Concluding Remarks. No Fijian custom has given me so much trouble to ascertain it as that of the Nanga; and, since the facts are new, and of considerable importance, it may be well to state the authority on which they rest. Vague rumours of strange practices among the hill tribes of Navitilevu reached me long ago from time to time, but it seemed to be impossible to get at any definite information concerning them. For several years I tried in vain to ascertain them through our native mission agents who were stationed in the hill country. They told me the people either pretended ignorance, or refused to speak because they were bound to silence by "oaths to the dead;" and yet tantalising disconnected bits of information leaked out now and then. The old men especially were very reticent. They were evidently afraid to speak, and when by persistent questioning I got some fragment of description from them as to the less important parts of the subject, they were very uneasy, and anxious to escape from me. About a year ago, Nakorovau, the artist who drew the ground plan of the Nanga, and who is an old pupil of mine, sent me an account of the initiation ceremony which he had managed to extract from a young Vilavóu, who had been made a Lewe ni Nanga at the last celebration before the tribe abandoned the practice. But the account was incomplete, there were many points in it which called for further inquiry, and it did not explain all the facts-e.g., those connected with circumcision -which I had already in my possession. After much fruitless inquiry, I happily fell in with an elderly Wainimala man, who had been one of the Vunilōlō matūa, and whose confidence I won by giving him a detailed account of the Australian Bora. He listened with all his ears, and with eyes opening ever more widely. Presently, while we were talking, a woman passed by, and lowering my voice, I said, "Hush! the women must not hear these things." This finished him. Covering his mouth with his open hand he said earnestly, in an awestruck tone, "Of a truth, sir, you are a Lewe ni Nanga. I will tell you all about it." And he poured out his soul. As he warmed in the narrative, his eyes sparkled, his lips quivered with excitement, his body swayed to and fro, and his arms waved in quick gesticulation such as is never seen among the coast tribes. Unfortunately, I had only one interview with him; and though I learned much from him, there are still many things to be sought out. ་ This rite seems at least open to interpretation as a remarkable case of consanguine marriage" being kept up as a ceremonial institution.-[E. B. T.] Sufficient, however, has been ascertained to convince me that the Wainimala Nanga is identical with the Australian Bora, and that, as I stated in my introductory remarks, it probably connects with the so-called club or secret society, which is found throughout Melanesia as far at least as New Britain, where it is called the Dukduk. But there is one important distinction between the Australian usage and that of Wainimala which must be noted. In some, at least, of the Australian tribes a young man is not permitted to take to himself a wife before initiation, nor indeed until he has served his subsequent probation of two years: whereas, among the Wainimala people, he may take possession of the girl who was betrothed to him in infancy, without any further ceremony, as soon as he considers her to be old enough. He announces his intention to his comrades, and they watch for an opportunity of seizing the girl in the forest, when the husband takes possession of her, but, (as far as I have been able to learn) without the "expiation for marriage"-to use Sir John Lubbock's convenient and appropriate term-which is accorded by so many Australian tribes. The customs of the clubs or secret societies of the New Hebrides, Banks, and Solomon groups, when ascertained, will doubtless throw more light upon the Nanga mysteries. For what we may call the Nanga tribes in Fiji are evidently, both in custom and in language, more Melanesian, so to speak, than the rest of the Fijians. According to their traditions they came into the hills from the West. Establishing themselves first at Nandi, they made their way throughout the country shown in the map until they formed a connection with the southern coast at Korolevu. And the fact that they found the Kai Muáirā in the Wainimala country before them shows that they were immigrants of a later date. The temporary licence between the sexes, and the suspension of proprietary rights in general, during some of the Nanga ceremonies, is well worthy of note. What does it mean? And why, both in Fiji and in Australia, is it resorted to as a means of expiation? We cannot for a moment believe that it is a mere licentious outbreak, without an underlying meaning and purpose. It is a part of a religious rite, and is supposed to be acceptable to the ancestors. But why should it be acceptable to them unless it were in accordance with their own practice in the far away past? There may be another solution of this difficult problem, but I confess myself unable to find any other which will cover all the corroborating facts. Description of Plates I and II. PLATE I. View of the Nanga, or Sacred Stone Enclosure, of Wainimala, Fiji. From a sketch by Mr. Leslie J. Walker, Civil Service, Fiji. Sketch-Map of Na Viti PLATE II. Levu, showing the boundaries of the Nanga Districts. Drawn by Mr. J. P. Thomson, C.E., M.I.S., Government Surveyor, Fiji. The Nanga districts are shaded. The following paper was read by the author: On the LANGUAGES of MELANESIA. By the REV. R. H. CODRINGTON, M.A. By Melanesia is here meant the chain of groups of islands of the West Pacific which stretch in a kind of curve from New Caledonia to New Guinea, west of Polynesia. The boundary eastwards is very well defined, Fiji being as plainly Melanesian as Tonga is Polynesian. New Guinea is not included in the present consideration, partly from want of knowledge of its languages, chiefly to keep away from the use of the term Papuan. I desire not to name the languages of Melanesia, with which I am acquainted, after a people or a country of whose languages I know very little. The groups of the Melanesian Islands are 1. New Caledonia, with the Loyalty Islands. 2. The New Hebrides. 3. The Banks' and Torres Islands. 4. Fiji. 5. Santa Cruz, and the Reef Islands. 6. The Solomon Islands. My own acquaintance with the languages of these islands is limited, but it extends to each of these groups, and covers a good deal of the ground from the Loyalty Islands to Ysabel in the Solomons. Beyond that I have the valuable addition of the language of Duke of York Island, between New Britain and New Ireland, by the kindness of Mr. Brown. From within the limits of Melanesia, as thus defined, certain places with their language have to be withdrawn from consideration. They are those in which the language is Polynesian, in fact Tongan; part of Three Hills, Futuna and Aniwa, Fila, in the New Hebrides; Tikopia, some of the Reef Islands of Santa Cruz; Rennell Island and Bellona Island, south of the Solomon group, and Ongtong Java to the north. The presence of these distinct Polynesian outliers in Melanesia presents no difficulty, and is interesting. For the purpose of the present paper it is important to observe that the Melanesian languages in immediate proximity to these Polynesian settlements, show no more Polynesian character than those that lie far away. The language of Mae, in Three Hills, is in fact Tongan; that of Sesake, two miles off, is known to me, and certainly in my view is no more Polynesian than the languages of the Banks' Islands; which again are very much more like Fijian, than that is like Tongan, its very much nearer neighbour on the Polynesian border. The object of the present paper is to set forth the view that the various tongues of Melanesia are homogeneous, belong to one common stock; and, secondly, that this stock is the same to which the other ocean languages belong: Malayan, Polynesian, the languages of the islands that connect Melanesia with the Indian Archipelago, and Malagasy. The view which is opposed is one according to which the original Melanesian stock of language is distinct from that to which Malay and Polynesian belong; the theory according to which whatever in Melanesian languages is found common with Malayan and Polynesian is said to be introduced from, or due to influence from, either Polynesian or Malay, as the case may be. I am very far from denying that words have been introduced, and language influenced, from the Polynesian or the Malayan side, though I do not think the modern Malay of commerce has reached Melanesia. What I believe is, that whatever has been introduced, from the one side or the other, into the Melanesian tongues, has been introduced not from a foreign but a kindred stock. I shall endeavour to give briefly my reasons for this belief. The first view of the Melanesian languages, as a whole, shows a surprising assemblage of tongues differing so widely among themselves that within very short distances they become mutually unintelligible. In these are found, by those who are acquainted with Malay on the one hand, or Polynesian languages on the other, a number of words and forms which they recognise as familiar, and naturally take to have been introduced. It is not difficult to conceive how the Melanesian languages may have become so very different, if the present inhabitants are supposed to have arrived at their present seats at different times by various routes, and to have had little intercourse between them selves. But the differences in some cases are so very great that it is not easy at once to believe them all of one stock. Further acquaintance, however, diminishes the sense of difference; languages to the ear very unlike are seen on paper to be varying forms of the same. In this way, as the circle of acquaintance widens, languages are learnt more and more to be alike. Finally, in my own case, I may say that of the more than thirty Melanesian languages I have examined, there is not more than one (of which I know very little indeed) which still seems to me to stand much outside the groups into which the others have arranged themselves. I may give as an example the language of Santa Cruz. Any Melanesian language seemed easy to Bishop Patteson, but he was never able, for lack of sufficient intercourse, to make acquaintance with that of Santa Cruz. Within the last few years it has become accessible; and though very strange at first, with a very different vocabulary, and with curious phonetic changes, it soon showed itself as familiar in its main structure, it arranged itself on the lines of the other Melanesian languages. The same has been the case with me in every Melanesian language I have become acquainted with. There are groups, as Fijian, Banks' Islands, the nearer and further Solomon Islands, which, differing among themselves, come on the whole near together. some, till they are examined, seem strange and widely different, such as the Loyalty Island languages, Ambrym, Santa Cruz, Savo; and, when they are examined, show their family connection. But There is one characteristic of some of the Melanesian languages which again causes them to appear of a very distinct family. If any one, for example, should approach the Southern New Hebrides from New Zealand, he finds not only a vocabulary generally very different, but a very rugged consonantal form of words, strongly contrasting with the very vocalic Maori he has left. The difference, therefore, between Melanesian and Polynesian languages seems extreme. But when the whole Melanesian language field is surveyed, it is found that Fijian and the Solomon Island languages generally refuse to close a syllable, and that some of the Solomon Island languages are as vocalic as the Polynesian poorest in consonants. More than this, in a little district of the Banks' Islands there is one language, Motlav, which throws out every vowel it can; while its neighbour within three or four miles, Volow, substantially the same, is almost as vocalic as Fijian. It is impossible, therefore, to regard vocalic character as a mark of difference. In the matter of vocabulary, the Melanesian languages are seen to have a great variety among themselves, and also to have VOL. XIV. D |