fangarere " sacrifices " in this house first and then does so again in a larger house near by. Mr Durrad suggests that the carving on the roof represents the iofa, the most important animal atua of the division, which is described as moving like a flying fox.

• Mr Durrad does not state definitely that this stone is the atua takaretoa used in the ceremonial described above but this would seem to be the case. 2 It may be noted that the man takes the name of the stone.

Chiefs and other ranks'^.

Each of the four divisions has a chief who takes the name of the group over which he rules, preceded by the syllable Pae which is the general prefix of male names ; thus, the chief of the Kafika is Pae te ariki Kafika. The chief is succeeded by his eldest son unless he is too young, in which case the brother succeeds. Thus, when the chief of the Tafua, who comes first in Pedigree IX, was killed in Santa Cruz, his son was too young to succeed and the younger brother became chief. The latter died in May, 1910, and was succeeded by his nephew, the son of his predecessor, who had been too young to succeed when his father was killed.

The brothers of the chief are called te paito ariki (literally, "the house chief"). This is a title which is not transmitted to the son of the paito ariki, but passes on to the son of the reigning chief. The paito ariki sees that the orders of the chief are carried out and his word is law to the people.

Another rank, that of te pure matua, would seem to belong to the heads of "families'," for the dignity is hereditary, a man being succeeded as te pure matua by his eldest son except when, owing to his youthfulness, his place is taken by the brother. If, on the other hand, there is no brother, there will be no successor till the son becomes old enough.

Those called te pure matua act as counsellors and protectors of the chief and it is their duty to provide him with food, each doing this for a day in turn, the chief giving a present in return. The larder of a chief is always well stocked and, if there is more than is wanted, the chief will give the excess to the pure ?natua.

Other people will also give food to the chiefs and one who sees anything especially choice in his garden will tell his family not to touch it but reserve it as food for the chief.

It would seem as if yet another differentiation in rank were taking place in Tikopian society. A man named Paefakofe of the Taumako (see Pedigree X), though not one of the four chiefs, is yet considered the most important man in the island. The reason for this is that in his youth he killed all the crew of a vessel which came to the island and was

• For a full account of the chiefs by Mr Durrad see Southern Cross Log (Melanesian Mission), 191 1, xvii, 75.

' For the sense in which this term is used, see p. 344.

therefore recognised to be an exceptional man. He acts as a kind of judge in the island. In the matter of putting strangers to death he can veto the decision of the four chiefs and it was owing to his intervention that John Maresere and his companions were hospitably received. Paefakofe is a tufunga and has already instructed his son, Paevatere, who shows great ability and is acquiring some of his father's authority. The family of Paefakofe is one of those which is regarded as Tongan (see pp. 334 and 355).

Relationship.

The following account is based on data collected by the genealogical method, i.e. on the terms which are applied to one another by various persons in the pedigrees collected by Mr Durrad : —

The father, his brothers and the husband of the mother's sister are all called pae^ and also toku tamana or toku niana, the own father being distinguished as toku tamana maori or " my father true."

The mother, mother's sister and the father's brother's wife are naue or naive. The mother's sister is also called toku nana and it is probable that this term is used for the true mother. It is probably a corruption of tinana]us,i as tamana becomes mana.

A child is called tama or foasa, both words taking the possessive toku.

Tama is used by a woman, i.e. by the mother and her sisters, etc., but foasa appears to be used by both men and women, though it is possible that it is properly a term used by men only.

Brothers call one another in general taina while the reciprocal term used between a brother and sister is kave. There are, however, special terms according to age, the eldest of a family being called te rumatua, the youngest te toetai, and the intermediate members te roto, " the inside." Brothers and sisters are distinguished by the words iangata dindfajine ; thus, the eldest brother is te rumatua tangata, the eldest sister te rumatua fafine, and so on.

The terms taina and kave are used for cousins of all kinds, i.e. for all those of the same generation with whom

• This is the vocative formed by the addition of e to the word pa.

kinship can be traced, including both categories of cousin. A distinction is made, however, between these two categories ; the children of the father's brother or mother's sister are called toku tama (or kave) fakalau or "my brother (or sister) make good," while the children of the mother's brother or father's sister are called toku taina (or kave) fakapariki, " my brother (or sister) make bad."

The mother's brother is called tuatina and he calls his sister's son iramutu.

The father's sister is mesakitanga, though she may also be called nana like the mother. She calls her brother's child tama.

There is some doubt whether the husband of the father's sister is called pae or tuatina, but the wife of the mother's brother is certainly naue.

All kinds of grandparents are tupuna, those of different sex being distinguished by the addition of the words tangata or fafine. The grandfather may also be C3.\ed putangata and the graindmother purine.

Grandchildren are called makupuna, though they may also be called tama, thus classing them with the children.

The husband is toku matua or is addressed and spoken of by name. The wife is toku nofine or is addressed by name or she may be spoken of as fine followed by the name of her husband. As we shall see, her personal name becomes on marriage the same as that of her husband but with a different prefix.

The father of husband or wife is tamana fongoai or pae fongoai, but is often spoken of simply as pae. Similarly, the mother-in-law is toku nana fongoai or simply nana or naue.

The consort of son or daughter is Q.2^^d fo7igoana or tautau pariki or may be addressed by name.

The wife's brother and sister's husband (m.s.) call one another tangata or ma and two sisters-in-law call one another fine or ma.

The four relatives of this category of different sex, i.e. the wife's sister and sister's husband (w.s.) on the one hand, and the husband's brother and brother's wife (m.s.) on the other hand, call one another taina, the word normally used between brothers, i.e. this word when denoting relatives by marriage is used between those of different sex. This term is also used by the husbands of two sisters and probably by the wives of two brothers.

All these terms are used very widely ; Mr Durrad obtained the terms applied by a man to the members of several pedigrees with whom no direct connections were traced and found that in most cases definite terms of relationship were employed. The classificatory principle evidently applies very widely in this island.

There are certain interesting relations between crosscousins who call one another taina fakapariki. They may not strike, injure, or speak evil of one another though those who call one another taina fakalau may do so. A child will be told by his father's sister never to strike his taina fakapariki, i.e. the child of the father's sister, and if he does so, the father's sister will call on her atua to injure the hand of the child. A sore will form and when it gets large the father of the child will pray to the atua to desist from injuring his child. The atua will appear to the father's sister in a dream and say that the father has forbidden him to continue to produce the sore. The father's sister then prays to the ghosts of her father and mother, calling them pae and naue, and they will appear to her in her sleep and ask what she wants of them. She will tell them to make the hand of the child still worse. The father, on the other hand, will pray to them to desist from hurting his son and the ghosts will be angry on account of the contrary requests made to them and will make the hand of the child worse. Then father and child will cook food and take it to the father's sister, and the child will say, " Mesakitanga ! Make my hand well again. Blow on my hand to make it recover. I am sorry I hurt your son." Then the father's sister will make an offering of food to her father and mother and pray to them to heal the child and they will again come to her in a dream and do what she asks.

If the child dies, there will ensue a quarrel between the father of the dead child and his sister. The father will curse his sister and forbid her to come to his house or gardenground. If the quarrel continues till the death of the father's sister, her ghost will kill all the remaining children of her brother, or she may spare one so that the family may be continued. When, however, the child of the father's sister dies, i.e. the cousin who was originally struck, his ghost will want to kill even this remaining child, but he will meet the ghosts of his father and mother and of his grandparents who will

embrace him and kiss him in the Tikopian fashion and beg him to desist from revenge, and he will respect their wishes.

A doubtful point in this story is the nature of the original atua who produced the sore. Mr Durrad thought it might be one of those animal atua especially connected with the social group but, if so, it seems clear that later the matter came entirely into the hands of the ancestral atua.

It is noteworthy that there is a similarity in the names for the cross-cousins and sons- and daughters-in-law who are called tautau pariki. If people are joking and the son- or daughter-in-law of one present comes, some one will say,

"5"z sei faefakata. Tautau pariki e nofo.'' " Do not laugh. The tautau pariki is here."

It is evident that the similarity of name carries with it similar ideas regarding the conduct to be observed towards these relatives.

Cousins, whether of the fakalau or fakapariki kind, are not allowed to marry ; they are considered tapu to one another.

The family as a whole is called te ganua paito, but by this is meant the family in the extended, not in the narrow, sense. Such a group may consist of a man and his brothers, their wives and children, together with the wives and children of the sons. As the group increases in number and the house becomes crowded, the family spreads into other houses close at hand. If a son marries and goes to live in a new house, some of his unmarried brothers and sisters may go to live with him and, if the brothers who marry remain with him, this house may form the starting point of a new extended family.

Married people take their names from the house in which they live. At marriage it is customary for both man and woman to take the name of the house in which they live, husband and wife being distinguished by different prefixes, and if they go to live in a new house they will both acquire new names. In connection with this custom of taking the name of a house, i one does not say, " I am going to the house of so and so," ' but simply to the house mentioning its name. A man will not say that he is going to see Paerongore but that he is going to Rongore.

Mr Durrad notes that the people have many children and the truth of this is well illustrated by the pedigrees, there

Marriage 345

being two cases of a family of nine, two of eight and several of five and six children. It would seem as if the families are growing larger in more recent generations. There is one case of twins, Vasingateva and Fonotia in Pedigree X, both children being boys.

Adoption.

Children are certainly adopted but probably not very frequently. The adoption of John Maresere and his companions has been already mentioned but this is obviously an exceptional occurrence. The only other case of which a record is given is that of the girl Tarangarua (8)\ The wife of Paemarinua is a lunatic and unable to look after her house and as Paemarinua has no girl to cook for him he adopted Tarangarua. With the exception of two strangers, John Maresere (9) and Patita (10), this is the only case of adoption in the pedigrees to which attention is called by Mr Durrad and in this case there was a definite motive.

Marriage.

The pedigrees collected by Mr Durrad give much information concerning marriage. Their analysis shows that people very frequently marry within the group ; thus, there are five marriages between members of the Kafika, five within the Tafua and four within the Taumako, while there are seven marriages between the Kafika and the Tafua, twelve between the Kafika and the Taumako and five between the Tafua and the Taumako. These figures would seem to show that members of these three divisions marry within their own group almost as often as they marry into either of the other two. The remarkable fact, however, is that in none of the nine pedigrees is there a single case in which marriage has taken place between the Fangarere and any of the other three divisions. It seems extremely unlikely that this should be a mere matter of chance, i.e. that Mr Durrad should have collected pedigrees of families which do not happen to have married with the Fangarere and it seems possible that the Fangarere form a more or less strictly endogamous group.

• The numbers in brackets following the names of persons refer to the pedigrees on pp. 358 — 62 in which the names occur.

On looking, however, at Mr Durrad's map it will be noted that in no case does he give the Fangarere as inhabitants of any village, and it is possible that this division is now nearly extinct. It is in favour of this that in a list compiled by Mr Durrad the Fangarere have only one pure matua while the other three divisions have 5, 6 and 8 respectively.

Mr Durrad notes that many people never marry. Children call the unmarried sister of their mother " toku nana laka,'' " my mother unmarried," or " nau^aka" followed by the personal name, and such a woman will help her married sister to look after her children and will take care of them when the mother goes out to work.

Marriage with the brother of the deceased husband is ^apu or forbidden. If such a marriage were to occur, it is believed that the ghost of the dead husband would injure the brother who has taken his place.

Divorce is sometimes practised in cases of sterility.

In Pedigrees VIII and IX there is a case in which a brother and sister have married a sister and brother, and in Pedigree XIII a man has married a woman who, as the daughter of his father's father's brother, would be his mesakztanga.

Death.

The body of a dead man is dressed in a new girdle of dyed bark-cloth and decorated with a necklace of leaves. A new mat and the bark-cloth with which the body is later to be covered are put on the floor of the hut and the body is laid thereon. The head is pillowed on scented leaves and the breast and head smeared thick with blood-red turmeric (see p. 312). The knees are raised so that they are bent in the position in which the body is to be interred. During the wailing the relatives come forward on their knees one by one and lay their cheeks beside the cheek or forehead of the corpse and tear their cheeks with their nails so that blood drips on the face of the dead.

Funeral songs called fuatanga are sung in a deep tone, now and then rising to a high note concluding in a wail consisting of the word seauwe^.

It is remarkable that no less than four of the persons included in the pedigrees committed suicide. It was arranged that Tofionga (5) was to marry Paefakofe (10) but being

• For a full account of the wailing see Southern Cross Log, 191 1> P- 52.

Names 347

afraid of getting sores after the birth of a child she killed herself. Tearikifono (9) killed herself on the discovery of an intrigue. The manner of the suicide of these two women is not given, but in the other two cases the victims, Paeraroifi and Nauraroifi (8) hanged themselves. The occurrence of these four cases in so limited a number of pedigrees suggests that suicide must be a frequent occurrence in Tikopia.

Names.

The pedigrees provide a mass of information concerning names. It seems clear that on marriage both man and woman change their names and assume a common name with different prefixes to distinguish man and wife, the former having the prefix pae, the word for father, while the woman takes the prefix nau, the word for mother. The common element in their names is that of the house in which they live, thus the house inhabited by a couple called Paeraroakau and Nauraroakau is Raroakau.

When a chief succeeds he takes the name of his division, such as Paetearikitafua ; thus, Paerangifure (9) became Paetearikitafua when his father died, and at the same time his son Paenokofure took the old name of his father, becoming Paerangifure. The adopted son of the present Paetearikitafua, John Maresere, married during his second stay in Tikopia and was then called Paerongore, but now that he is dead he is again known by his bachelor name and his widow is no longer known as Naurongore. The second son of the present Paetearikitafua used to be called Taumua, and in the ordinary course of events he would have become Paerangitakai, probably after the name of the house or place where the second son of the chief would naturally live. It is evident that names may be given by a chief, for this man who would naturally have become Paerangitakai was called Paerongore by the wish of his father after the death of John Maresere.

This case seems also to show that a widow is no longer denoted by the name she assumed on marriage, for Naurongore is no longer so called now that her husband is dead.

Another example of change of name is that one of the

pure matua of the Tafua named Paesau has died recently.

His son is too young to succeed to the dignity, so the brother

of Paesau, named Paenotoa, will succeed and will take his brother's name.

An examination of the pedigrees shows that names given to children have sometimes been used in the same family; thus, Paepaka (7) has called one of his sons Poature, a name which had previously been borne by an elder brother of Paepaka, now dead. The practice suggests the absence of any prohibition of the use of the names of the dead. Several names are obviously English, such as Fourpenny and Kaliko, while others such as Sereman, Mistere and Misiroos suggest such an origin. It will be noted, however, that such names occur in the present generation, and there is no obvious evidence that the Europeans who were settled on the island a century ago (see p. 353) have influenced personal nomenclature.

Dress.

The bark-cloth used as dress is made from the mami tree and PI. XXIV, Fig. i shows the bark being heated before it is stripped from the tree.

The dress of the men is shown in the photographs. The fold of the loin-cloth which hangs down in front is often long enough to be turned between the legs and put on the ground so that the wearer can sit on it. PL XXIV, Fig. 2, in which a man is loosening the earth in a taro-garden by means of a short stick, represents the fold resting on the ground so that it can be used as a seat, though the man is not actually sitting on it.

The girdle of cloth worn by women becomes useful in carrying an infant, for the child will stand on the girdle while it clings to its mother's shoulders, as shown in PI. XX, Fig. 2.

When a man dies his loin-cloth and hair are worn by his wife or sisters. The loin-cloth is worn as a twisted band round the neck (see PI. XXV, Fig. i and PI. XXI, Fig. 2) and is used as a kind of secret pocket. The cloth is worn till the woman dies or till it becomes so rotten that it falls off. The hair is worn twisted round the head as a fillet (see PI. XXV, Fig. I and PI. XXI, Fig. 2). A woman will not only wear the hair of a dead husband or brother in this way, but also any hair which these men have shaved or cut off at times of mourning, the only occasion on which their hair is cut. The hair of men is sacred and is always worn in this way. Women usually wear the hair short, and at times of mourning

Plate XXIV

Fig. I. Heating bark of tree in making cloth.

Fig. 2. Digging taru-garden.

J' I ate XXV

Fig. I. Women wearing loin-cloths round their necks and fillets of hair round their heads.

Fisj-. 2. Bovs showing method of cutting hair.

Games

so short that the head seems to be shaved, but their hair is not kept.

Some of the fishing lines of a dead man will be cut up and worn by his relatives, two young women wearing such lines in addition to loin-cloths being shown in PL XXI, Fig. 2.

People also wear the dropped or decayed teeth of elderly relatives ; of the father because he has brought in food from the garden ; of the mother because she has suckled her child ; and of the father's sister (mesakitaitgd) because she has fed the child with chewed food. Children in PI. XX, Fig. 2 and in PI. XXV, Fig. i are wearing teeth in this way.

Boys wear their hair in tufts, often reddened with lime, as shown in PI. XXV, Fig. 2. In times of mourning these tufts are cut off and worn by their sisters.

Games.

One very popular game is a kind of mock fight with sticks in which, however, no one is hit. Two parties approach one another from a distance waving their sticks above their heads with a figure of eight movement of the hands. Each party moves forward with a rapid shuffling movement till they meet and clash their sticks together. Mr Durrad could not discover that there was any definite conclusion to the game, which is played with the greatest zest and enjoyment.

A kind of single-stick is played by boys, using the midrib of the coconut leaf The point of the game appeared to consist in hitting the sticks together rather than in hitting the opponent, but Mr Durrad remarks that it may be more difficult to get through the guard than is apparent.

Another game resembled "Hunt the slipper," with a stone as the object to be sought, and it was said that the stone represented a woman. Mr Durrad suggests that the game may be a playful imitation of seizing a wife from the father's house.

Wrestling is also a favourite game.

Fishing,

PI. XXIII, Fig. 2 shows a man shooting fish on the lake with a bow and arrow, a method which was not mentioned by John Maresere.

Comparison of the two accounts.

It is evident that Mr Durrad's notes confirm the general correctness of the account of John Maresere. On two topics the records cover the same ground, viz. those dealing with relationship and with the atua especially associated with the social divisions. The difference between the accounts of relationship probably gives us a good indication of the extent to which John's record is faithful. There is not a single statement in his account which is contradicted by the information obtained genealogically by Mr Durrad. The difference between the two records is that, where John gives only one simple term, Mr Durrad is often able to give alternative terms or qualifications which distinguish certain relatives who would by John's account appear to be grouped together. Thus, to take two examples : according to John all cousins are called taina or kave, including both cross-cousins and the children of father's brother or mother's sister. Mr Durrad agrees that they are all called taina but goes on to say that cousins of one category are distinguished as taina fakalau and those of the other as taina fakapariki. Similarly, while John tells us simply that the father- and mother-in-law are addressed in the same way as the father and mother, Mr Durrad is able to give alternative terms in which the words for father and mother are qualified by the addition of the word fongoai. Though John's account is the truth, it is not the whole truth, but this is only to be expected. Indeed, it may be that the deficiency is not the fault of John but of myself; it is possible that I was content with simple terms for each relationship and did not push the matter further to inquire for qualifications and variants.

The discrepancy between the accounts of the animal atua of the four social groups is more serious, but here again we probably have to do with defect of information rather than with actual error. Mr Durrad's account is the more detailed but in three cases, those of the Taumako, Tafua and Fangarere, most of the atua given by John are included in the list given by Mr Durrad. Thus, to take the Tafua as an example, John gives the fresh -water eel, the flying fox and the turtle. Mr Durrad gives a very large eel of the lake and the flying fox together with others, but omits the turtle. It is evident from John's account, however, that the relation of

Agreement of evidence 351

the turtle to the Tafua is not so definite as that of the other two atua. The most serious discrepancy is between the two accounts of the atua of the Kafika, Mr Durrad not mentioning either of those given by John. It may be noted, however, that Mr Durrad's information on this subject was obtained, not directly from Tikopians, but from one of his Motlav teachers, and it may be that both accounts are incomplete. It is possible that each division has many atua and that one man has given examples which the other has omitted.

Other minor discrepancies may be mentioned. The two accounts differ in their record of the position in which the dead are buried. Among the modes of fishing John did not mention shooting with the bow and arrow, but Mr Durrad has taken a photograph of a man using this method to kill fish in the lake {see PI. XXIII, Fig. 2). Further, according to John the bark of the mami tree is put in the sun to dry before being stripped off to make cloth, while Mr Durrad's photograph (PI. XXIV, Fig. i) shows the trees being heated over a fire before being stripped. In other cases where there is disagreement it is probable that we have to do with changing conditions (see below). Where the two accounts cover the same ground the agreement is sufficiently close to indicate the general trustworthiness of John's record, though there can be no doubt that it is little more than a skeleton compared with the full body of fact which would be obtained by a thorough investigation.

Two previous accounts of Tikopia have been given by Peter Dillon^ and by Dumont d'Urville and Gaimard^ and a few features of their records may be mentioned here, partly to indicate the general resemblance with present conditions, partly to suggest that certain changes have taken place during the century which has elapsed since these accounts were written.

One important difference between the present day accounts and those of the early writers is almost certainly due to modification in custom. Both Dillon* and Gaimard^ stated that

• Narrative of Voyage in the South Seas to ascertain the fate of La Pdrouse^s expedition^ London, 1829, Vol. 11, pp. 112 et seq.

2 Voyage autour du Monde et a la Recherches de la Pe'rouse, Paris, 1833, Tome V, pp. 108 et seq. ; also the journal of Dr Gaimard, published in tliis volume, pp. 304 — 312.

3 op. cit. p. 134. * op. cit. p. 308.

women were more numerous than men, Dillon going so far as to estimate the number of the women as treble that of the men. According to Dillon all the male children of each female, except the two first, are strangled the moment after their birth, but three Englishmen found on the island by Dumont d'Urville^ denied this, while Gaimard speaks of the number of children as varying from three to eighth

The account of John Maresere is that the family is limited in size to four, any beyond this number being buried alive as a general rule, and the only indication in his account that boys rather than girls are done away with is that, if four girls are born in succession, one or more of them will be killed in the hope that succeeding children may be boys. Mr Durrad, on the other hand, not only states that the people have large families, but his pedigrees show families of eight and nine children, while families numbering five and six are frequent. Further, there is an obvious excess of males, 42 families recorded in the pedigrees showing a proportion of 96 males to 55 females, the excess of males being present even in the earlier generations of the record.

There can be little doubt that we have to do here with change in custom. Both Dillon and Gaimard can hardly have been mistaken in observing a numerical superiority of the women and Dillon agrees with John Maresere as to the existence of infanticide, though from the account of the latter it would appear that, with one exception, boys and girls are equally the victims of the practice. The actual genealogical evidence would seem to indicate the disappearance of infanticide if it were not for the great excess of males, which may even be taken to indicate that female infanticide has taken the place of the ancient killing of male children, but we know so little about the possibilities of disproportion of the sexes due to purely biological causes that we should hesitate before drawing this conclusion.

The evidence concerning polygyny also indicates a change in the proportion of the sexes. Both Dillon^ and Gaimard* speak of plurality of wives as a custom of the island, the latter saying that the number is limited to four, though from another part of his account' it would appear that at the time of his visit only one of the four chiefs had more than one wife.

1 op. cit. p. 113. 2 op. cit. p. 309. « op. cit. p. 135.

• op. cit. p. 308. * p. 306.

Old Accounts of Tikopia 353

John Maresere was quite well aware of the ancient polygyny, but said that it had disappeared, even a chief at the present time having only one wife\ It is unlikely that any external influence has been in action to do away with polygyny, and it is probable that its disappearance has been the natural result of a change in the proportion of the sexes.

Of other subjects on which light is thrown by the ancient records, the following may be mentioned :

The present vegetarian habits of the people are stated to have been in existence in the time of Dillon* and Dumont d'Urville' who are, however, able to tell us that this was not always so. They state that the people at one time had pigs and fowls but had destroyed them on account of the harm done to their gardens. According to John Maresere there are now fowls on the island, perhaps due to recent introduction. There is incidental confirmation of John's statement that formerly the turtle was not eaten, for Dillon* records that a German named Martin Bushart spent many years on the island without tasting animal food except now and then a little fish, while two Englishmen taken from the island by Dumont d'Urville' complained that they had had nothing to eat but vegetables and fish.

Gaimard" gives several details concerning the chiefs in his time which are very interesting in relation to the accounts of the present day. He speaks of the population as distributed in four "villages," Laven-ha, Namo, Outa and Faea, three of these evidently corresponding to the Ravenga, Namu and Faea of Mr Durrad. His names for the four chiefs also correspond, these being given as Kafeka, Tafoua, Fan-harere and Taoumako, in the order of their importance. Kafeka is given as belonging to Laven-ha and Tafoua to Namo\ but the other two chiefs are not assigned to districts and the actual dwelling-place of the four chiefs given in the next paragraph of his account suggests that some mistake has been made in the order, for the chief of Namo is made to live in a village of Faea.

Gaimard's account is interesting in relation to a discrepancy between the two accounts of the island I have recorded. John Maresere stated that each division has its own district

^ See p. 309. 2 op. cit. p. 134.

• op. cit. pp. 114 and 305. * op. cit. p. 134.

^ p. 114. ^ op. cit. p. 306.

^ It may be noted that Mr Durrad records that the late chief of the Tafua lived at Namu.

while Mr Durrad's map shows clearly that there is no such se gregation so far as houses are concerned. It would seem that formerly each division occupied a separate district and John's account indicates that this localisation still persists in some respects although the people live scattered all over the island.

According to John Maresere the chiefs differ in importance. He gave Kafika as the most important and Fangarere as the least, Tafua and Taumako lying between. The order of Dumont d'Urville is Kafeka, Tafoua, Fan-harere and Taoumako, but, as already indicated, the succeeding paragraph of his account suggests that a mistake may have been made in the order of the latter three names. The point on which there is definite agreement is the priority of Kafika, and there is much which suggests that John is right in ascribing to the Fangarere the lowest place.

Both Dillon^ and Gaimard^ speak of a high priest, who according to the latter is the minister of the head chief and has three other priests under his orders. I was told nothing of such office or offices by John, but Mr Durrad speaks of Paefakofe (lo) having surrendered his "sacrificial" powers to the chief of the Taumako and as not acting any longer as high priest. We have thus definite confirmation of the early accounts, though the exact nature of the office remains uncertain.

Gaimard mentions the animal atua or "gods" of the chiefs. He gives the sea-eel as the "god" of the Taumako and says it is regarded as the "god" of the sea, atoua de tai, thus agreeing with both the accounts of the present day. The "god" of the Tafua is given as the rousette {chauve-souris), which may certainly be taken to be the peka or flying fox of our records. The Fangarere are said to possess only the "god" of the sea, called simply atoua, while the Kafika have a fish the name of which was unknown.

Further, an incident is recorded* which illustrates the attitude of the people towards these atua. Some Tikopians coming on board the Astrolabe with flying fish fled (secarterent avec une sorte de terreur religieuse) when they saw a sea-eel on the capstan.

The accounts of Dillon and Dumont d'Urville are especially valuable for their record of the influence of Europeans and other people in Tikopia. When Dillon first went to

' op. cit. p. 122. - Op. cit. p. 307. 3 op. cit. p. 119.

External Influence 355

the island in 1 8 1 3 it seems improbable that it had ever been visited by Europeans. On this occasion a Prussian refugee from Fiji, Martin Bushart, already mentioned, and a "Lascar" were landed there to be found again by Dillon thirteen years later. On Dillon's third visit in 1827 he found five Englishmen on the island who had landed there shortly before. It is therefore clear that Europeans have settled in small numbers on the island at various times.

Further, Dillon^ states that in his time there was a tradition that both Tikopia and Anudha had been invaded by five large double canoes from Tongatabu, the crews of which committed various outrages, destroyed plantations, robbed houses, violated women and murdered men. Gaimard^ also records an invasion from Tongatabu which is evidently the same as that mentioned by Dillon. It occurred in the time of the father of his informant, a very old man. Mr Durrad was told of an invasion from Tonga which probably refers to the same occurrence, and he states that there are still certain people on the island who are regarded as Tongans and these are possibly the survivors of these visitors. Dillon also records* the drifting of a canoe and four men from Rotuma and doubtless there have been many such additions to the population.

The fact that Dillon found on Tikopia relics of La Perouse who had been wrecked on Vanikolo is evidence of the intercourse with this island. The accounts of both Dillon and Dumont d'Urville as well as that given by John Maresere show that there has been extensive intercourse between Tikopia and Vanikolo and other islands of the Santa Cruz group, though the hostile nature of many visits makes it improbable that this intercourse has been such as to allow any great amount of mutual influence.

Anudha.

According to John Maresere the customs of Anudha, Anuta or Cherry Island are like those of Tikopia, kava and betel being used there in the same manner as in Tikopia. The Tikopians say that they had the belief that there must be land in the east from which the sun rose and on going to look for this land they discovered Anudha. The Tikopians used to go freely to Anudha but do so no longer. Not long

^ op. cit. Vol. n, p. 112. * p. 309. ' op. cit. p. 133.

23—2

ago some people of Anudha were going in their canoes to Tikopia but drifted away and finally reached Motlav. They were taken back to Tikopia by Bishop John Selwyn and some of them still live on Tikopia. Two marriages between people of Tikopia and Anudha are recorded in pedigree I X.

Appendix.

In the account given by John Maresere reference was made to his use of remedies which he had been taught by his parents. Soon after his arrival in Tikopia John was seriously ill and going about the island one day he found a tree which his parents had used as a remedy. He drank an infusion of its leaves and was soon well. The Tikopians were much impressed by his recovery and from that time went to him when their own methods of treatment had failed. John's measures are derived partly from his own island and partly from Tonga where his parents had learnt several remedies when on a visit. The chief interest of this account is that it illustrates how elements of culture may spread from one place to another.

Methods of Treatment.

For pain in the head. The leaves of a tree, called in Tikopia maire, are chewed and wrapped in a piece of barkcloth, the whole is squeezed, and the liquid which exudes is dropped into the ears of the patient. The forehead and upper part of the nose are pressed tightly (massaged '^) and some of the same kind of leaves are bound on the forehead of the patient who is then told to sleep. If the pain lessens, the treatment is repeated on successive days till it has been done five times. If on the fifth day recovery has taken place, the patient washes in hot water and pays for his treatment. If the pain is not improved by the first treatment nothing more is done ; the patient is told to go elsewhere. The bathing in hot water is a Tongan practice.

For abdominal pain. The leaves of a tree called in Tikopia nonu and the bark of the mangare tree are wrapped in coconut fibre, soaked in water for a time and then the whole is squeezed into a cup of water, and the liquid so obtained is drunk by the patient. This treatment is also carried out five times but during convalescence the treatment differs

Medical Remedies 357

from that for headache in that the patient is not allowed to eat coconut, banana, Papaia or red yams. Roasted fish is also forbidden though boiled fish may be eaten, all these restrictions on diet being derived from Tonga.

Procedure for a man injured by falling from a tree. First there is prepared some coconut-cream (sinu) which is added to some boiling sea-water. The patient is then asked where he feels the pain and it is noticed that something is out of place which is put into its proper position. Then some leaves of the orange (lime ?) tree are procured and wrapped in coconut fibre. The whole is beaten to a pulp, dipped in the mixture of coconut-cream and sea-water and squeezed into a cup of water. This is drunk by the patient who is then told to micturate. He will pass blood which will carry off the pain. It is believed that the putting straight of the tendon or other structure which was out of place has allowed the blood to pass freely and that if the blood is not so passed, it will thicken and corrupt and the patient will die in five days\

Bleeding. The arm is held out straight so that the front of the joint is tense and the arm and wrist are bound tightly. A shark's tooth is then bound tightly in the cleft end of the reed from which arrows are made and the tooth is driven sharply into the vein in front of the elbow. The blood squirts out to a great distance and is caught in a coconut cup. Two cups of blood are taken and the thumb is placed over the wound and when the binding of the arm and wrist is removed the flow of blood is found to have stopped.

This method of bleeding resembles our own so closely as to leave little doubt that it has been learnt from Europeans. It may be noted, however, that it was said that this method of bleeding is only used for affections of the arm. John also practised bleeding by making cuts in the back, while he himself had been treated when a child in Wallis Island by cutting the abdomen.

Method of hastening delivery in childbirth. Leaves of a hibiscus (kaute) are pounded ; the juice is squeezed into a cup of water and then drunk by the woman. The liquid should be drunk as quickly as possible.

^ In connection with this treatment I may call attention to the ^Uocolosi" treatment of Tonga (Mariner's Tonga, first edition, ll, 254). It is possible that we have here ideas concerning pathology which may explain this mode of treatment.

I

2^

< <i

a, '<^ eL,3

<

<

"^

^1

C/3 ^

IS

"i

O

a

<

o

z

<^-~.

O <3

!5rS

fc^-'

w

<

'fo

0^

»-1

rt^

2 «

<x:

•£ i

'5

^

b ^

!:

<

1^

3 v2-

nJ

1— !^

ni

II

II

II

< <

<

o <

<

rangarua F adopted

^.

ra

^

^-^

3 «i

■v,

C<«;

asr**

;?

■c s

r-'';

^

b

ii

s

£

§

^

ft;

II

<

JD

v?S

p^-""^

2^

Sv

-IS

s

<

o ^

C 0

\

ft.

,^

tq

^

»:

o r H 1

■2 1^ . r^

J

^ 11

I 1^

r^

^ M S

•^

-g

— rt «

■^

<

< 2 II

J

^

.«

i4

^

^

Cu

< '^

■3

:

<a

S

Oh /n

^

§

-S

Q^

K.

«

• i~i

a

C

a

(2

venutapu

nudha

Tafua)

•J

St t,

II ■■ "1

1^

<

g 2

11-

II — % 3 3

Paetearik

(Tafua

d. May i6

Paevenutap^ {Tafua)

J

t/3

Q

ft;

!'«

^

**

^

^

-1

Q

z

!^

%»

^

II

V

^

«

!-:

^

g

'to

s

''I c

-^

II-

t* ,

^

i

"1

»«4

S

—

•^

5;

k ^

5 ^ S<

t lb

o

c

5^

Q.

•<

ft:

Jo

1 ^

i;^

II '

St

mrangifur

now Nautafua {Kafika)

1!

-8?

2i

t

II

-^b

ft; Sm

'to

V

^ '

^

o

^

)-■

II —

■■

s

bo

1 §

V

Si

'5)

c

^•^

Z> ■>.

t*

II

o a

z

II

II

ft; t«,

II

• S

^ ^:

_k.

u

L.<K,

o

2 ^

"*) r

Z <fl

.pS

<

2 '«j

as o '— >

< S

r "s

-II —

Si

uS

I

•<

r

is;

■^ ft

a,—

lb

§0^

J3

^

ft! <a

ft.

^

^i

§

^

a S

^

o+

a, o

2'^

I

rt Si

a- <3

I -^ a

_ ^

K

^

ft.

ft.

Si u5

3 2

O Si

-11

Si ft.

Melaitesian Society

*o

^

H

■k

♦O

"^ o

o S OSS

^

< '^

< Ph

s

Oh

<3

ei

Q

K

<3

f--^

K

«

^

ti

ys

«<

-«

5

^

•»5

\r\

f^

^^

'2 "^

"S-^

ti

S O-

£<,

=5 •<

<

^

-«

ti

s

<

'^

0 2

^'

t^^^

fc; S

"^

r-"^ S

■35 <<

Pft; «

ti

.iS^

kj Si

\

3 ^^

^

rt ^

1=^

p

2^

<<«

Fn

'^

^1

^1

(5

CHAPTER XIII

TONGA, SAMOA AND NIUE

During a few days spent in the Tongan and Samoan Islands inquiries were made into one or two matters, attention being devoted especially to the systems of relationship. I was also able to obtain the system of relationship of Niue from some natives of the island who formed part of the crew of a vessel on which I was travelling. It was only possible in the time at my disposal to make superficial inquiries, but the systems of relationship are probably fairly complete and correct. The Tongan system agrees closely with that recorded by Morgan^ and collected for him by the Rev. Lorimer Fison. The fact of this close agreement makes one hope that, in spite of the brief time available, the other systems I have recorded are also correct.

Tonga.

The information which follows was collected during a short stay of two days at Nukualofa on the island of Tongatabu. Through the kindness of Dr Maguire I was able to avail myself of the services of a very intelligent Tongan named Uga with a good knowledge of English, and with him and an older native of the island I was able to go into matters connected with relationship fairly thoroughly, though doubtless I obtained only a small fraction of the information which a longer stay would have brought to light. I did not make any extended inquiry into the details of social grouping but was told that each "family''" has connected with it a sacred

1 System of Consanguinity and Affinity of the Human Family^ p. 569.

2 This word was used by Uga and certainly refers to more than the family in the strict sense, but I was not able to define the social groups with which the otua are connected.

object called an otua. Some otua are animals, examples being the octopus, flying fox and pigeon, while others are stones. No example was known of a "family" which had a plant as otua. When the otua is an animal there is a belief in descent from it and it is not eaten.

The following are the terms of relationship with their meanings : —

Tamai. Father and father's brother and other men of their generation according to the usual rules of the classificatory system. Also the husbands of the mother's sister and of the father's sister. It is also applied by a man to the father of his wife and by a woman to the father of her husband. The real father may be distinguished from other relatives to whom the same term is applied by the expression tamai inooni.

Fae. Mother, mother's sister and the wives of the father's brother and of the mother's brother ; also the mother of the husband or wife. The mother's brother also is sometimes called fae, the word for mother being thus used to designate a man.

Foha. Applied by a man to his son and also to the son of his brother or of others whom he would call brother.

Tama. Used by a woman for her son and for her sister's son. The sons of her husband's brothers would also be called by this name.

Ofefine. The term for daughter used by both men and women and applied according to the same rules as foha and tama.

Fanau. Used generally for a child, either male or female.

Tokoua. Used by a man for his brother and by a woman for her sister ; also applied in the corresponding way by persons of the same sex to the children of the father's brother, of the father's sister, of the mother's brother and of the mother's sister. It is thus used for all cousins, whether they are related through the father or through the mother. The husbands of two sisters or the wives of two brothers also call one another tokoua.

Taukete and teJima. Terms for elder and younger brother (m.s.) and for elder and younger sister (w.s.). In the family in the limited sense these terms are applied according to the relative age of the persons themselves, but for more distant relatives their correct use depends on the respective ages of

Relationship in Tonga 365

their parents, thus, a man would call the son of his father's younger brother tehina, even if the latter were older than himself, and similarly in the case of children of a brother and sister. At the present time the expressions tokoua lahi and tokoiia sit are often used in place of tauketi and tekina, but the latter words are those proper to the old Tongan system according to my informants.

Tungaane. Used by a woman of her brother.

Tuafajine. Used by a man of his sister. Both tungaane and tuafajine are used according to the same rules as tokoua but the terms tauketi and tehina are never used between men and women.

Tuasina. Mother's brother and others whom the mother would call brother. As we have seen, the mother's brother may also be called y^^. The wife of a tuasina \sfae.

Ilaniutu. Applied by a man to the child, whether son or daughter, of his sister. If his sister is older than himself, the man will call her child ilamutu eiki or tama eiki{eiki meaning chief). Ilamutu is also used in the classificatory sense for the children of any woman who would be called sister and in this case the expression ilamutu eiki is applied to the child of a woman descended from an older branch of the family.

Mehikitanga. The father's sister and others whom the father would call sister. Her husband is tamai.

Fakafotu. The term reciprocal to mehikitanga, applied by a woman to the child, son or daughter, of her brother.

Kui. This term applies to all grandparents, both male and female, and to other relatives of their generation. The parents of the father and mother may be distinguished as kui ihe tamai and kui ihe fae respectively.

Mokopuna. Grandchild, the child of either son or daughter.

Unoho. This is the old term for either husband or wife, but owing to its also being the term denoting sexual intercourse, it is not now used. There is another ancient term hoa, meaning partner, but this also is not used at the present time and the customary term is maliy which is the Tongan form of the English word "marry." The wife's sister, the sister's husband (w. s.), the brother's wife (m.s.).and the husband's brother were all in the old time called unoho.

Matapuli. The wife's brother and sister's husband (m.s.).

Maa. Husband's sister and brother's wife (w.s.).

Properly matapuli should only be used between men and maa between women, but this distinction is not always now preserved.

As already mentioned, the wives of two brothers and the husbands of two sisters, call one another tokoua.

I could not find that there was any term for the relationship between the parents of a husband and wife.

When a man had several wives, they were collectively sinifu and the chief wife was mataki.

This relatively simple system has several interesting features. There are special terms for the mother's brother and the father's sister and each term has also a special reciprocal, but the consort of the maternal uncle and paternal aunt are not distinguished in nomenclature from the mother and father respectively, and no distinction is made between their children and those of the father's brother and mother's sister.

Another feature is that each parent uses a special term for a son while they share a term for their daughter. Further, a brother and sister do not use one reciprocal term but a man calls his sister by one term and she addresses him by another. Still another noteworthy feature is that the parents-in-law are not distinguished in nomenclature from the parents.

Functions of Relatives.

The Mother s Brother^ In Tonga the tuasina or mother's brother is " the same as a servant for his sister's son. He has to listen to the boy all the time." The sister's son is especially honoured by his uncle and can take anything he likes from his uncle's place ; his pig, his canoe or anything he chooses, and no objection will be raised. Indeed, I was told that a man would be very glad to see his nephew take any of his goods. A man looks after his sister's son while he is young and, when the latter goes to his uncle's house, he can do as he chooses and is subject to no restrictions on his conduct such as we shall find affect him when he visits his father's sister. On the other hand, a boy has not to be especially obedient to his maternal uncle ; if the boy's father tells him to do anything, he will do it, but, if told by his uncle, he will do it or not as he pleases. There are no restrictions on conversation with the tuasina nor is there any prohibition

I

The Father s Sister 367

on the use of his name, and indeed in this as in the case of all other relationships it is customary to use the personal name in address. I could not discover that the tuasina would take any special part in ceremonial connected with his nephew either in connection with naming, incision or marriage. Though he might assist in providing the feasts on such occasions, his part did not appear to be more prominent or even as prominent as that of the father or father's brothers.

A woman can also take anything from the house of her maternal uncle and has the same rights and privileges as her brother.

The ilamutu eiki is more highly honoured than other ilamutu, but the only practical result of this which I could discover was that if the uncle had two pigs, a large and a small, the former would go to the ilamutu eiki.

The Father s Sister. My informant described the position of this relative by saying that the relation of the paternal aunt or mehikitanga to her fakafotu or brother's child is exactly the same as that of the ilamutu to his tuasina. Just as the tuasina honours his ilamutu, so does the fakafotu honour his mehikitanga ; just as the ilamutu can take anything from his tuasina, so can the mehikitanga take anything from her fakafotu, i.e. the aunt may take anything from her nephew. A man honours his father's sister more than any other relative, even more than his father or his father's elder brother, and it was believed in the old time that, if anyone offended his father's sister, he would die. The mehikitanga usually arranged the marriage of her nephew or could veto one arranged by the parents or by the man himself. Even at the present time a man will usually take the woman whom his father's sister wishes him to marry, though he now often pleases himself. As already mentioned, there are many restrictions on the intercourse of a man with his father's sister ; he may talk to her and address her by name, but he may not eat in the same place with her nor may he eat anything which she has carried. He will not sit on her bed and if she comes into a house in which he is present, he will at once go out.

The mehikitanga also arranges the marriage of her brother's daughter and may take any of her possessions. I could not discover that the aunt took any part in ceremonial

368 Melaitesian Society

connected with her nephew, but at a feast^ on the occasion of the first menstruation of her niece, the mehikitanga took the chief part and the menstrual blood was given to her in a piece of bark-cloth.

The husband of a mehikitanga shares in his wife's rights, but it seemed clear that this was due to the fact that a woman would wish her husband to share her privileges and was not due to his own relationship.

Samoa.

The following information was obtained partly at Apia, partly from a Samoan whom I met in Fiji.

The people regard certain animals as sacred and call them atua. One informant told me that each district had its atua ; that the atua of his district on the outskirts of Apia was the fee or octopus ; that this animal was the atua of about 400 people ; and that there was a house for the animal at a place called Falipouma. The atua of the next district was the lulu or owl ; this bird belonged to about 500 people and lived at Sagata. In another district the atua was a big kind of shell. In each case in the old time the people did not eat the animal which was their atua, but my informant was quite decided on the point that there was no belief in descent from it. In the old war times the atua, whether octopus, owl or shell, used to call out so that all the people could hear, and the people would then be glad and know that there was going to be a fight. The owl used to fly in front of its people as they went to battle. Another informant spoke of each family having an atua, but I was not able to find out what he meant by a family. His own atua was a large stone called Kalinga i, and he gave the shark and garfish as other examples, while in some cases the atzia was a tree. In the old time it was believed that those who had the shark for an atua were in no danger of being troubled by a shark when they went into the sea.

The system of relationship was obtained from several natives independently, partly by the genealogical method. I put the data on record with much more hesitation than in the case of any other system, for they are very anomalous and fall

• This feast is called 7igata or end, owing to its marking the end of the girl's growth.

Relationship in Samoa 369

so little into line with our knowledge of the mode of expressing relationships found elsewhere in Polynesia. I am uncertain about the correctness of many details and record the system only in the hope that my mistakes may induce others to carry out a full investigation of the very exceptional mode of denoting relationship which seems to exist in these islands.

The following are the terms with their meanings : —

Tama. Father. I was assured by more than one native that it was only properly used for the father's brother after the death of the father when his brother had taken over the guardianship of his nephew, but on the other hand I was told by Europeans that the term is habitually used by the natives, not only for the father's brother and for others of his generation, but also for the grandparents and others of their generation.

Tina. This term was said to be limited to the relationship of mother but it was probably also generally used in the same wide sense as tama.

Matua. This was said by some to be limited to the proper parents and grandparents of a person but it seemed that it was also generally used for any elder.

Tupunga. Said to be used in the same sense as matua. It also means 'ancestor.'

Atalii. The word for son, probably only used by a man for his sons and for the sons of those he would call brother.

Tama. Another term for child which probably should only be properly used by a woman of her children. Atalii and tama, certainly the latter, appear to be used for grandchildren. So far as I could discover, there is no special term for grandchild.

Afajine is the word for daughter.

Uso. This term is applied by a man to his brother and by a woman to her sister, and it appears to have been used by each sex in a very wide sense, probably for all those of the same generation with whom relationship could be traced.

Tuangane. The term used by a woman for her brother.

Tuafafine. This is applied by a man to his sister, both tuangane and tuafafine being used in the same wide sense as uso.

I could learn of very few terms for relatives by marriage. A wife was called ava and there were also distinctive terms for the wife of a chief and of an ordinary man, the former )^\Vi<g faletua and the latter tausi. The wife's father was said to be Q.2i}^^^i faletua.

There are thus very definite terms for brothers and sisters, viz, one word for the brother-brother relationship and other words for the relation of a brother to his sister and for the relation of a sister to her brother. There are also distinctive terms for father and mother, while the frequent Polynesian term for parents and grandparents, though present, seems to have become very indefinite in meaning. There are no distinctive terms for the relationships of grandparent and grandchild, and definite terms for relatives by marriage are either absent or very indefinite in meaning. The system is so incomplete and its scanty terms are so indefinite in meaning that it would seem that the people must have some other means of expressing relationship and there is no doubt that this is the case.

In my first attempt to obtain the system of this island from an educated Samoan living in Fiji I was given the terms for father, mother, son, daughter, brother and sister as I have given them above, but all other relationships were expressed in a purely descriptive manner ; thus, the father's brother was called uso o le tama ; the father's sister's husband was tane a le Uiafafine o le tama ; the mother's brother's son was atalii 0 le uso 0 le tama, and so on. On continuing the investigation in Samoa there seemed to be no doubt that the people were well acquainted with this descriptive method of expressing relationship, and for any occasion on which relationships have to be expressed with any degree of definiteness, I think there can be no doubt that they resort to these descriptive terms.

The marriage regulations are of the same kind as those usual in Polynesia, a man not being allowed to marry the daughter of the brother or sister of either father or mother, or more simply and correctly he may not marry anyone whom he would call tuafajine, and it is probably the existence of this marriage regulation which has helped to preserve the special method of denoting the brother-sister relationship while so much else has been lost.

There are no restrictions on the use of personal names

Nine 37 1

and even the nearest relatives are addressed in this way. One of my first experiences in Samoa was to hear a child a few years old addressing her father by his personal name instead of using a term of relationship.

NiuE OR Savage Island.

One of the Niue men from whom I obtained the system which follows, Viri, was so intelligent that it was possible to use the genealogical method in spite of his very scanty knowledge of English. Without this method it would have been impossible to obtain information of any value whatever, whereas the data actually obtained are so consistent that I have little doubt that I obtained the essential features of the system, though no doubt there may be errors, especially of omission, in detail.

The following are the terms with their meanings : —

Matua kane. The father, the brothers of both father and mother and other relatives of their generation, including the husbands of the father's sister and of the mother's sister.

Matua fifine. The corresponding term for the mother and her sisters, the father's sisters and the wives of the brothers of both father and mother.

Fanau and tatna. Terms used for children both male and female, used generally as terms reciprocal to inatua kane and matua fifine. I did not discover whether either or both of these terms are used by one parent only.

Taukesi. A term applied by a man to his elder brother and by a woman to her elder sister ; also used for elder relatives of the same generation, as for the children of the father's brothers and sisters and of the mother's brothers and sisters. I could not learn with certainty whether this term is used for those actually older than the speaker or whether its use depended on the relative ages of the fathers of the persons concerned.

Seihina. The term for the younger brother of a man and the younger sister of a woman, used in a similar manner to tuakesi.

Tungaane. Used by a woman of her brother in the same wide sense as previous terms.

Mahakitanga. The corresponding term given by a man to his sister.

24 — 2

Matua tupuna. A general term for the grandparents and for other relatives of their generation. Those of the female sex are distinguished as matua tupuna fifine.

Mokopuna. The corresponding term for grandchildren, for whom however the term mangafaua (see below) is also often used.

Matakainanga and mangafaua. These are terms applied to relatives in general. I was at first very much puzzled by them, one or other of which was used of various relatives. Thus, I was first given matakainanga as the term for the mother's brother and mangafaua as that for the child of the father's sister and thought that they were special terms for these relatives, but it became clear that they were properly matua and taukesi respectively and that the terms matakainanga and mangafaua had only been used of them as they might be used of many other relatives. It seemed that it was only customary to use these terms of relatives other than those of the speaker's own family in the restricted sense\

Matua vungavai. The term for the father-in-law and mother-in-law and for the brothers and sisters of these.

Fingona. The term reciprocal to vungavai, used for the son's wife or the daughter's husband. Those of the female sex are distinguished as fingona fifine.

Hoana is the term applied to a wife. It is also used by a man for a brother's wife but I could not ascertain that it would be used for a wife's sister.

Kane. This, used generally in order to distinguish a male relative, is also the term for a husband.

Maa. This term may be used by a man for a brother's wife in addition to the term hoana ; thus, a man may call the wife of his elder brother either maa or hoana taukesi or hoana alone. Maa is also used for the wife's sister, for the sister's husband (w. s.), and for the brother's wife (w.s.). There seemed to be little doubt that it was also applied to one another by the husbands of two sisters and by the wives of two brothers.

I could not discover that there was any term for the relationship between the parents of a man and the parents of his wife.

• I may note that Turner in his comparative vocabulary {Samoa, London,

1884. gives mata keinanga as the word for brother. If I am right he must have been misled by the tendency to give this name to any relative which at first misled me.

Marriage 373

The marriage regulations seem to be of the kind usual in Polynesia and may be summed up in the law that a man may not marry his makakitanga, the daughter of his father's brother, his mother's brother, his father's sister or his mother's sister.

So far as I could ascertain, there were no restrictions of any kind on the use of personal names and it was the custom to address all relatives by name rather than by the terms expressive of relationship.

CHAPTER XIV

THE HAWAIAN ISLANDS

This chapter differs from those which have preceded it in that it deals with islands now so permeated by external influence that little remains of the institutions and customs proper to the people. Few pure-blooded natives are to be found, at any rate in the only two islands, Oahu and Kauai, which I was able to visit. Further, many of those who are partly or wholly of Hawaian origin and take an interest in their past history are so concerned in demonstrating the high character of their former culture and their racial affinities with Aryans and Semites that little trustworthy information is to be obtained from them, although its value is rated so highly by themselves that I was once asked to give twenty-five dollars for the privilege of a conversation.

In consequence of these difficulties the actual amount of concrete information which I was able to collect was very small, but I have ventured to supplement it by giving the impressions I gained from conversations with those who had any knowledge of the past condition of the people.

I worked chiefly with a number of old men, inmates of the Lunalilo Home for old and infirm Hawaians, and I am much indebted to Mr Joseph Emerson for introducing me to this sphere of work and to Mrs Ellen Armstrong Weaver, in charge of the Home, who did everything in her power to help me. Among Hawaians who helped me I am particularly indebted to Mr J. M. Poepoe who has a genuine interest in the history of his people, though his information has necessarily to be derived from written documents rather than from direct knowledge of Hawaian custom. I am also greatly indebted to Mr Aubrey Robinson of Kauai who knew the people when their condition was very different from that of the present time.

Genealogies.

Most of the old men in the LunaHlo Home were able to give me extensive pedigrees similar to those generally to be obtained in Oceania. The most extensive included six generations, the oldest being that of the grandfather of my in- formant, who was himself over eighty years of age.

The inmates of the Home come from all parts of the Hawaian group, so that the pedigrees were disconnected and of little value except in the investigation of the system of relationship, and I have not therefore thought it necessary to publish them.

One general point of interest is that the pedigrees of both father and mother were well known, and I gained the impression that the relatives of the mother were known somewhat more fully than those of the father ; certainly there was no inferiority in the knowledge of the mother's pedigree. Certain other features of interest will be mentioned later.

The System of Relationship.

Previous accounts have been published by Morgan^ and Hyde*^ and the data obtained by myself agree in general with those given by these writers. I have, however, been able to make a few additions and to describe more exactly the full connotation of the terms. These are as follows.

Makiia. This term applies to the parents and to all those of their generation with whom direct relationship can be traced, i.e. father's brothers and sisters and mother's brothers and sisters, both in the classificatory sense. It may often be used figuratively, but the distinction is quite clear between the usages in this and in the more direct sense. Sex is distinguished by adding the words ka^ie or wahine.

Luaui. A term used only for the real parents in distinction from those whom we should call uncles or aunts. There seems to be no doubt that it is an old word which is now rarely used and I found some Hawaians hardly knew of its existence. There was a question whether it should not properly be limited to the mother, but it seemed clear that it

1 System of Consangumity and Affitiity of the Human Family^ 1871, pp. 451 and 523, and Aticieftt Society^ p. 419.

2 Hawata?t Almanac and Annual, Honolulu, 1884, p. 42.

376 Meianesian Society

was used for both parents who were distinguished, as in other cases, by the terms kane and wahine.

Step-parents are now distinguished as makua-kolea. According to Hyde, kolea originally meant real parent and its use for step-parent is recent.

Kaiki. The term for child, sex being distinguished by the words kane and wahine as in other relationships. It is also applied to the children of all those whom the speaker calls kaikua-ana, kaikaina, kaikunane or kaikuhine.

Kaikua-ana is the term for elder brother (m.s.) and for elder sister (w. s.) and those of the same generation who belong to an elder branch of the family, so that it may be applied to a person who is actually younger than the speaker. The son of an elder brother of the father or mother would thus be kaikua-ana, even if younger than the speaker.

Kaikaina. Younger brother (m.s.) or sister (w.s.) or child of a younger brother or sister of the father or mother.

If there are three sons in a family the two younger will call the eldest kaikua-ana and will be called by him kaikaina. The third will call the second kaikua-ana. The second will call his elder brother kaikua-ana and his younger kaikaina.

Kaikunane and kaikuhine. These are terms used between those of opposite sex. A woman calls her brother kaikunane^ and gives this name also to all those men of her own generation with whom she can trace relationship. Kaikuhine is the corresponding term applied by a man to his sister or other female relative of his own generation.

Kupuna. The parents of both father and mother and others of their generation with whom definite relationship can be traced. It is no doubt often applied figuratively to other persons, but in such a case there seemed to be little doubt that the distinction between the ordinary and figurative uses was definitely recognised. Ancestors of the preceding generation, corresponding to our great-grandfathers and great-grandmothers are also kupuna, but are distinguished as kupu7ia-kua-kahi. Similarly, there are terms for still older generations, kupuna- kua-lua for great-great-grandparents and kupuna- kua-kolu, kupuna- kua- ha, etc., for still older generations [kahi, lua, kolu, ha are i, 2, 3, 4).

The grandfather and grandmother are distinguished from one another by means of the additions kane and wahine or

' According lo Mr Poepoe, this term is a euphonic form oi kaikuakane.

hine, and sex was distinguished in this manner only in all cases in which the term was used,

Moopuna. Grandchild and others of that generation, being used as the reciprocal of ktipuna. Moopuna is also used for more remote descendants on the same lines as in the case of kupuna, thus the great-grandchild of a man will be his nid opuna-kua-kahi.

The above are the only terms for relatives by blood, there being thus, with the exception of lumii, only one term for the people of each generation older or younger than a person, while for relatives of the same generation there are four terms the use of which depends on sex and seniority.

There are several terms expressing relationship by marriage : —

Kane. Husband, also applied by a woman to the brothers of her husband. It was probably used by a woman for all whom her husband would call kaikua-ana or kaikaina, but it is now difficult to obtain satisfactory evidence on this point. A woman also calls her sister's husband kane.

Wahine, The term for wife and for the sisters of the wife, probably using sister in the wide sense. The wife of the brother is also wahine.

Hunoai or makua hunoai. Parent-in-law, sex being distinguished as in other cases by means of kane and wahine. The terms are not limited to the real parents of the husband or wife but are applied to all who would be called makua,

Hunona is the corresponding term for son-in-law or daughter-in-law.

Punalua. Applied to one another by the husbands of two sisters or by the wives of two brothers. If a woman has two husbands, they are punalua to one another and any paramour of the wife receives the same designation. I could not ascertain definitely whether the husband of everyone whom a man's wife would call kaikua-ana or kaikaina would be his punalua, but probably at one time this was the case, and similarly it is probable that the punalua of a woman would include the wives of all those men whom her husband would call by the above terms.

Kaikoeke. Used by a man of his wife's brother and his sister's husband, brother and sister being used in the wide sense. Similarly, the kaikoeke of a woman are her brother's wife and her husband's sister.

Punalua and kaikoeke are thus terms only applied to one another by persons of the same sex, while kane and wahine are only used between persons of opposite sex.

On Oahu brothers-in-law and sisters-in-law are often distinguished as makua or opio according to the age of their consorts ; thus, a man will call the husband of his elder sister kaikoeke makua and the husband of his younger sister kaikoeke opio, while a woman will call the husband of her elder sister kane makua and of her younger sister kane opio. Similarly, the husband of an elder sister is called by a woman kane makua while the husband of the younger sister is kane opio. In Kauai the term nui is used in place of makua.

Puluna. A term applied to one another by the parents of a married couple. The father and mother of a man are the puluna of the father and mother of the man's wife, the term being used reciprocally.

Some other terms occasionally used were : —

Hanai, a term for an adopted child or a child brought up by a person, but it is occasionally used by a man for his own child, in which case it seemed that it was regarded as a title of honour. It is probable that it was originally a term only used by the chiefs and has thus come to carry with it an idea of distinction.

Pokii seems to mean properly the youngest of a family but is sometimes used in the same sense as kaikaina.

I endeavoured to find out if there were any means of distinguishing in nomenclature between own fathers, brothers, and sisters, etc., and those more distant relatives to whom the same kinship terms are applied. It seemed that the nearer relatives, i.e., those whom we should call father, mother, brother or sister, etc. may be distinguished by the addition of the word ponui, so that the real father would be makua ka?ie pomii to distinguish him from those we should call uncle, but it was said that this usage is recent. Similarly, I was told that the word hanauna might be used to distinguish those related in the more distant way. Thus, a cousin (elder) would be kaikua-a?ia hanauna, a term which would not be used for an own brother, but it was doubtful whether this usage is not also recent.

In the foregfoinpf account the extension of the terms has been described as limited by the power of tracing relationship. The degree of extension of the terms is a point which

both Morgan and Hyde left uncertain and I had much difficulty in obtaining satisfactory evidence on this question. Owing to there being no clan-organisation among the Hawaians, there is an absence of the machinery for the limitation of the use of the relationship terms which is found among many peoples who use the classificatory system, and the account given by Morgan would seem to imply that every Hawaian is either the kupuna, inakua, kaikzta-ana, etc., keikt or mdopmta of every other Hawaian. In my earliest inquiries it almost seemed as if this was the case, but it soon became clear that this impression was only due to the widespread practices of adoption and of forming bonds of artificial brotherhood together with the frequent use of terms of relationship in a figurative sense. With these exceptions, the use of the terms is limited to those with whom direct relationship can be traced. I think it probable that, at any rate in the case of especially important people, it was not necessary that this relationship should be capable of being traced out in full detail. The Hawaians are great genealogists and it is probable that two men meeting and finding that they had a common ancestor would apply the suitable designation to one another, but this use was probably distinguished from that in which a definite near relationship could be traced.

Several old men were very emphatic that they would not apply the terms of relationship to any persons except those included in the pedigrees which I obtained from them. This evidence was not very satisfactory, and it is of course possible that there had come about in recent times a limitation in the use of the terms. I can only record the fact that in all my inquiries I found little to differentiate the use of the terms of relationship from that which I have found in other places, where there often exists a wide and more or less figurative use of the terms of relationship side by side with their use in a limited sense for those actually related by blood, adoption or marriage.

Marriage.

In any consideration of Hawaian marriage, as indeed of nearly all Hawaian social institutions, a distinction has to be made between the customs of the chiefs or alii and those of the ordinary people or makaainana. The two will therefore be considered separately.

Marriage among the chiefs. It seems to be quite clear that among the chiefs, and certainly among the higher chiefs, there existed the institution of individual marriage, in which the union of two persons was accompanied by definite ceremonial. Of this marriage to which the term ho-ao was applied, there were several grades, depending on the relation to one another of the pair to be united. According to an account given to me by Mr J. M. Poepoe, there were, at any rate in the island of Oahu, nine grades of marriage.

The highest kind of marriage was that between a brother and sister, children of the same father and mother of the highest rank. Such a marriage was called /z'^ (bending) and its offspring niau-pio or branch of the pio. Such chiefs were so sacred that all had to prostrate themselves in their presence and they were not allowed to take off their clothes in the presence of others. As examples of this grade of marriage Mr Poepoe gave me the following : Kamahanu, a chief of the highest rank or alii niau-pto, married Lonukahikini, a chiefess also of the highest rank or niau-pto. They had a son, Kaneoneo, and a daughter, Kapueo-kalani, who married one another. Another example of the pto marriage was that of Kamehameha-nui of Maui who married his sister, Kalola, and their daughter Kalani-kauio-kikilo was a niau-pio.

Naha. This name was given to the next grade of marriage in which the husband and wife were half brother and sister. The usual case was that a woman had two husbands ; the first child would be regarded as the offspring of the first husband and the second child as belonging to the second husband and, if they were male and female and married one another, the marriage would be naha\ if there was a third child, it would not be royal and would be given away. People had to sit in the presence of chiefs of this rank. As an example of such a marriage, Kalola-nui, a chiefess or alii niau-pi 0 married first Kalaniopuu, also alii niau-pio, and had by him a son Kiwalao. She then married (or took as punalua) Kaoua by whom she had a daughter, Keikuipoiwa or Liliha. Kiwalao married Keikuipoiwa and this marriage was of the grade called naha.

Wohi. If a chiefess who was a niau-pio married a chief who was naha, the marriage would be of the wohi kind and the issue would receive this name.

Marriage 381

Lo alii. A marriage limited to the island of Oahu, where there were certain chiefs living in the forests called alii lo. If the chiefs of the other parts of the island wanted new blood, they married with the alii lo and this kind of marriage was named after them.

Alii papa. A marriage between a chief who was either niau-pio or naha with a chiefess who had the prostrating taboo \

Lokea alii. A marriage in which an alii lo married one of low chiefly rank.

Lda-uli alii. A marriage in which a chief who was either niau-pio or naha married a chiefess of lower rank. i^Uli means blue-black and the marriage was said to be so called because the blood was not clear.)

Kaukau alii. A marriage in which a chief not of the highest rank married a chiefess, either niau-pio or naha.

Lepo-popolo or kukae-popolo. A marriage in which the wife was of high rank and the husband one of the ordinary people {lepo = ea.rth), i.e. a marriage of an alii wife with a man of the makaainana. In this case there was a distinct slur on the birth of the child though it would rank as a chief owing to the condition of the mother^

All these marriages were accompanied by ceremonial of a simple kind. For those of the lower grades, the pair came together and lay under a piece of bark-cloth and all present called out three times ''hoao-e' or "hoao-e na alii.'' There was an exchange of presents and there might be sacrifices, the number of presents and sacrifices increasing with the rank of the newly married pair. In the case of a marriage of the highest grade there would be a great concourse of people and the pair would unite in a tent of bark-cloth of which the bark-cloth above mentioned was probably the representative in marriages of lower grade'. In the high grade marriages of the chiefs there was the closest bond of blood-kinship between the husband and wife. Marriage between own brother and sister was the most desirable union and it was said that marriage between father and daughter

^ I give this marriage as related to me but its exact nature is doubtful.

2 For another account of the grades of marriage among the chiefs see David Malo, Hawaian Antiquities^ translated by N. B. Emerson, Honolulu, 1903, p. 80. This account is less complete than that of Mr Poepoe from which it differs in some details.

^ See David Malo, p. 179.

might also occur. How far such marriages occurred among chiefs of low rank, it is now impossible to ascertain.

It seems to be quite clear that these consanguineous marriages were carried out with the very definite aim of maintaining the purity of the royal blood. According to David Malo (p. 93) only a few of those about a court lived in marriage and it seems that even in the case of the highest chiefs marriage was only an institution to obtain offspring and as soon as two children of different sex were obtained, the union had no further motive and might be dissolved. The chiefs of the Hawaian Islands only maintained the institution of marriage to an extent just sufficient to ensure sue cessors to those of the highest rank whose purity of descent could be guaranteed.

Marriage among the ordinary people. I had the greatest difficulty in obtaining any satisfactory information about the institution of marriage among the makaainana. It seemed quite clear that it was inaugurated by no definite ceremony. There was general agreement that consanguineous marriages were not allowed among the ordinary people, and that, if they ever occurred, they were looked upon with general disfavour, this kind of marriage being regarded as the special privilege of the chiefs. It also seemed clear that the objection to marriage with a sister extended to all those who received the same designation as a sister, viz. kaikuhine. There thus seems to have existed among the ordinary people just the same kind of restriction on marriage as is found among most peoples of rude culture who possess wide restrictions on marriage dependent on kinship, and I could not ascertain that it made any difference whether the relationship was on the father's or the mother's side and, since both kinds of cousins receive the same name, the prohibition probably extended equally to both.

The Hawaian commoners appear thus to have had exactly the same kind of restriction on marriage as is found among other Polynesians, and I was told that the fact that the first Christian teachers allowed marriage between cousins was a distinct stumbling-block when the new religion was introduced.

It seems to be quite certain that there is no trace of a clan-organisation among the Hawaians but I endeavoured to ascertain if there was any indication of restriction on marriage

Marriage 383

dependent on locality which might possibly be the vestige of such an organisation. It was said that the people of a district acted together in certain circumstances, as in warfare, but I could not discover that the locality of a person in any way influenced the choice of a wife. In collecting pedigrees I usually obtained the name of the district and place to which each person belonged. Every person was given as the inhabitant of a district such as Ewa or Waialua on Oahu, and his locality could also be given more precisely within such districts so that the exact place to which every person belonged was known. It was quite clear that marriages might occur between inhabitants of the same district, but I cannot speak definitely about the influence of the smaller divisions of a district, and it is possible that they may have played some part in the regulation of marriage. It is perhaps significant that the place to which everyone belonged should have been preserved, even although their descendants might have moved to other islands.

There seems to be distinct evidence, at any rate among the ordinary people, for the existence of trial marriages. Dr Hyde^ derived the Hawaian term for marriage hoao from a root meaning "to try" but Fornander- objected to this derivation and Dr N. B. Emerson^ has pointed out that the word meaning "to try" is ho-do and not ko-ao\ Though the derivation of the word for Hawaian marriage thus appears not to carry with it any idea of a temporary nature, several independent witnesses, men at the Lunalilo Home, stated that in the old time, at any rate among the ordinary people, a couple came together for a term of weeks or months, after which the union was dissolved if not found to be satisfactory.

In connection with marriage one or two words may be said on the system of descent. It has often been said that the Hawaians practise female descent but this expression has a far smaller significance than in most cases in which it is

^ op. cit.

2 Hawaiati Almanac and Annual, 1885, p. 46.

3 Report 0/ the Hawaiati Historical Society, 1898, p. 16.

• Mr Poepoe suggested to me a derivation of ho-ao for which there is much to be said. Before a marriage of the highest grade the betrothed couple were placed under keepers in separate establishments and were kept very stri'ctly till, at the time of the marriage, they came before the people and it is suggested that it was this bringing before the people, or bringing to light those who had previously been secluded, which gave its name to the ceremony. (See David Malo, op. cit. p. 179.) Mr Ray suggests that according to this derivation the word should be hci'o-ao.

used\ Among the chiefs it seems that royal succession was counted in both lines and, as we have seen, chieftainship of the highest rank depended on both parents having been of royal rank and in these cases it is certainly not correct to speak of succession as matrilineal. As we have seen also, the grade of the marriage and the rank of the children depends quite as much on the rank of the father as on that of the mother, and the only case in which it seems applicable to use the term matrilineal succession is when a child has only one parent of chiefs rank. When it is the mother who has the rank of a chief it seems that the offspring belongs to the chiefs, though it is recognised that the chiefly rank is stained by the plebeian condition of the father. If, on the other hand, it is the father who is of higher rank in a mixed marriage, it seems that the offspring is not counted as a chief at all, and the Hawaians seem themselves to recognise that this is due to the uncertainty of male parentage.

It may be noted that in the account of Mr Poepoe, the marriage called laa-uli alii, in which a chief of high rank marries a chiefess of lower rank, is given before that in which a chiefess of high rank marries one of lower rank, the kaukau alii marriage. If this order is significant it would seem to imply that it is the rank of the husband rather than that of the wife which determines the position in the series of royal or princely Hawaian marriages. I do not know, however, whether any importance can be attached to the exact order in which the marriages in question were placed as they were recounted to me.

I am not aware of the existence of any data which would show whether there were other features of social organisation which can have made the mode of succession important. One would suppose that there must have been laws to regulate the inheritance of property among the chiefs but I do not know how far this kind of inheritance was on the same lines as the succession to royal rank.

Among the ordinary people it seems to be doubtful whether there existed anything which could be called descent or inheritance because there was little or nothing to descend or be inherited. There were no clans and there seems to have been nothing of the nature of individual property,

• According to my definition (see p. 15), it would be incorrect to use the term " descent " in this case.

Individual Marriage 385

the whole of the land and all goods being absolutely at the disposal of the chiefs. Nevertheless, there must have been something which was passed on from person to person, even though it may have been liable to the claims of the chiefs, but I failed wholly to discover what was done in such case.

The evidence thus shows clearly that in Hawaian society in its ancient condition there existed the institution of individual marriage, though undoubtedly accompanied by much sexual laxity and with the possession of marital rights by others than the husband. This conclusion is supported by several other features of the social order. There is the great richness of the relationship system in terms expressive of connections by marriage. There are definite terms for many relationships by marriage and further there is the t&vmpuluna which can have had no meaning except in a community in which marriage was a definite institution. There is also the equal certainty about male and female parentage to which I have already referred. Further, I obtained an account of an incident from an old man which is difficult to reconcile with any other state of society than one in which there was a very definite union between a man and a woman.

This incident was related to me by an old man named Naonohielua. The mother of this man, Kekaula, was regarded as the daughter of Kikipani and Malai and was so described in the pedigree as it was given to me.

MALUAi = Paku Kikipani = Malai

I KUMAIKU = Kekaula

I Naonohielua

Her actual parentage was, however, different. Kikipani and Malai after some years of married life had no child, so Maluai, the husband of Kikipani's sister, was called upon and the result was the birth of Kekaula. Maluai was the kaikoeke of Kikipani but became his punalua when his services were thus used. Naonohielua is now about 74 years of age and his mother, if alive, would be probably about 100. We have thus the record of an incident which occurred a hundred years ago and the fact that another man was expressly called in to beget a child for a couple without issue seems to show

that sexual relations external to marriage cannot have been so constant a feature of Hawaian marriage as is usually supposed. Naonohielua is an old man whose memory was failing to some extent and in order to get a clear idea of the exact relations of the five persons concerned he arranged five sandals, two pairs representing Kikipani and his wife on the one hand and Maluai and his wife on the other, while the fifth sandal represented Kekaula. His relation of the whole incident was quite spontaneous and he was obviously much interested in this story of the old days.

I refer to this matter at some length because it has been supposed that Hawaian society was characterised by relations between the sexes which approached a condition of complete promiscuity without any of the restrictions which are almost universal among other races, a view which was supported by the undoubted existence of consanguineous marriages among the Hawaians. We have seen, however, that there is every reason to believe that consanguineous marriages only took place among the chiefs and among them perhaps only for a special purpose, and there is no reason whatever to suppose that we have in the consanguineous marriages of the Hawaians any indication of a state of society characterised by the general existence of such marriages. Further, there seems to be definite evidence of the existence of wide restrictions on marriage, much wider than those of the civilised world, which were dependent on ties of kinship of the same kind as those found among many peoples of rude culture.

At the same time there can be no doubt that in general the sexual relations of the Hawaians were lax and that side by side with the presence of individual marriage as a social institution there existed, especially among the chiefs, a state of very general laxity, while the claims of the chiefs on the women of the ordinary people (and of the women of chiefly rank on the men of the lower class) extended this laxity to the relations of the whole community.

Part of this laxity was due to the existence of certain definite institutions. One was the claim which a chief had over the virgins of his district, a right allied to that of the/W primae noctis which appears to have existed in the Hawaian Islands, but without the limitation which is implied in the title of that right. Another and more important institution is that of the punalua. There existed among the Hawaians a definite

Cicisbeism

Z^l

system of cicisbeism in which the paramours had a recognised status. Of these paramours those who would seem to have had the most definite status were certain relatives, v\z. the brothers of the husband and the sisters of the wife. These formed a group within which all the males had marital rights over all the females, and I was told that even now, nearly a century after the general acceptance of Christianity, these rights of a punalua are still sometimes recognised, and give rise to cases which come before the law courts where they are treated as cases of adultery. In addition to these punalua who had a recognised status owing to their relationship to the married couple, there were often other paramours apparently chosen freely at the will of husband or wife.