MELANESIAN LANGUAGES CODRINGTON \^ ITonbott HENRY FROWDE Oxford University Press Warehouse Amen Corner, E.G. :R W E S I A MARSHALL I iko° KllSrGSMILL IS s'li.cauz GH,otrp TOEBESrj^ I? BAUKS I» MEW V li HEB iltlES .LOYALTY K Zi/hO 0-; Codrvn^tfTTLa Mela>Lesu3L7h. Ltuujua^i CZarendoTv Ji"«ss. Cto%na THE MELANESIAN LANGUAGES BY E. H. CODEINGTON, D.D. OP TBCE MELANESIAN MISSION FELLOW OF WADHAM COLLEGE, OXFOED AT THE CLARENDON PRESS M DCCC LXXXV [All rights rescrved'\ 150° CAELOIIK :fE§iiA 350° THE MELANESIAN LANGUAGES E. H. CODEINGTON, D.D. » f OF THE MELANESIA^ MISSION FELLOW OF WADHAM COLLEGE, OXFORD or THE L'NIVERSITY >itf0rb AT THE CLARENDON PRESS M DCCC LXXXV [All rights reservcd'\ GENERAL PLCZ0 5 /^ PREFACE. I HAVE endeavoured in the following pages to cany on the work of Bishop Patteson. He brought to the philological study of the Melanesian languages an extraordinary linguistic faculty,'^ which enabled him to use very many of them with ease, but he left little behind him in print or in manu- script. In the year 1864 he printed privately some outline grammars or grammatical notes, and in 1866 phrase-books, and Vocabularies of Mahaga, Bauro^ and Sesake, which have furnished material for the Melanesischen Sprachen of Von der Gabelentz. In the latter year also phrase-books were printed in some of the languages of the Banks' Islands. I have not taken these as the foundation for my own work. I never had the advantage of studying them with Bishop Patteson, and I know that he considered them im- perfect and tentative. It seemed better to work indepen- dently on materials obtained directly from natives of Mela- nesia, and afterwards to compare my conclusions with those of the Bishop where the subjects were the same. Bishop Patteson, therefore, is not answerable for the general views concerning the Melanesian languages here put forward, nor for the structure and arrangement of the Grammars ; but I can never forget that I owe any knowledge of these languages that I may possess to the impulse towards the study of them and to the direction which I received from him in the first instance ; and I desire to make all that I have been able to do a memorial of gratitude and affection to him. We have in our Mission school in Norfolk Island from time to time boys and young men from many of the Melanesian vi Preface. Islands, who all come to know and use^ more or less, the Mota language. From these natives of the various islands, and by the medium generally of the Mota language, I have ob- tained what I now offer ; with the exception of what concerns the Duke of York, Rotuma, Sesake, and Fate languages. There is an evident advantage in a method that is throughout more or less comparative ; and in some cases my interpreters have been able to speak more languages than their own and Mota. In this way I received the greatest assistance from my friend and pupil the late Edward Wogale, a native Deacon, who used with much intelligence his knowledge of the languages of the Eanks' Islands, Torres Islands, Florida and Fiji. Great, however, as are the advantages of Norfolk Island for gaining knowledge of the Melanesian languages, the want of books and of communication with scholars is much felt in so remote and isolated a place ; and I hope in consideration of such difficulties that some indulgence will be allowed to the many defects of which I am sensible. Since my return to England I owe much to the kindness and learning of my friend the Rev. A. L. Mayhew, of Wadham College, who has saved me from many errors and helped me in many difficulties. In conclusion, I venture to say that I put forth my con- tribution to the knowledge of these languages with a certain desire to show that in my long absence from Oxford I have not been altogether idle or unmindful of my connexion with the University and College to which I owe so much. Wadham College, Aug. I, 1885. CONTENTS. I. INTBODUCTION. The Groups of Melanesian Islands. 2. The Melanesian languages here in view. 3. Polynesian settlements in Melanesia; their language not in view. 4. The Melanesian languages homogeneous, and of a common stock with the Ocean languages, those of the Indian Archipelago and Polynesia. 5. Great and conspicuous differences between the various Melanesian languages tend to disappear on closer view. The most ex- ceptional among them not Australian. 6. Proof of kinship to be found in Vocabulary,'Grammar, and Phonology, whether between the Melanesian languages themselves or between these and the Ocean languages gene- rally. Examples of the use of Vocabulary. 7. Common Numerals and systems of Numeration, 8. Examples of proof from Grammar; suffixed Pronouns, and Verbs. 9. Great differences among Melanesian languages to be accounted for by the probable history of the population of the islands. Note. Theory of the connexion of the Melanesians with the Ocean races pp. 3-35 II. VOCABULAEIES. Mr. Wallace's Vocabularies in his ' Malay Archipelago.' Examination of his nine words in fifty-nine languages. Seventy words in forty Melanesian languages. Notes on the Vocabularies .... pp. 36-100 III. SHOBT COMPARATIVE GRAMMAR OF MELANESIAN LANGUAGES, Comparison of Melanesian languages with Malay, Malagasy, and the Maori of New Zealand. Absence of Inflexion and formal Parts of Speech, i. Demonstrative Particles. 2. Articles. 3. Personal Articles. 4. Pro- nouns. 5. Personal Pronouns. 6. Suffixed Pronouns. 7. Possessives. 8. Interrogative Pronouns. 9. Demonstrative Pronouns. 10. Nouns, II. Independent forms of Nouns. 12. Classes of Nouns. 13. Construct Nouns. 14. Prefixes to Nouns. 15. Collective Nouns. 16. Redu- plication of Nouns. 17. Plural of Nouns. 18. Prepositions. 19. Ad- verbs. 20. Adjectives. 21. Verbs. 22. Verbal Particles. 23. Verbal Suffixes. 24. Prefixes to Verbs. 25. Reduplication of Verbs. 26. Pas- sive Verbs pp. 101-192 IV. PHONOLOGY. I. Alphabet. II. Phonetic Changes: — i. Gutturals; 2. Dentals; 3. La- bials; 4. Compound Consonant ; 5. Nasals; 6. Liquids; 7. Sibilants; 8. Aspirates; 9. Metathesis; 10. Vowels; n. Phonetic Character pp. 193-219 Vlll Contents. V. NUMEKATION AND NUMEBALS. I. Systems of Numeration. 2. Grammar of Numeration. 3. Peculiar Methods and Terms. 4. Numerals pp. 220-251 VI. GRAMMARS. Banks' Islands — 1. Sugarloaf Island, Mota 2. Saddle Island, Motlav . II. Volow Pak „ Leon and Sasar „ Vuras . „ Mosin . Alo Teqel Star Island, Merlav Sta. Maria, Gog . „ Lakon Bligh Island, Norbarbar Torres Islands — 14. Lo 4. Vanua Lava, 5. 9. 10. 11. 12. III. North of Fiji— 15. Rotuma IV. Neio Hebrides — 16. Aurora Island, Maewo 17. Lepers' Island, Oba 18. Pentecost Island, Arag 19. Espiritu Santo, Marina 20. Ambrym 21. Three Hills Island, Sesake Tasiko, Lemaroro, Tonoa. 22. Sandwich Island, Fate 'Nuna, Anaiteum, V. Loyalty Islands — 23. Britannia Island, Nengone . VI. Santa Cruz — 24. Sta. Cruz, Deni . 25. Nifilole .... VII. Solomon Islands — 26. San Cristoval, Fagani . 27. „ Wawo . 28. Contrari^t(5 Island, Ulawa 29. Malanta, Saa 30. Florida, G'ela 31. Guadalcanar, Vaturawa 32. Ysabel Island, Bugotu 33. ,, to the east. Fiji lies detached to the eastwards, and ap- proaches very nearly the limit which divides Melanesia from Polynesia. Althoug-h Polynesia is often made to include the Melanesian islands as far as New Guinea, yet if Melanesia is to be the name of the reg-ion defined above, as undoubtedly the lang-uag'es and the people are separated by a clear line of division from their eastern neig-hbours, it is desirable to use the term Polynesia strictly to indicate the reg*ion of the East Pacific, to the west of which Melanesia begins mth the Pijian g-roup. To the north of Melanesia lies the region of small and scattered islands which are comprised in the name of Micronesia. The attempt here made to give an account of the languag"es of Melanesia does not include, except in the way of occasional reference, the languages of New Guinea. Whether the in- habitants of Melanesia can be all called Papuans or not, it is clearly desirable to avoid the use of the name Papuan when the lang'uag-es of Melanesia and not of New Guinea are in view. That some of the languages of New Guinea, e. g*. Motu, are Melanesian is clear, the vocabulary of a very distant part, such as Mafoor, contains a larg-e proportion of words common in Melanesia, and by no means all of these Malayan or Polynesian ; but the languag-es of New Guinea have not been available for consideration and examination tog-ether with those of Melanesia in the narrower acceptation of the word. Of these languages, those of the great curve stretching- from New Guinea, beginning with New Britain and New Ireland and ending in the Loyalty Islands, it may perhaps be said that a g*eneral representation is here g-iven. The lan- guage of Duke of York Island, which has been kindly communi- cated by the Rev. George Brown, may be taken to represent those of the great islands between which it lies. The likeness of this language to those of the Solomon Islands and of the New Hebrides, and its wholly Melanesian character, tog-ether with the Melanesian character of such a New Guinea lan- g-uag'C as that of Motu, warrant the assumption that the space between New Britain and Ysabel is occupied by not Introduction. 5 dissimilar languages. The personal knowledge and enquiries of the writer begin with Ysabel to the north, and extend, with serious deficiencies here and there, to the Loyalty Islands. The Fiji language is within easy reach in Hazlewood's Gram- mar and the translation of the Scriptures ^ A Grammar and Dictionary of the Anaiteum language has been put forth by the Rev. John Inglis^. The two treatises on the Melanesian Speech of the elder von der Gabelentz deal with many of the languages included within the limits above specified^. The materials were supplied to him to some extent by Bishop Patteson, and the same materials have been employed here ; but in whatever case the same language has been dealt with, what is put forth here is either, as in the case of Nengone, the result of independent enquiry from natives of the place, or, as in the case of Wawo in San Cristoval, the representation of a dialect not the same as that which has been given by von der Gabelentz. It may be confidently hoped that a view of languages taken from within, that is, by means of a native language in which Melanesians give an account of their own speech, has certain advantages over a view taken, with greater intelligence and more knowledge of language generally, from without, that is, from printed books. A Nengone man, for example, who can speak Mota will probably be able to ex- plain some things to an European who can speak Mota, which may be misunderstood even by an European who can speak Nengone himself. The Melanesian languages, like all kindred languages, explain one another, and appear in the light when they are viewed one with another. At the least, all the groups of Melanesia are represented here, at the furthest point west- wards by the Duke of York Island, at the fui-thest southern ex- tremity by Nengone ; and if the regions towards the extremities are comparatively unexplored, there is a tolerably complete ' I am bound to take the earliest opportunity of expressing my sense of what I owe to the assistance of the Eev. Lorimer Fison, late Missionary in Fiji, in the discussion of the various problems that arise in the comparison of the Fijian with other Melanesian languages. ^ AVilliams and Norgate. London, 18S2. ^ Die Melanesischen Sprachen. Leipzig. 1873. Melanesian Languages. investig-ation of the central part in the languages of tlie Northern New Hebrides, the Banks' Islands, Santa Cruz, and the Southern Solomon Islands. 2. It will be convenient to give a list here of the languag-es of Melanesia which are brought into comparison in the intro- ductory treatises of this book, and of which Grammars, or outlines of Grammars, are subjoined. Beginning at the ex- tremity furthest from New Guinea : — 1. Loyalty Islands. 1, Nengone or Mare. 2. New Ilehrides. 2. Fate, Sandwich I.^ 3. Sesake, Three Hills. 4. Ambrym. 5. Espiritu Santo. 6. Araga, Whitsuntide or Pentecost. 7. Oba, Lepers' I. 8. Maewo, Aurora. 9. Merlav, Star I. 10. Santa Maria, Gaua, or Gog. 11. „ Lakon. 3. Banks Islands. 12. Vanua Lava, Pak. 13. „ Sasar. 14. ,5 Vureas. 15. „ Mosina. 16. „ alo Teqel. 17. Mota, Sugarloaf I. 18. Saddle I., Motlav. 19. „ Volow, 20. Ureparapai'a, Bligh I. 21. Lo. 22. Rotuma. 23. Deni, Santa Cruz. 24. Nifilole. 25. Ulawa, Contrariety I. 26. Malanta, Saa. 27. San Cristoval, Wa«o. 28. „ Fagani. Tlie sketch of this Grammar is drawn from the translation of a Gospel, Torres Islands. N. of Fiji. Santa Cniz. 7. Solomon Islands. Introduction. 7 Solomon Islands 29. Guadalcanal", Vaturawa. {continued). 30. Florida. 31. Savo. 32. Ysabel, Bngotu. 33. „ (?ao. 34. Duke of York. Some of these are but dialects differing not much from one another, as those of Vanua Lava here given ; but there is much instruction in the comparison even of dialects philo- logically and geographically very close. Vanua Lava, an island fifteen miles long, had, before its depopulation by the labour trade, fifteen dialects recognised, as distinct by its in- habitants : it was worth while to preserve as much as possible of so characteristic a specimen of Melanesia. Other lan- guages, though very near together in one island, as those of Gog and Lakon on Santa Maria, are not less valuable or less characteristic because they differ so widely one from the other. The absence of the Fijian language from the above list leaves, no doubt, a great incompleteness in that general view of the Melanesian languages which might otherwise be thought to be given. But the language of Fiji, so much the most important of all, is so well known as not to need what it would be a presumption on the part of one not practically acquainted with it to offer. Much, no doubt, remains to be learnt about it by the study of dialects and by the com- parison of other Melanesian languages, for which materials may be here supplied. 3. From the limits of the Melanesian languages as defined above, the language of the Polynesian settlements in Me- lanesia has to be -withdrawn. The distinction between this and the Melanesian is ever}"^herc plain, and there is very little distinction apparently to be made of dialect in the speech of one settlement and another. These Pol}Tiesian out- liers are to be found in Uea, one of the Loyalty Islands ; in Futuna, a small island of the New Hebrides ; in Fate, Sand- wich Island ; in some of the islets of the Sheppard group, and 8 Melanesian Languages. notably in the settlement of Mae in Three Hills ; in Tikopia, north of the Banks' Islands, and in several of the Swallow- group near Santa Cruz ; in Rennell and Bellona, south of the Solomon 'Islands, and in Ontong Java, near Ysabel. The language of these is said, on good authority, to be substantially that of Tonga, and the same throughout ; speakers of the Maori of New Zealand can understand it and make themselves understood ; it has nothing directly to do with the Melanesian languages^. The existence of these Polynesian settlements, however, in the midst of Melanesia cannot fail to suggest ques- tions of interest and importance which it is impossible to dis- miss without consideration. As to their origin, it is not difficult to conjecture what it has been. Canoes accidentally drifting or blown away, or expeditions purposely directed to known islands, have landed small parties of Polynesian people either on uninhabited places or on islands occupied l)y Melanesians. Some at least of such settlements may be supposed compara- tively modern. If such islands as Rennell, Bellona, or Tikopia have been reached, remote from any large Melanesian island, the colonists naturally remain purely Polynesian in language, habits, and physical characteristics, for there is no admixture. If a single canoe, or a small male party, has found its way to an inhabited Melanesian island, the Polynesian element has been absorbed, leaving perhaps only some fairer and more straight-haired children as an evidence of mixed blood ^. In the case of such a settlement as Mae the case is diiferent. The middle part of that island, one only about six miles long, is occupied by people whose speech is that common to all these Polynesian settlers, but who physically are not dis- tinguishable from their neighbours who are Melanesian both * Some few years ago a whaler picked up in the Solomon Islands and brought down to Norfolk Island some natives of Mae and of Fate, nirvivors of a crew massacred in Ongtong Java. They belonged to the Polynesian settlements, and they told me that they, the Mae and Fate men, spoke the same language, and also understood that of the Ongtong Java people. "^ I have seen myself in Ureparapara a man and woman with a son, drifted thither from some Polynesian island; and I have noticed straight-haired chil- dren in Saddle Island who were known to be descendants of Polynesian cast- aways. Introduction. 9 in language and physical character. The same is the case in the Swallow Islands : the inhabitants of islands close together speak either a language like that of Santa Cruz or the Poly- nesian ; but they are all alike Melanesians in appearance. The Tikopians, an isolated Polynesian settlement, are wholly unlike Melanesians, — tall, hea\y, light-coloured men, with straight hair. The reason why \^q Polynesian-speaking people of Mae, for example, are Melanesian in appearance clearly is that the INIelanesian blood in them has overborne the Polynesian element ; that is to say, the Polynesian settlers have, generation after generation, taken Melanesian wives into their villages in which the speech was Polynesian. The speech, the descent of chiefs, certain relig-ious practices, have remained Polynesian, the physical aspect has gradually lost its original character. Under such circumstances the speech which will be permanent is the speech of the settlement ; the physical character that will prevail will be that of the blood. Hence the Tikopian is physically and in language purely Polynesian, the Fileni man of the Swallow group is in speech Polynesian but physically Melanesian. The phenomena of the case are thus explained^. It remains to state another remarkable fact. In Three Hills Island, Mae, the Polynesian settlement above men- tioned is about two miles distant from Sesake, at one end of the island, occupied by those who may be called the aborigines. The Mae language is Polynesian, if not purely at least de- cidedly so ; the Sesake language is Melanesian decidedly, and at any rate has nothing that makes it appear more influenced by its Polynesian neighbour than if Sesake and ]Mae were in ' Some fifty years ago the Banks' Islands were visited in two successive years by double canoes. The people in these canoes said they came from Tonga. They settled the first year for a time on the Islet of Qakea, close to Vanua Lava, quaiTelled after a time with their neighbours, and went off. When they returned next year they were attacked by the natives and driven off. There were women with them. If they had settled on Qakea there would be there now a Polynesian-speaking people, but Melanesian wives from Vanua Lava would be continually bringing in Melanesian phj'sical characteristics. If Qakea had been an isolated place like Tikopia, there would have been then a small purely Polynesian colony. lo Melanesian Languages. different and distant islands. This cannot be too positively stated, and the importance of the fact is very great. It is an exemplification, in a very narrow field, of what is found also to be the case with regard to Fiji. The Fijian group is only some 200 miles west of the Friendly Islands, which are de- cidedly part of Polynesia. There has been a considerable intercourse between the two groups, and no doubt a great infusion of Tongan, Friendly Islands, blood among the higher classes of Fijians. There has been also, according to native le- gends, a considerable intercourse between Fiji and the purely Polynesian Samoa ^. Yet the Fiji language is most decidedly Melanesian ; it has no doubt something directly derived from Tonga, but it is no more Polynesian than the languages of the Banks' Islands, which lie far away to the west, out of reach of any but the most casual and insignificant intercourse with Tongans or other Polynesians. Intercourse therefore and close neighbourhood with Polynesians do not as a matter of fact materially affect the language of Melanesians. - . 4. The view of the Melanesian languages here proposed is, ' in the first place, that they are homogeneous ; and secondly, that they belong to a common stock with the Ocean tongues generally — those of the Indian Archipelago and of Polynesia. The view which is opposed is one which would make the Melanesian stock of languages originally distinct from that to which Malayan and Polynesian languages belong, and would pronounce all that is found in Melanesian languages common with Malay and Polynesian to be borrowed from these tongues, or due to influence received from them. In opposition to this latter view, it is by no means denied that the Melanesian languages have borrowed from those of the Indian or Malay Archipelago on the one side, and from those of the Eastern Pacific on the other, or that they have been influenced in various ways ; allowing this, what is maintained is, that whatever has been introduced has been brought from ' Turner's Samoa, Macmillan, 1SS4, pp. 41, 123, 225, 228, 230, 256. In these stories the Fijians are by no means represented as inferior to the Samoans. See Vocabulary Notes under the word ' Bow.' Introdtutioji. 1 1 languages of a kindred, not a distinct stock. By way of illustration, not of exact comparison, the English language has borrowed largely, directly and indirectly, from Latin and Greek, but still what has been borrowed has been taken from kindred languages. It is not as when Arabic is found in Spanish. There was an archaic flow of Aryan language over Europe, and over a great part of the region so covered more recent waves of the same have passed. There is differ- ence enough between Celtic, Slavonic, Italic, and Teutonic, but they are members of one family. Let as much difference be allowed between the various Ocean families, and let not one be said to borrow from the other without good reason. The Melanesian languages, which are very little known,"" come geographically between the Malay and Polynesian languages, which are well known. Any observer of the Melanesian languages who apj)roaches from the "West and sees in them much that is the same with the Malay, calls that a Malay element, and calls that which he does not recognise the native Melanesian or Papuan element. One who approaches the Melanesian languages from the East finds much that is common with the Pol}Tiesian, and he calls that the Polynesian element, and again what he does not recognise the Melanesian or Papuan. But suppose an ob- server to begin with the Melanesian languages, and, being familiar with them, to advance on the one side to the Pohmesian regions and on the other to the Malayan. He will find in the islands of the Eastern Pacific people of a brown colour, using a language very much of which is familiar to him, but one poorer in sounds, poorer in grammatical forms. He will say that they speak a kind of Melanesian dialect. If, as is conceivable, it had so happened that an English occupation of Australia had made the Fijian language familiar to Australian merchants, officials, and scholars, before English- men had advanced far enough to the East by India to have come in contact with Malay; then as Australian commerce advanced westwards from Fiji, and the native languages were found more or less to resemble Fijian, it would certainly 12 Melanesian Languages. have happened that the Melanesian, the Indian Archipelago languag-es, Malay, Malagasy itself, would have all been found marked by Fijian character, would even by Australian scholars have been said to belong to the Fijian family. It is a matter of chance or circumstance after what member of a family a family of languages is called, just as it is by what name foreigners call another country. The first that comes perhaps is the one that gains the place, and it is very likely that it is quite inappropriate. At any rate, there is a certain fallacy in the natural conclusion that the language after v>'hich a family of languages is named is the standard, the characteristic, and to the unlearned the original, language of them all. If an- other supposition may be ventured on ; suppose America to have discovered Europe and not Europe America, and the American discovei'ers to be in a superior state of civilisation to the discovered Europeans, so that they should extend their acqviaintance with them and observe their languages: they would discover England first it may be supposed, find the Dutch language a form of English, recognise in French an English element in that which was found in common ; when their learned men knew more they would distinguish the languages of Northern Europe from the Southern, but the Northern languages would take their name from English ; German would be classed as Anglic, and not English as Teutonic. It is possible, by the correction of a similar mis- take, that, instead of speaking of Malay or Polynesian elements in Melanesian languages, it would be rig*ht to class INIalay and Polynesian languages as Melanesian. The Melanesian people have the misfortune to be black, to be much darker, at least, than either Malays or PoljTiesians ; and because they are black it is presumed that their original language cannot be of the same family with that spoken by their broMTi neighbours ; that where their language has a general resemblance to that of their neighbours they must have cast off their own and taken another in the lump, and that where the resemblance is not conspicuously apparent they must have borrowed words and expressions in com- Introduction. 13 mercial or other intercourse. With reg'ard to colour it is enouo-h to say, if the matter is to be considered at all in a question of lang-uag-e, that between the black of Melanesians and the brown of Polynesians the difference is not so very great, and that the colour of the inhabitants of the Indian or Malay Archipelago is much the same as that of the Melanesians. There is a great variety of shades of colour in Polynesians, Melanesians, and others ; with quite enough of general distinctive character to throw them into classes, but yet such that among Polynesians are to be seen very dark individuals, as among Malays those who might be taken for Chinese. Given an original dark and frizzly-haired stock, it is not difficult to conceive such a cross with straight-haired and light-coloured men as would' produce a brown and wavy- haired progeny, and beyond that such a series and confusion of inter-crossings as would give a great variety of inter- mediate shades of colour, straightness or curliness of haii", and other physical characteristics. There is no doubt a certain reluctance on the brown side to acknowledge the kindred of the black. The Melanesians are the poor relations, at the best, of their more civilised and stronger neighbours ; but a question of language must be discussed on its own merits, and degrees of complexion or cultivation may be put on one side. Any one who approaches the Melanesian languages with~^ some knowledge of a Polynesian or Malayan language cannot ) fail to find a certain resemblance ; he will find words, perhaps/ very many words, the same ; he may find the Melanesiaa language so much like the one he has been before acquainted with that he will hazard the assertion that it is a corrupt Samoan for instance, or will conclude that commercial inter- course with INIalays has had a great effect upon the native language. Whether, except in the Polynesian colonies above mentioned, he will ever find a language that he can think taken over by Melanesians as a whole to supersede their original tongue is very doubtful. That such a language has been taken over from Polynesians one may say is certainly not 14 Melanesian Languages. the case. In a Melanesian language many words will be the same as the Polynesian, many grammatical forms will be the same ; but undoubtedly, as a rule, the phonetic character of the Melanesian will be fuller than that of the Polynesian word, and the Melanesian grammar will fail in some feature con- spicuous in the Polynesian, the Passive for instance, and will have in vigorous life some power which is at work, but com- paratively little at work, among the Polynesians ; for ex- ample, the definite transitive power of certain forms of verbs. If the inhabitants of any Melanesian island have cast off their old tongue and taken to a new one, which is not here denied, at any rate it has not been one of the Polynesian languages of the East Pacific that they have taken. If the people, for example, of the Banks' Islands have, either in their present seats or in some place from which they came in a distant time, given up their old speech and taken to another, we should not expect to find what they had given up, but we should have every reason to expect to find the source whence they have obtained their present language. Cer- tainly this is not Polynesian, and certainly not Malayan ; looking round for a language resembling the Banks' Islands languag'es we may fix on Fiji as being very similar on the whole. But in Fiji the same process has to be gone through. If the Fijians, Melanesians, have thrown off" their old speech and taken another, whence did they get their modern lan- guage? Most cei"tainly not from their Polynesian neigh- bours. The most remarkable characteristic of the Fiji language, that very efficient suffix of a transitive termination to a verb, which flourishes in this even more than in any other Melanesian language, is absolutely unknown to their Tongan neighbours, and to the Malays also, if in consideration of the immense space which divides them it is necessary to mention the Malays. But the fact that Melanesians, pre- sumed to have cast off" their orig'inal language and taken another, have not taken a Malayan or Polynesian one instead, is no proof that they have not taken any other. It may be that the languages here treated of, those of which a list has Introdttction. 1 5 been given above, and here called Melanesian, because the islands in which they are spoken are Melanesian, are not the original languages of the race that now speaks them. It may be that a third term should be used for the languages which, not Malay, not Polynesian, are now spoken in the Indian Ai'chipelag'o and in the Melanesian islands. Those now in use are the languages here called Melanesian ; and if the ancient language once possibly belonging to the people now inhabiting Melanesia could be found, there would be a great difficulty about nomenclature ; a difficulty which only does not arise because the languages here called Melanesian are the only languages, with the exception always understood of what is spoken in the Polynesian settlements, now found to be spoken in the Melanesian islands. This last assertion, resting on no authority, requires some explanation ; the truth of it can only be assented to, or dissented from, after the study of the languages themselves. 5. The first view of the Melanesian languages no doubt shows great differences between some languages and others ; the learner who has acquired one approaches another as if it were quite a foreign language, with some words only in common with that which he knows. But languages which are mutually unintelligible to natives of parts of the small islands on which they are spoken are often perceived without much trouble to be really not far apart, when once acquaintance has been made with them ; for, without any substantial differ- ence in vocabulary and with little difference in grammar, two languages as closely allied as Dutch and English may have so far diverged in pronunciation and in the use of the vocabulary as to be very different to the ear. As acquaintance with the languages increases, the likeness of one to another becomes more apparent ; a sort of ideal standard is established to which they conform, the specific differences become subordinated to the general character, the difficulty presented by the multi- plicity of forms diminishes more and more. But there will remain some among the Melanesian languages which seem exceptional and hard, not fitting into the neighbouring 1 6 Melanesian Langtiagcs. groups, and having" forms which cannot well be matched with parallel examples. Supposing, for example, that the languages of the Northern New Hebrides are to be learnt by one who has an acquaintance with Mota : it is not long before he sees that the languages of Aurora, Pentecost, Lepers' Island, and Espiritu Santo belong to a type with which he is acquainted, though a knowledge of Mota does not enable him to understand what he hears. But when he reaches Ambrym he finds that the language is much more hard ; there are words indeed that he knows, but he seems a stranger. Beyond Ambrym, in such a language as Sesake, he is comparatively again at home. The same is the case in the Solomon Islands, where round Florida, in Guadalcanar Malanta and Ysabel, the languages which are heard are not very far apart ; but at Savo, which lies in the midst of them, the language seems altogether puzzling: many words indeed are the same and the people identical in customs and appearance, and yet to make out the speech is very difficult. Far beyond Savo, the language of Duke of York Island is again familiar, with hardly anything in it which has not an easily-perceived analogy with the Solomon Islands, Banks' Islands, Fiji, and New Hebrides tongues. If difficult ex- ceptional languages are to be named in the list of thirty-four Melanesian languages given above, they would be Nengone, Ambrym, Santa Cruz, Savo. The very important question then arises whether these differ so fundamentally from the other languages that they cannot group with them as mem- bers of the same family. The lesser question, whether they agree together in difference from other Melanesian languages, is easily answered in the negative : they sometimes agree, but generally do not ; each has its own peculiarities. On the main question the judgment of the writer has been already given when the Melanesian languages have been pronounced to be'^-^homogeneous. As with the Melanesian languages generally, very much as their differences are conspicuous at the first view, greater acquaintance with them makes them appear more and more alike, so with these exceptional Ian- Introduction. 1 7 guages, the difficulties of them do not disappear, they do not range themselves by the side of the others in orderly groups, but the more they are known the more their features show the family likeness. It cannot be said of these that they are the remains of the old Melanesian speech, now in many islands thrown off to make room for a foreign language to take its place. More archaic they well may be, belonging to an earlier movement of population, carried forwards by an earlier wave of speech passing onwards among the islands, but having somewhere a common origin with those which have since and successively passed among them. It is equally clear that these languages, which are rather exceptional among those of Melanesia, do not, in the points in which the exceptions appear, agree with the neighbouring language of Australia. It would seem natural, on the sup- position that there was a language in the Melanesian islands originally which has since been exchanged for another and that some traces of the original still remain here and there, that we should look to Australia with the expectation of finding in the native people and languages the kindred of the orig-inal Melanesian. But in reg-ard to lans-uag-e it must be said that any one familiar with the Melanesian tongues finds nothing but what is strange to him in Australian grammar and vocabulary. Illustrations of this will be ofiered hereafter, but the statement is necessary here that, with regard to words at least, whereas a strange word appearing in a Melanesian language is very often to be found in some distant Ocean vocabulary, no correspondence whatever seems to occur between Melanesian and Australian vocabularies. If then two distinct families of language do not appear in Melanesia at the present time, not including New Guinea in the consideration, and the Melanesian languages are homo- geneous, it further has to be established that they belong to the same family with the Ocean languages generally, that is to say, with the Polynesian, the Malay, the Malagasy, and those of the Indian Archipelago generally. This can otx\j be done here, so far as a very limited acquaintance with those c i8 Melanesian Lanma^es. 2> ' lang-uages can enable it to be done at all, by a comparison of Vocabularies and Grammar. 6. The proof of the kinship of languages must be made in consideration of Vocabulary, Grammar, and Phonology. Some material is here offered towards this end. Seventy words in forty languages of Melanesia are given, which can be com- pared among themselves, to see what proof they give of the substantial unity of the Melanesian languages ; they can be compared further with the same list of words given in thirty- three languages of the Malay Archipelago by Mr. "Wallace in his book on that subject ; and they can be compared also with the corresponding Malagasy and Polynesian words, which will make comparison with the Ocean languages tolerably complete. A comparative sketch of the Grammar of the Melanesian languag^es is given, with reference also to the Ocean languages as exemplified in Malay, Malagasy, and Maori, and in the Marshall group in Micronesia. There is also added something on the phonology of the Melanesian languages, the vocabularies and grammar of which are the subjects of consideration. With regard to Vocabulary it may be observed that the use of it requires always careful con- sideration when proof of the kinship of languages is sought in it. There is always a certain element of chance corre- spondence to be expected, such as may be seen in the words tain, as, si, if, the same in Latin and Mota, or s'lke in Mota, the same in sense with the English seek. There is also to be calculated on the presence of borrowed words. It proves nothing as to the kinship of two languages that many words are found common to both. On the other hand, the pre- sence of very many common words must be allowed to prove something in favour of relationship, and disagreement to a very great extent in vocabulary does nothing to disprove relationship. No one could refuse the evidence of vocabulary in proving the kinship of English and German, and no one is shaken in the belief of such relationship by complete disagreement in the ordinary words for the commonest things, horse, dog, or pig, woman, boy, or girl. If there- Introduction, 1 9 fore in comparing ]\Ielanesian vocabularies among" themselves we find a great deal of agreement, that is not without its value in proving them homogeneous ; if we find difier- ences where we might look for agreement, that does not prove any fundamental distinction. The same is the case in comparing- Melanesian vocabularies with those of the other Ocean languages. For example, if we find iga the common Melanesian word for a fish, it argues something for the re- lationship of the languages which have it ; and if we find the same word common in the Pol3'nesian and Malay Archipelago vocabularies, as it is in the Melanesian, it argues something for the relationship of all these Ocean languages together. But if we find two neighbouring Melanesian languages like Merlav of the Banks' Islands and Aurora of the New Hebrides, one having the word for fish \g and the other mas'i^ there is no proof whatever of radical difference between the two^. There is no reason why a language should not have taken into its use a word for fish which is unknown elsewhere. When, how- ever, in Borneo a fish is masik, as it is masi in Aurora, and many common words are found in Borneo and in Aurora be- sides, although this particular coincidence may be accidental, it is impossible not to admit the fact as going some way to prove that a common stock of words is found in the Malay Archipelago and in Melanesia. Exceptional words have thus their value, as have those which are so commonly the same. It is certain, however, that the presence of words the same in form and in meaning, in two or many languages, does not do anything like so much to prove a common stock as the presence of words either the same in form but differing in signification, or the same in signification but with a change of form. With regard, in the first place, to words which are substantially the same in form in different languages but different in signification, this may be rather apparent than real, it may be only a difference of application ^. W^ords also ^ The words for fish in the nearly related languages of Latin and Greek are radically distinct. ^ ' Would not a man, not well versed in the Teutonic languages, infer from C 2 20 Melanesian Languages. are often present in two languages, or more, but when a vocabulary is being compiled they will appear in one and not in another, because they lie in different levels of the language^. What is the general term in one language is specific in another, what is the common word in one is an out-of-the- way word in another. If English and German vocabularies are compared in the few words mentioned above, Horse and Pferd, Dog and Hund, Pig and Schwein, Woman and Weib, Boy and Knabe, Girl and Madchen, are wholly different words, and the effect of them set side by side is to make the two languages seem distinct. But there are in the English language prad^, hound, swine, wife, knave, maiden, on other levels in the language, some of them only different in application, some with a specific instead of a general signi- fication. Vocabularies such as are gathered by travellers are made up of words taken from the surface of language, and are therefore apt to deceive. Very often a word is in a language but in another meaning, very often it is there but is not ordinarily used : a common stock of words is held by various languages, but in the lapse of time and changes of use they get sorted and re-sorted into various applications and employ- ments, though they may lose very little of their original form. A few examples may be useful, taken from Melanesian and Ocean tongues. The word rangi in Maori is the sky, as in other forms it is in all the Polynesian languages ; the same rani in San Cristoval is rain ; the same in Mota, Ian, is wind ; the same in Fiji, cagi [dhangi) is wind and, more generally, the atmosphere. If the Maori ua, rain, is put against the San Cristoval ram, or the Maori hau, wind, against the Mota Ian., there is nothing to show a likeness in vocabulary. But ^lsa for hut, maar, sondern, and to, naar, zu, that English, Dutch, and German are no kindred languages ? Such words seem different, because they are very liable to different applications.' Outlines of Malagasy Grammar, H. N. van der Tuuk. ' Mr. Fiaon has used the expression that ' words are not in the same focus, microscopically speaking,' using the illustration of infusoria in a drop of water, one of which may be invisible until a turn of the screw brings it into focus. ^ The word palfrey, I am informed, contains the same root ; and wife is in fact contained in the word woman, wife-man. Introduction. 2 1 rain is used in the next island to San Cristoval, and saw in Mota is the blowing of the wind; the words are there, though in the latter ease not in the same application. The common word for blood in Melanesia, as in the Malay Archipelago and in Malagasy, is ra, dara, nara ; the common Polynesian word is toio. But tofo appears in the Solomon Islands in the name of a disease, and as congealed blood, and no doubt is the same with the iofoa, toto, of the Banks' Islands, w^here it signifies the sap or juice of trees. The word tasi, ta/n, tai, is common in Melanesia, Polynesia, and the Malay Archipelago for salt, salt-water, the sea, and a lake. In the Banks' Islands, though tas has gone out of use as meaning salt, or salt-water, the verb tasiff is used for seasoning food with salt-water ; fas is no longer the sea or sea-water, but the names of places on the lee or weather side of islands retain it in that sense : Tasmate, where the sea-water, the surf, is dead or still; Tasnianr, where it is alive or breaking. The lake in Santa Maria is the Tas ^. Again, the Maori tvhehi, Samoan fetii, is no doubt the Mota v'ltu, a star, which, with the termination commonly suffixed, is in the Banks' Islands also vitugl. In Celebes hitui is clearly the same and the meaning the same ; but in Dayak of Borneo hetucJi is 'eye ;' as in the compound betuch anuli, the eye of day, the sun. The word is the same, and the expression is paralleled in the Malagasy maso andro, eye of day, the sun, masoe being a star also in the Banks' Islands, and maso, eye, in Malagasy. To add one more, the word for mouth in Maori, waJia, in Batak baha, Malagasy vava, in Lepers' Island tcaum, does not appear in Mota as a mouth, but in a verbal form is wajijan, to open the mouth. Another Maori word for mouth, mangm, is mana, mouth, in Florida, which is also a word or speech ; and in Mota again is formed into the verb manasag, to make a word or speech about a thing, to make known ^. ^ In Madagascar one of tlie large lakes is Itasy, and on the weather S. E. coast is Tasinioro. These names Tasmate and Tasmaur, in varying forms, are applied to places on weather and lee sides of islands in the New Hebrides and Solomon group. ^ See further these words in the Vocabulary Notes, 2 2 Melancsian Languages. To observe words which are evidently the same under changes of form is equally interesting", and more effective as a proof of kinship between langTiag-es. The word for star, mentioned above, whetu, vitu, hetvch, lifni, is no doubt the Malay Linfang, the Malagasy viniana, n having been intro- duced to strengthen t. It is plain that the Malay bintang cannot have been introduced in the form vitu, vitugi ; the word has not been borrowed by the Melanesians from Malay. Has it then been borrowed from the Polynesians ? There is one objection to the supposition, in the fact that in Fiji, which is nearest to the Polynesians in Tonga, halohalo is the word for star, not vitu ; but, as has been observed, there is no greater nearness in speech coinciding with geographical near- ness to the Polynesians. It is in the likeness of the Dayak form to the Melanesian that the argument for the common property of the Ocean tongues in this word mainly lies. The Dayak nouns have this termination ch where the Banks' Islands languages have gi or g\ hurucli, feathers = ««'?unti in SumbaAva, not given by Mr. Wallace. From this word, common to the Indian Archipelago, Madagascar, and Melanesia, it will be observed that the Malay jjizsawgr is altogether different. The Samoan is fa'i, of which no form is Melanesian. In Tongan, the nearest geographically to Melanesia of any Polynesian group, the word is fugi, which certainly is far from the Fiji vudi. But it is the same with the San Cristoval Jmgi and fuki in the Solomon Islands. It is possible that these words hugi, fugi may be another form of the Ceram fud and the Malagasy ontsi, by the substitution of k for t. In languages so close together as Ulawa and Wa7io there may be that very ancient difference between the two forms of the word they have taken up : or the Tongan word may somehow have been conveyed to that small district, being itself distinct ; or again the two words may be distinct, and Tonga and San Cristoval have received fugi from some common source. But it will be observed that if forms with t and k ai'e ancient forms of the original root, the isolated presence of the form with k in San Cristoval and Tonga goes to show that each received the word from a common source. If so, in the great area over which the word is Notes oil the Vocabularies. 55 spread it has divided into three distinct forms, with t, with n, with h., all of which are in Melanesia. The common word in the Banks' Islands, vetal \ is local. The Ambrym vi is remarkable as corresponding to jia of the Sula Islands, and lios of Anaiteum may well be husa of Sanguir. Agree- ment in exceptions between the Vocabularies of the Malayan and Melanesian Archij)elagos is as instructive as agreement in a generally common word ; it tends to show that of an original stock of words carried hither and thither abroad some have survived here, some there, as a witness to original unity. 4. Belly. — Let it first be observed that this English woi'd means that which bulges ; and also that when a word is got by pointing at an object there is often a confusion, which in this case may be between the protuberance in view and its contents, between, that is, belly and bowels. In Mr. Wallace's Vocabulary for this word there are many words like tia, and these evidently got by pointing, for they have suffixed Pronouns, tiaka my belly, tiamo thine, Man his, tiare theirs. In Malagasy tsinay is bowels, as tinae is in the Banks' Islands ; it is natural enough to use the word in one sense or the other : but no doubt the word tinai originally belongs to the bowels. In the same way it is certain that the Banks' Islands toq^ai refers to the protuberance or curve, though at Saa in Malanta, where ii is belly, 'ocia is bowels, as toha is in Florida. Both in the Malay Archipelago and in Melanesia tia is used for the belly, though in Melanesia only in Espiritu Santo in the New Hebrides, and in two parts in the Solomon Islands. It is not, however, in the secondary sense of the words but in the primary that the Avords used in the sense of belly are interesting, and important in this enquiry. A word which is used in widely distant parts of the same language area, in such a way as to imply in each use some pi'imary significance, is a very good proof of kinship between lan- guages. Thus assuming the meaning of toqal, toha, to be that which bulges, outwards or inwards, convex or concave, we can understand that the Mota toqalaui the calf of the leg, is parallel to the Mala- gasy kibondranjo. In Malagasy kibo is belly and ranjo (Mota iriiw) is leg, the words for belly are different, for leg the same, but in each case the calf of the leg is called the bulge. So in Fiji, kete is ' The Motlav word is given vtel, an unpronounceable form. This repre- sents the true word without the Article, which in this language coalesces with it, navtel for na vetel. The Anaiteum and Fate words are in the same way divested of the Article. 56 Melanesiaii Languages. belly', but tuba, the Florida form of toqa, is a harbour, where, that is, the shore curves inwards. In Duke of Yoi-k hala is belly, and the same figure is used, hala na waga is a harbour. 5. Bird. — Out of Mr. Wallace's thirty-three words twenty-four are forms of manu, out of forty Melanesian words thirty-three are forms of the same manu, which is also the Polynesian word ^. Tlie agreement over so large an area is very remarkable ; it can hardly be supposed that almost all Melanesian languages and most of the Malayan have borrowed the word from without. The exceptions, however, are still more remarkable, and the consideration of them is very instructive. The Malay vocabulary has not manu ; the languages, therefore, of that Archipelago have not boi-rowed it from thence. The Malay word hurung is the Malagasy vorona, and has no knoAvn corresponding form in J\Ielanesia. There are only two other exceptional words in Mr. Wallace's Vocabulary. The excej)tious in Melanesia, though not numerous, are very interesting. In the first place it must be remarked that there is a remarkable indifctinctness of meaning. This is the case in Fiji, where, as in Polynesia, all beasts as well as birds are called vianumanu, which, in the absence of almost all land animals, is not perhaps so wonderful. In the rest of ]\Ielanesia they do not call beasts birds, except in the Duke of York, where pika is the generic term for all animals, which are various enough comparatively, and pika rowo flying animal, is a bird, rowo, as in Mota, being ' to fly.' But there is in Melanesia an astonishing case of the same word meaning bird and fish ; the word 7na7i in Lakona. And as manu, oven if used for a beast, must be thought to mean properly a bird, so mah, though used of a bird, must be thought to be pi'operly a fish, because in another island of the Banks' Group, and in the New Hebrides, meh, mes, and masi mean fish. We are coubtraiued, there- fore, to think that in these words, manu and muh, we have woi'ds so ancient as not yet, in the absence perbaps of quadrupeds, to be particularized. If a Lakona man were asked how it is known what he means when he says mah, he would answer that everj' one would oindei'stand because they would know what he was talking about ; and if there were any doubt he would add ' of the sea,' ' of the wood,' to explain. It is nearly the same thing in Nengone of the ' It is quite likely that this word is the same with the Mota gete a bag or basket, the Maori kete ; as toqa also in Mota is a particular kind of woven bag. ^ Maan at Mafoor, New Guinea. Notes on the Vocabularies. 57 Loyalty Islands, where ia is u bird and ie is a fish : the words arc too much alike, and too much like i(ja, to he thought altofjcther separate. Wc have to ask what the primitive idea conveyed by manu, mah, iga is, a creature with wings, or variegated in colour or what ? The Mota word for a nose and beak, manui, may suggest that manu means a creature with a beak. Beyond this another question is raised when the locality of the exceptions among the Melanesian words for 'bii'd' is considered. Almost everywhere is 7nanu; but in the Loyalty ia, in Ambrym bwehel, in Santa Cruz kio and deguluo, in Savo kosu, in Gao naji, in New Georgia oloJco, are all very different from one another and from the common word. There is no doubt but that, if there arc jMelanesian languages which stand aj)art from the more common type of language spoken in Melanesia, they are those of the Loyalty Islands, Ambrym, Santa Cruz, and Savo : and it cannot well be doubted that exceptional words like these in the Vocabulary confirm, if they do not iinpi-ess, the belief that these districts have somehow not been reached by a later flow of words. But then in this particular they are not so very unlike Malay itself in having exceptional words for bird. The changes of manu into Espiritu Santo nanu and Alite malu are regular in those languages, as will be shown in the Chapter on Phonology. 6. Black. — The word which in Malay is itam, and with the prefix ma of quality is maita, miti, mete in the Indian Archipelago, occurs in twenty-seven places in Mr. "Wallace's Vocabulary. TJiere is probably no distinction between black and dark colour. The Malagasy mainty is the same as the Malay. No form of the w-ord is at any rate common in Polynesia. In Melanesia three islands of the Northern New Hebrides have it as maeto, meto, naeto, and Sesake maeto, with the meaning of black. In Mota maeto is black volcanic stone ; in Florida m,eto is dirty. In Micronesia, in the Marshall Group, the same appears in the compound kilmed black, probably black skin. The word, therefore, is widely spread, and it is plain that it did not spread from Malay. The chief interest of the Vocabulary is grammatical ; the list of Melanesian words shows a good deal of the form of the Melanesian Adjective. The charac- teristics are three : (i) reduplication, (2) the prefix of quality ma, (3) the adjectival termination ga. Something may be noted concerning individual words in the Melanesian list. The natural connection between night and black- 58 Melanesiaii Languages. ness is shown in the Santa Cruz bo, which may be ^jo, honi, and tlie Ulawa and Malanta roto, and the Anaiteum ajng^. In Mota, siliga dark, is often used for night. The Vureas korJcor is the same word, made adjective by reduplication, as kor a dried bread-fruit, in Mota ; in which language indeed kor, at least in the slang of nicknames, is used for black, l^us-kor Blacklip. The Gaua word wirwiHga is used in Mota, and is indeed another form of silsiliga. This is shown by the parallel forms of the Mota word sinaga food, in Motlav hinag. The change is from h to it), shown in Duke of York whiagan, and as very commonly from I to r. It is out of the question that the change has been made in Gaua, where the language is very like that of Mota. What the word in these two forms points to is an ancient source, fi'om which by once diverging channels the two forms have come in different shape into these neighbouring islands. It is an exemplification of what we may well believe to have been the way in which neighbouring languages have come to differ so often and so much -. 7. Blood. — The word represented by the Malay darah appears more than twenty times in Mr. Wallace's Vocabulary, and again twenty-three times among the forty Melanesian words. The prob- able root is the Malagasy ra, which has become the Nengone and Fiji dra, ndra, and so nara, and dara. The change from r to d and t is shown in Fate and Sesake. We have here, then, a word in very varying form spread over the Indian and ^lelanesian Archi- pelagos, in such a manner that it is impossible to point to any existing source ; for the root form is found in Madagascar. Tliere is hardly a word which does more to show the kindred origin of the tongues. The exceptions are not less interesting and instructive. In Menado of Celebes and in Sauguir the word for blood is dalia, which, having nothing in it of r, cannot be counted as akin to darah. In Araga, Whitsuntide, of the New Hebrides, there is daga, to w'hich probably dai of the neighbouring islands is allied. These exceptions, so widely separated in place, one would hesitate to suppose the same, were it not that identical forms appear else- whei'e : daga in the Philippines, taga in Formosa, daya in Pam- ' In this word a is probably a Verbal Particle, a,ui\. pig night, g = ng, is the Banks' Islands qon. ^ Tlius the English heart and French cceur, hound and ehien, are in fact forms of the same words. Notes on ike Vocabularies. 59 pango, ihijia in Dayak of Sarawak. "We liave, therefore, to believe that this is one of the many words for blood which have been obscured by the general use of r«, and that it has survived in Melanesia isolated in the New Hebrides : dwja has been no more imported into Araga from the Philippines than ra has from Mada- gascar to Fiji. The Polynesian word for blood is toto, which does not appear in the Malay Archipelago at all, and in the Melanesian Vocabulary only as tot in Eotuma, where it is probably an importation. But toto is used in Melanesia, and in such a way as to show that it is at home there and cannot posi^ibly be an importation from Polynesia. In San Cristoval toto is congealed blood ; in Florida the disease hgematuria is mimi-toto ; and though it is not unreasonable to maintain that toto may have been borrowed in those parts, it would be very hard to conjecture how it had been done. But toto in the Banks' Islands is a poisoned arrow, and this can be shown to be the same word. The arrow is called after the tree with the viscous saji of which it has been smeared, and the tree has its name from the abundance of its sap, in Mota totoai, in a dialect of Fiji dotoa. The sap of a tree is its blood, and it is very easy to conceive a word at one time more general in its meaning being particularized to signify in one set of languages blood and in another sap. It assists this view to observe that toto in San Cris- toval is clotted thick blood, like the thick sap totoa. There is another word so common in the Solomon Islands as to deserve notice, in the forms of gahu, habu, abu, apu, ka]), the variations of which argue that the word is no purely local one. 8. Boat. — This word has been taken to mean a native boat or canoe, and this as far as possible generically. In Melanesia each kind of canoe has its own name, so as to obscure in some places the use of the generic name. Thus in the Solomon Islands no native canoe is called vaka or haJca, though that the word is native there is shown by its variation in form. An European vessel is called vala or Jiaka, as in the Malay regions Crawfurd says wang- kang is used for foreign junks. The Melanesian terms in this Vocabulary are the native equivalent to this word, aka, vaka, &c. The words in the Vocabulary of the Malay Archipelago which are forms of this are in number thirteen, and the forms are various ; haka, umga, waha, waa, waim, sakaen, tvug, and bunka, which last is questionable. The terminations im and en are probably suffixed Pronouns. Of the forty Melanesian words thirty-one are corre- 6o Melanesian Languages. spending forms, corresponding in variation as much as in re- f^emblance ; aha^ liaka, vaka, faka, aga, vaga, tvaga, ak, ok, og. Here again this very common word fails in the Loyalty Islands and Southern New Hebrides, and in Santa Cruz. If then we take the very common use of aka in Polynesia, Indonesia, and Melanesia as a proof that canoes were made and called by some such name before the original race was scattered far away, it is still j)os&ible to suppose that some of the family had wandered oif before the word arose, and have reached the extremity of Melanesia without being overtaken by this word. It rather tends to encourage such a supposition that only canoes constructed with planks are pro- perly called aka or vaka, &c. The Santa Cri;z canoes are elaborate sea-going vessels, and they are called loju, which, as j takes the jDlace of t, appears to be the same as lotu of the Sula Islands. It is possible also that the lakatoi of Motu, New Guinea, is the same with Malagasy lakana. 9. Body. — A considerable number of the words in Mr. Wallace's list agree, fourteen out of thirty-three ; but the word badan is not at all represented in the Melanesian list, Malagasy and Polynesian words agree, tena, tinana, Una. There is nothing of interest in the word ; but the Sula Islands koli and Florida hull are probably the same. 10. Bone. — In the Vocabulary of Malay Archipelago words there is nothing at all of general agreement ; the Malay tulang is the Malagasy taola^ia. The Polj-nesian word is iwi. Neither of these words is found in the Melanesian Vocabulary; and there is hardly an agreement of any two words in the two lists. There are two Melanesian words common, hui or sui, huri and suri : the first confined to Fiji and the Northern New Hebrides, the other spread from the same group to Duke of York, and beyond that to Mafoor in the north-west of New Guinea, where bone is kur. The first of these is probably the same with hoi of the Sula Islands, in the language of which remarkable coincidences have been observed in the words for banana, boat, and body. As concerns the word su7-i the interest lies partly in the phonetic changes, and paitly in the distribution of the word and the ex- ceptions to it. We again observe in it a word very common in Me- lanesia generally, which fails to ai)pear in the Loyalty Islands, in the Southern New Hebrides, in Santa Cruz, and in Savo. The Fate word vatu is stone. Notes oti the Vocabularies. 6i 11. Bow. — This is a word of extraordinary interest. The !N[ah\y fanah is said to be the Sanskrit vdna, the Malay meaning bow, the Sanskrit arrow. That very many Sanskrit words are in use in Malay is certain ; most of them words belonging to tlie higher state of civilisation which intercourse witli India has assisted the Malays to attain. Supposing /;««« in all its forms to be indeed Sanskrit, there is no reason for supposing also that the Malays learnt the use of the bow from India ; as with many other words, the native term may have been superseded by the foreign one. Nor would there be any great difficulty in supposing that the Sanskrit woi'd has peneti-ated to the Pliilippine Islands, or even into Polynesia. But the presence of the word in Melanesia cannot be thought devoid of difficulty if only the distance between India and the Loyalty Islands is considered. Yet, as immigration from Polynesia has certainly within a few generations reached the Loyalty Islands ^, it is not at all impossible that the Nengone 2)ehna may have been a late importation there. In the same way the Eotuma fan may be thought to have come from Tonga ^. But there is another thing to be considered which can hardly be thought easy to reconcile with the belief that the Sanskrit word has reached Melanesia. In the Vocabulary it may be seen that pehna in t^engone, /ana in Ana'iteum, fan in Rotuma, are the only forms in which pana appears in Melanesia, and they may be easily understood to be recent ; as may the Santa Cruz nejma an arrow. These words signify a bow ; but in Fiji and in Florida the word does not appear as a bow or arrow, but as a Verb, to shoot ; vana in Fiji to shoot with a bow or gun, and also to bore a hole or /«erce through ; vanahi in Florida to shoot with bow or gun. In either language there is a native word for bow, dakai and hage. The fact that in Fiji vana means to pierce through as well as to shoot is well woi-thy of consideration; it seems that the sense of piercing is primary and that of shooting secondary, rather than the reverse. But, apart from that, it seems strange that in islands whei-e the word does not signify a bow or arrow, it should mean to shoot ; on the sujiposition that ' The Polynesian immigrants in Uea in i860 counted four or five genera- tions of chiefs. Notes grammaticales sur la Langue de Lifu. ^ ' The Tongans got the bow and arrow from Fiji, as they got pottery, salt, and their improved canoes. They say they had not the wordyVtwa until they got the bow, and they imported the word (Fijian vana) with the thing. They call the bow kaufana shooting stick.' Rev. L. Fison. 62 Melanesian Languages. it has been imported from the Sanskrit. We have to make the supposition that a Sanskrit word meaning arrow has been carried to ishmds at a vast distance, and certainly never reached by direct commerce of the Malays, and has there not supplanted the native words for arrow or bow, but been taken up as a Verb, to shoot. It is of course possible : but the date of Indian intercourse with the Malayan peoples being generally put within the historical period, it gives little time for so great a change and journey. It is much more difficult to account for the Verb in Florida than for the Noun in Nengone and Anaiteum. Still it is rash to put forth a counter theory, and presumptuous to disbelieve the identity of the words ; and perhaps only pardon for some hesita- tion can be expected. Yet it may be that the resemblance is accidental ; it may even be that the word has been borrowed by the Sanskrit. For myself I cannot easily believe that Florida has got the word from Sanskrit \ Among the forty Melanesian words for a bow, eighteen are forms of vus^ varying very widely indeed to take in yu and /7t, but still easily recognisable. In the Malay Archipelago there are many words which are evidently the same, husu, pusu, husu, osio and others. Whatever, then, may be the history of pana, here is a word common to the Indian and Melanesian Islands. The word bage used at Florida seems confined to the Solomon Islands, and there are other local words. The Malagasy antsahy and the Maori kopere also stand quite apart. 12. Butterfly. — This is one of those words in which there is the danger of getting the species instead of the genus, and much agreement in the names cannot be expected. There is, however, a certain amount of interest about the word ^^e^^e, and its compounds. This is the common Polynesian word, and it appears twenty-five times in the Melanesian Vocabulary 2. More than this, there may * In Mr. Van der Tuuk's very instructive Outlines of a Grammar of the Malagasy Language he gives voatavo pumpkin, as ' an interesting proof that the Sanskrit words came into Malagasy from the Indian Archipelago,' making tavo to be the Malay labu, the Sanskrit alahu. But in Mota a native pumpkin is too tavai, and wo is tvoa the same with Malagasy voa fruit. The likeness of favai and tavo is not of the same value as the identity of voa and ■wo ; but is the likeness of Malagasy tavo to the Sanskrit alabii sufficiently great to make us extend the Sanskrit word into the Banks' Islands ? If not, as one may venture to think, the Malagasy tavo and Mota tavai are the same, and quite distinct from Sanskrit alalti. ' Mafoor, in New Guinea, apo;p. Notes on the Vocabtdaries. 63 be a question wliether the latter syllable of the Banks' Islands word rupe (also rup, rom) is not the pe which is reduplicated in the common word. The word pe;;e itself is used in Mota, though not as the name of a butterfly ; they call flying sparks ^;e/)e- roworowo ' flying butterflies,' and two canoes sailing side by side are counted as 2>^P^ i"^^ 'butterfly-two.' In j\Ir. Wallace's list the word occurs only once, in Morella of Amboyna, as pepcid. This is identical with the Savo woi'd beheuJa ; and it is further evident that the same elements reversed make up the Florida uleulebe, uJe and he, and the New Hebrides lepepe. The Malagasy lolo, being u, may be the same word ul. We have therefore not only a very wide distribution of pepe in Polynesia and Melanesia, but an evidence that the Melanesiaus have not borrowed the word from Poljoiesia, in the presence of the word in Amboyna in precisely the combination in which it occurs in Savo. Observe also the variation of a single root in G^ao kohou and Duke of York toto. 13. Child. — This word is subject to confusion between the meanings of cliild as a j'oung person of either sex, and as in relation to the pai-ents. A native is Likely, on the one hand, to speak of his cliild as his boy, and, on the other, to speak of a grown-up son as his child. Nearly half the words in Mr. Wallace's Vocabulary are forms of what in Malay is anak ; and this is also the Malagasy zatuika. The Polynesian word is tama ; tamaiti, tamariki, tamachi, with an adjective meaning small. Neither of these words appear in the Melanesian list. The word ana is, however, in use in the Banks' Islands, and in such a way as to suggest that the word is the same as the Malay, and also that the root meaning of the word is to be found in the Melanesian use. It is assumed of course that the primary sense of the ISIalay and Malagasy word is ' child,' and the expressions are interpreted as metaphorical by which in both languages an arrow is called anak 2>(inaJi and zanali antsaky ' child of the bow.' The word, however, as used in the Banks' Islands, if it be the same, means primarily an append- age or belonging ; my boy is rereynera anak, the boy belonging to me, tanun anai is a man of the place, not a visitor, tanun anak a man of mine, a dependent. The last thing of a series is the paspasoanai, the hundredth mel nol anai. It is therefore an interesting suj^position, in view of the unity at bottom of all these languages, that in the isolated Banks' Islands anai we may have the primary signification of the common word ; and 64 Melanesian Languages. that anah conies in a secondary sense to mean a man's child, and an arrow to be called anah i)anah, as belonging in the way of an ajipendage to the father and to the bow. The most common Melanesian word, fifteen out of forty, is natu ; a word the primary meaning of Avhich is no doubt ' little.' This word clusters about the North New Hebrides and Banks* Islands, but there are outliers in Eromanga and Duke of York. It may be doubted whether tu, without na, which ajDjoears in places so far apart as G^ao, Ureparapara, and perhaps Nengone, is not a separate word. In the Malay Archij)elago in nanat and naanati of Bouru, and iintuna of Gilolo, we have no doubt the Melanesian word ; as so often haj)pens, the word which is common in one Vocabulary appears as exceptional in another. There is another Melanesian word, which, though common only in the Solomon Islands, appears also in the New Hebrides, Tcari, gari, gale. The word given as ' girl ' in New Britain is probably the same, yarra vafini', since gari mane is a boy in Florida. Many examples occur of the interchange of h and iv ; and it may therefore be assumed that this word is represented in the JNIalay Archipelago by wari in Amboyna. A very interesting word for child is vnera, used in Mota with Tiat, natmera, a small child. Ordinarily, however, the word is reremera, with the reduplicated plural sign re. It is to be observed that this plural word is used to designate a single boy; one boy is called ' children.' This is parallel to the use of raveve for mother (see No. 43), and rasoai for husband or wife ; it points back to the time when the children were the children generally of the community, and not individualised. In Teste Island, New Guinea, a boy is meramera, as sometimes in Mota ; mela in Malanta is the same. 14. Cocoanut. — In the various stages of growth and ripeness very different names are given to the nut ; it is possible, therefore, that Mr. Wallace's list contains words which describe the pai"- ticular cocoanut in view, and not the nut generally. However, what is certainly the common word for cocoanut in Polynesia and Melanesia occurs in his Vocabulary very often — the woi'd niu ; which is also the Malagasy nihu. There are two words which nearly divide the Melanesian list between them, m« and matiy. The second is quite local : the first extends from New Guinea to the Loyalty Islands. In Micronesia the Marshall Island word is the same, ni. Notes 071 the Vocadittaries. 65 15. Cold. — In this word, as is tlie case witli other Adjectives, there is no general agreement in the Vocabularies, nor between them. There are some words the same ; the Matabello mariri is the Mota mamarir ; the Ceram makariki is perhaps the Maori makariri. The two Vocabularies, however, agree in exliil)it- ing the characteristic prefix ma of Adjectives, and the similar prefix da. 16. Door. — There is in this word piobably some confusion between door and doorway. The door and the doorway, the opening and that which closes it, are more distinct in the native mind than they are in our common speech. Neither in the Malay Archipelago vocabulary, nor in the Melanesian or the Polynesian, is there any general agreement. In the Melanesian list a great number of the words are compounded with 7nata, or ma, the common word for eye or opening : and this has followed perhaps from the use of the Mota mateima, as the word to which equivalents have in most instances been got. The same compound appears in several words from Amboyna and its neighbouring islands, so far at least as regards mata. The eye of the house is the common expression in the Banks' Islands and Northern New Hebrides, and this describes the opening and not the shutter. In the Whitsuntide mat gatava the meaning is different, gatava is not the house but the door ; as in Mota, palegetava is the shutter that closes the doorway. 17. Ear. — The word which in Malay is telinga is the most common in Mr. Wallace's Vocabulary, and this is the taringa of Maori, and, with variation, of Polynesia generally. The ordinary Malagasy word is different, but the Sakalava tadiny is a change from talina. This word also is very common in Melanesia, occur- ring twenty-four times at least in the list. It is also in Micronesia, in Marshall Islands, lo-jeling. The great variation in the Melane- sian forms goes far to show that the word is not an importation from Polynesia or elsewhere. There is the very common change between r and I, and of d and t, and from t to ts and s. There is also the dropping of t, which is characteristic of some dialects. But the presence of forms in which k stands in the place of t points to a much more ancient common origin ; karina being no doubt the same as talina. The change between I and k certainly sometimes occurs, and it may be that tikga of ^Anaiteum {g being ng, in the other words written 71) is iikin(ja = telinga. The most curious variation, however, is that of the Rotuma /o?mn. The two changes in this are in accordance with the practice of the language ; F 66 Melanesian Languages. t has turned into /, and the last syllable is reversed, an for na. Observe the Motu taH'a for taliga. There is another word which occurs locally in the Northern New Hebrides and Banks' Islands, qoroi in Mota. The word talina signifies the orifice of the ear : this signifies its joointedness and projection ^ This word, as usual, is not without its representa- tive in the Indian Archipelago ; it is no doubt the same with horoncja of North Celebes. Another word again, kuli, has a narrow range in the Solomon Islands. 18. Egg. — The two Vocabularies of the Malay Archipelago and of Melanesia agree in the most common word, the Malay tulor ; which, with variations, occurs in nineteen places of the former and eighteen of the latter. It is the same word with the Malagasy tody, the change being regular from I to d, and proved in this particular case by the Betsileo pronunciation toly. The Poly- nesian words are quite different, hua, fua ; the same word which is used for fruit, having the root meaning of something round ; and the same which appears in the Nengone wa tei and the Vureas wowese. The exceptional words in neither Vocabulary seem to agree. 19. Face. — In Mr. Wallace's Vocabulary there is no word which occurs in more than three places. The Malay muka stands alone. This, however, appears in Melanesia as an Adverb, muka being ' first ' in Duke of York. The Javanese word is the same as the Fate rai. In the Melanesian list the words mata and nago appear respectively sixteen and seventeen times. The former of these is the word so commonly, almost universally, used for the eye, and used also for the face in the Polynesian languages. The Eotuma again shows t as /, maf for 7nat. The second word, nago, has no repre- sentative in the Malay Archipelago. In Melanesia it has a wide range, from Sesake in the New Hebrides to Alite in the Solomon Islands, for lao is a form of 7iago. In Santa Cruz the same word is in use for * mouth,' nao. In fact, in Melanesia it seems common to use the word which properly designates some feature for the face generally. The word used in Ureparapara for the face, naregi, is mouth in Volow, and lip, snout, beak, in Vanua Lava, and in Mota is used for a point of land. In languages where nago is not face, as in e. g. Florida, it is used as an Adverb, ' before.' 20. Father. — The very simplest form of word naturally serves as ' Qorqorosa is said of a plant when the buds begin to show ; a tendril is the qoroi of a climbing plant. i Notes 071 the Vocabularies. 67 a vocative for father, and this may easily become a common noun. Thus in the Mahiy region hajya^ haha, are no doubt of the same ch\ss of words with the English papa, and the very common Melanesian mama. Another prevalent word in the Malay Archi- l^elago is amxi, which may probably be the same with the common !Melanesian tama, which is common also in the Gulf of Papua, New Guinea. The example of Mota may probably stand for other languages; mama is the vocative, to?«a the common noun: yet mama is used also as a common noun, though never quite as tamu is. They will say mama inau for ' my father,' but never suffix a Pronoun, as in tamak. Tlie Malagasy ray is quite distinct. The Maori matua tane is a periphrasis; ^ja and ^;«pa are voca- tives. In Samoa, however, tama is used, as in Tonga. The Nengone caca {c = ch) is the same as Aurora tata, a word which is a voca- tive for an uncle in Mota. In Eotuma again ta appears as fa. 21. Finger. — There is no word at all common in the Malay Archipelago : some few are compound with lima a hand, but there is nothing to remark. In Melanesia, however, there is one word common to the Banks' Islands and Northern New Hebrides. That the word has a wider range now, (a witness perhaps to a much wider ancient use,) is shown by its presence in Nifilole near Santa Cruz, where the finger nails are bisi nime, and in Savo where fingers are karakara bizi. Both these latter differ generally from the more common type of ISIelanesian languages, and the presence of a word in them may be thought to argue an older connection. There is another Melanesian woi'd which will deserve attention under the head of ' hand,' kakau. Another is the Florida gigiri, which is remarkable as being the Vatvirajia ririki in another shape by metathesis. It is not at all unusual for the syllables of words to be reversed, and in this instance k has also been nasalised ; gigiri is kikiri, reduplicated kiri, as ririki is reduplicated riki ; the Saa riini is the same. The Nengone tubenine means the 'I'ow* of the hand, as tube nengoce means the ' row ' of the lips. 22. Fire. — There is for this an important word, because it is one of those which are so very widely spread in the Malayan, Polynesian, and Melanesian languages. The forms indeed are very different, which argues an ancient distribution. There are closely resembling one another the Malay api, the Polynesian aid and ajl, and also the forms eji, aif, yaf, yap, given by Mr. Wallace. The forms in Bouru, Amblaw, and Ceram, afu, aim, yafo, unite with the Malagasy afo, and lead on to aoiv and hao. In Melanesia the F 2 68 Melanesimi La7igtiagcs. variation is not so great, avi and ev differ little ; but if, as is probable, ha'pi and gapi, kaim, gcqyu, caj), are the same word, there is enough. In both the Malay Archipelago and in Melanesia there are many- exceptions, few of which agree together. Of the exceptions one of the most remarkable is hana in Bouru, taking it to be the same with the Malay panas and Malagasy fmia hot ; because it may be connected with the Fiji word it/aga (^q^ngg)ai\A the Y&iefaga{g=^ng), which are both used to express fire. It is an exti'aordinary thing that no word equivalent to fire should be in use in Fiji in any dialect \ To express fire words for firewood and burning brands are used. There must be a reason for this; which may be that the word for fire became tahu, or that, fire being always ready for use in fire-sticks, the word ' brand ' came to be used for ' fire.' In the latter case there is a certain parallel to the use of /et4 and fuoco for fire in the Romance languages. There remain words quite distinct from one another, used in one, or two or three languages, such as tuwa, laJce, joto, &c. To account for the use of peculiar words is impossible : but the same causes which caused the word for fire to disappear in Fiji, may have operated to bring in some new word in other places. Generally, however, it cannot be argued that kindred languages must needs have a common name for so necessary and primitive a common possession of their race as fire. Greek and Latin languages are closely allied, yet one lias nip, the other ignis. In the Romance languages the Latin ig7iis has disappeared. From what- ever source the Greek nvp came, we have its English coguate in 'fire.' In the same way the Ocean languages may have more than two or three distinct words for fire among them without being on that account forbidden to claim common descent. 23. Fish. — This is again a word of great interest. There is one very common and far-spread word for fish, the Malay ikun, Maori ika, Marshall Island iek, IMota iga, in the Malay A^rchipelago, Polynesia, Micronesia, and Melanesia. The changes in form are great enough to show an ancient disjjersion ; even between j\Ialay and Javanese there is the difference between ika and iwa. In Melanesia the consonant is often omitted, and we have ia, ie ; but in one form or another the word extends from one end to the other ; from ie in Nengone to ian in Duke of York. The con- nection is kept up in New Guinea in the Mafoor iji>n. The ' Kev. Loi'iiner I'isoii. Notes oil tJie Vocabidaries. 69 exceptions are comparatively fewer in the Malayan Vocabulary than in the !Melanesian, iha appealing in twenty-seven places out of thirtj'-three in the former, and in twenty-seven out of forty in the latter : but the very general consent is remarkable. The exceptions in these two Vocabulaiies in no case correspond, unless nau of Galela be no of Santa Cruz, which is also in New Caledonia ; but there is still much of interest to repay investigation. In the first place, the Hova dialect of ^Malagasy does not appear to have a word for fish at all, hazandrano, the expression in use, meaning water-game ; and this is sufficiently remarkable in itself, corresponding as it does to the deficiency in Fiji of a word for fire. In the Sakalava language, however, of Madagascar the word for fish is fiana. The root is Jia, and this comes near to one of the Melanesiau exceptions, fei of Bugotu. The unaccountable isolated apj)earance of exceptions in this and other words is remarkable. The general character of the Bugotu language is that of its neighbour at Florida, but it has many words in the vocabulary" of common things extreme]}^ unlike ; asjoto for fire, and fei for fish, although the common words aid and iga are also in use. The Savo language, on the other hand, is very unlike its neighbours ; and here an exception is not unexpected. The extraordinary use of one word, nmh, in Lakona, for bird and fish, has been noted. The same word is used for fish, but not for bird, in Aurora also and in Vanua Lava, in the forms masi and mes. In central Borneo masik means a fish, and in Koiara of New Guinea mesia, and these probably are the same word with the Melanesian. 24. Flesh. — In Mr. Wallace's Vocabulary isi and words I'e- sembling it are very numerous ; in the Melanesian Vocabulary visogoi and words like it are seventeen out of forty. In Santa Maria, on the one side, in Gaua the word is moswivin, on the other, in Lakona, ^;iAw ; in which two words it appears as if wivin were another form of filii, and compounded in each form with some other word. From this the conclusion may possibly be that vis, so common at the beginning of visogoi, vis-igoi, vihigogi, is the same as 2>ihi of Lakona, ivivi of Gaua, though in another combina- tion. Between visi and isi the difference is slight. 25. Fly. — The bluebottle fly is so much more likely to present itself than any other, that, though we may have in some cases the name of a species and not of the genus, the words in the Vocabularies are still likely to be names for the same thing. The name of that kind of fly is very widely the same in Polynesia and Melanesia, as 70 Melanesiaii Languages. lawo in varying forms. The same word occurs, but not often, in Mr. Wallace's Vocabulary ; in Celebes, Sanguir, Ceram, and Baju ; and though the common ]\Ialay word is lala, yet langau is also there. In the Maori of New Zealand the common form is ngaro, but ranyo is also used ; an example of the transposition of syllables not uncommon also in Melanesia. In the Melanesian languages the range of the word is from lag (lang) of Anaiteum, through lauo, rauo, len, thano, glccno, to lao of Motu and ran of Mafoor in New Guinea. 26. Fowl. — The domestic fowl has been no doubt introduced into the Melanesian islands : one might expect therefore to be able to trace the source from which it was introduced by the name by which it is called. It is disappointing, however, not to find anything in Mr. Wallace's Vocabulary which corresponds to the Melanesian words, unless toivim of Ceram may have to do with toa, which is not likely. In the Malay Archipelago languages, in twenty places out of thirty-three, the common name for a bird, a?^ come naturally together with tlie Banks' Islands qat and its changes into qotu, qidugi and qiliji. In Formosa vau is the word, which may very well be the same, and in Kings- mill it is atu. The root meaning of this widely spreading word can be ascertained in the Banks' Island languages : it means a knob, as in Mota a qat here is a knob stick. In better known languages the same metaphor occurs. The Malagasy is loha, and it is very remarkable that the same word is found in the Solomon Islands, lova head in Vatura?ia, and forehead in Savo. Here again a word isolated in Melanesia finds its kindred in some very remote language ; and it is plain that loha and lova have reached Madagascar and the Solomon Islands from a common source, neither Malay nor Polynesian. 33. Hot. — The Malay panas is a well-known word, of which mention has been made under ' Fire.' With the prefix of quality ma or ba it makes mofanas and hafanat in Ceram, mafana in Malagasy, and mahana in Maori. In Melanesia it is only found in words used for fire. There is another Maori word, wera, in Samoan vevela, which in ^lota as vevera is used of red-hot stones. In Fileni, one of the Polynesian outliers in the Santa Cruz group, vela is the sun. An exceptional word in I\Ir. Wallace's list is sasahu reduplicated, dasaJio with prefix of quality, in Tidore and Gilolo. This is no doubt the siahu of ]\Iotu, New Guinea, the sawsaw, sousou, seuseu, seseu, of the Banks' Islands, and is again a good example of the extensive occurrence of words which in their own regions are exceptions. The prevailing word about the Banks' Islands is tutun, titin, which in Vanua Lava, where t is left out, becomes i'in. The Eotuma sunu, Api pisusunu, is probably the same. Like so many Adjectives in this and other lists, tutun is reduplicated, tiie root is tun, a word which in Mota and also in Duke of York means to roast. 34. House. — This is an interesting and important word. The very wide range of the word, which in INIalay is ruma, and the great variety of its forms point to the great antiquity of this as a common possession of these languages. As is the case with the very widely prevailing name for a canoe, we may argue that a word which has spread so far and changed so much goes to show that the thing which it names was known to the undivided people whose dispersion spread the word so widely abroad. If the presence of certain common words in Aryan languages shows that the Aryans 78 Melanesian Languages. did not separate till certain arts were known and practised by the common ancestors, so we may argue that the Ocean languages testify that the ancient speakers made canoes, built houses, culti- vated gardens, before the time came when their posterity branched off on their way to Madagascar and Fiji, The word now immediately in view as the name of a house ranges from the Malay Peninsula, through the islands of the Indian Archi- pelago, to the very extremity of Melanesia in the Loyalty Islands. It has not a continuous range, it appears and disappears at intei'vals, but in that line and chain of islands it is never absent long. It appears in Mafoor at the north-west of New Guinea, and in Motu at the south-east, and in the Marshall Islands of Micronesia. In Polynesian languages it does not appear ; in the Kingsmill it is im. The fact that the word in this way has established itself generally, but not universally, at intervals and not in a continuous line, shows that it is not one which can be traced to one centime, from whence it may be thought to have been introduced by commerce or modern intercoui'se. The same conclusion is enforced by the consideration of the great variety of the form of the word, which ranges from ruma to en (eng). If a word appearing in its full form in Malay were to appear corrupted and changed as it receded in distance from the region in which Malay is spoken, we might well suppose the ilalay the original. But when the changes in form bear no certain relation to the distance from Malayan regions, and the variations are local and disconnected, it is not so ; some centre there must have been, but it cannot now be pointed out. The geographical range of the word must be obsei-ved bj^ com- paring the Vocabularies with the map. The variation of the form can be seen in the Vocabularies. In Mr. Wallace's list the Malay rumali and the Javanese umah give at once typical forms, one with and the other without an initial consonant. Of the first type there are also luma and huma, of the second um and probably om'^. Out of thirty-three words twenty-two are forms of these types. The variety of forms in Melanesia is greater, but the tyjies are the same ; rwma is in Duke of York and San Cristoval, uma in Api and Lakona. The vowel also changes, and ruma, with changes of initial consonant and vowel, becomes luma, nume, huina, rima, nima. By similar change uma becomes ima, ema, and dropjnng the vowels at the beginning or end, 'ma, im, eom, em, en. To ' The Ceram word y», used by Alfuros, is probably om, a form of inna, with the collective prefixybi ^Fiji vei. Notes on tJic Vocabularies. 79 account lor this last cliaiige it is enough to say that, in the neigh- bourhood where it is made at any rate, tlie m is tlie nasal one which, as mentioned above, regulaily changes into ncj : mm, im, makes en, as lima ' a hand ' makes Fiji liga, Maori ringa. This m in Nengone is written 'm, and the Nengoue 'ma is identical with the Santa Cruz ma. A tabulated view of the forms of this word and its distribution may be useful : — Malay Arch'rpelayu. Melanesia, riima . . . Malay, Aniboyna Motu, New Guinea Duke of York, Malanta, San Cristoval. rum Mafoor, New Guinea Ittma . . . Bouru, Amblaw, Amboyna, Ceram Malanta. ritna San Cristoval. numa, nima Malanta, Ulawa. hvma . . . Bouru. suma Fate. ttma , , . Java, Sula, Ceram Lakona, Api. wia Mota, Araga, Espiritu Santo, Ainbryiu. vm, om . . Gilolo, Ceram Anaiteum. -im, em, en Kingsmill Islands . Banks' Islands, Marsliall Islands, ma Santa Cruz, Nengone. The common word for a house in Polynesia is the Maori whare, Samoan and Tongan fale. This appears also, but rarely, in the Malay Archipelago, hali and bai'eh in Sanguir and Salibabo. In Melanesia it is not common, but it is the prevailing word in some parts. In New Britain and Duke of Yoi-k it is^^a^, in the Solomon Islands vale and vadhe, in Fiji and the Northern New Hebrides vale, in AmbrjTn Jiale. In Duke of York 2^1 is an outhouse, while im is a house. In Mafoor of New Guinea siim is a chamber, while rum is a house. In the New Hebrides ima is knoAvn and used for some particular buildings where rale is a house. Thus these two words to some extent overlap ; yet it may be said that the one belongs to the Eastern and the other to the Western Pacific. There are other words which are exceptions in all parts of this area of language — the Malagasy trano, for example. It is singular that in Vanua Lava in the Banks' Islands, an island twelve miles long, there are three words used for ' house ' so perfectly distinct as im or en, qeqek, and ijovur. 35. Large. — There is not any common word. The Ceram ilahe 8o Melanesian Lancriiazes. Vi is no doubt the Maori rali% and another Ceram word, mama, may be the Solomon Islands paina. The Malagasy lava is ' long,' the Maori 7'aha ' open, extended,' yet these are no doubt identical, and the difference in particular signification encourages the belief that these are the Mota lava large, of which the San Cristoval raha and rafa are forms, as well as Marshall Island lajy, and probably Duke of York galajn. There is very little ground for comparison between the words of one region and another. Within Melanesian limits the Fiji and Santa Cruz levu is probably the liwoa, luwo, of the Banks' Islands. In Malanta the haila of Alite, which con- stantly changes n to Z, is paina of Bululaha, of which latter name the last part is itself another form oiraha, lava. In fact languages have more than one word in common use, as Mota poa, liwoa, lava. 36. Leaf. — The word most commonly used for a leaf is very widely spread, and has a number of forms ; the root of it we may take to be rati. Between this and lau thei'e is no difference. By taking on d, as is often done before r, we have drau, and by another process d takes the place of r. Beyond this d, as is also common, becomes n ; and the forms rau, lau, drau, dau, nau are made. These are shortened into ro, ru, ri, &c. To these stems then have to be added the terminations na, n, gi, i, belonging to substantives, and the great number of forms in the Malay Archipelago, Me- lanesia, and Polynesia are accounted for ; the Malay daun, Malagasy ravina, Maori rau, Samoa lau, the Duke of York dono, the Lepers' Islands raugi, Lakona draivi, Mota naui, Ureparapara dugi, Motlav ron, Volow raven, Santa Cruz leu, Mosina no, Gaua do, even the Anaiteum ri and Nengone ru. The Vanua Lava togi is in accord- ance with a change commonly made there of n to t. In this case again the wide distribution and great variation of the word point to its antiquity in the languages, and agree very little with the theory of imported or bori'owed words. It has been already mentioned that this word is used in the sense of hair. In Mota little lapping waves are called nono naivo saltwater leaves, a phrase which shows the root notion to be that of flakes^. The same is shown by the Ulawa a^xiapa ni ai, Bulu- laha apapa ie, apaapa being used for the wing of a bird. The Ysabel eloelo i gai is perhaps akin to their word alo a wing. The G&o klakla it has been said is hair or leaf, as ulu in Fate. ' 'The complete form in Fiji is drmi ni fcau, in another dialect, vo )ii kai, leaf of tree, as drau ni ulu, ro ni vulu, is the hair of the head.' liev. L. Fison. The word druu is tlius shown to be hardly yet enough specialized in meaning not to recpiire some further indication whetlier hair or leaf is meant. Notes on the Vocabularies. 8i In many languages the word ran is used for a hundred, from the habit of using a leaf as a tally. 37. Little. — As is the case with most adjectives, there is but scanty agreement in the words meaning ' little.' Taking Polynesia, Micronesia, Melanesia, there is a word riki widely distributed, and, what is of more importance, not lying on the suiface. In Maori riki is present, but not commonly used ; in Mota it is obso- lete, but remains as rig in names and phi'ases. It is in the Mar- shall Islands as lik, and in eight places of the Melanesian Vocabulary. Agreement between widely separated places is seen in laki of Nifilole, near Santa Cruz, and malaki of Motu, New Guinea, the latter with the prefix ma of quality. 38. Louse. — In the name as in the thing there is a very general agreement. Out of ]\Ir. Wallace's thirty-three words twenty-seven are forms of the Malay kutu, and twenty-eight Melanesian places have the same. The foi'm differs very considerably. The Malay kutu is no doubt, by the common change of k to h, the Teor hut, and, by dropping the initial consonant, is utu and ut. The Maori kutu is Samoan 'utu. The Micronesian kid of Marshall Islands is no doubt the same. In Melanesia the changes are more consider- able, and certainly do not favour the theory of a recent importa- tion. The change from k to g, and to w is regular, and gives the Fiji kutu, Whitsuntide gutu, Mota wutu, and the shorter forms git^ wu. The Xengone ote is perhaps, and Anaiteum cet (c = hard g) certainly, the tame ; and Rotuma, with its usual change of t to /, makes uta into ufa. In the Solomon Islands the simplest form is reached in u, u'u is utu, fu of Fagani is hu. The Ulawa, Wawo, and Saa pote, bote, are not likely to be the same, but they may go with the Nengone ote. 39. Man. — Thei'e is not in this word so great an agreement as might perhaps be expected : and there is a certain confusion Ukely between man and male. In Mr. Wallace's Vocabulary mon, omani, manesh, are very likely the Solomon Islands mane male. The Malay orang ^ is the Malagasy olona ; elsewhere there is nothing answering to it. There is one root, however, that by itself or in composition makes often the name for man, ta, tau ; in Celebes tau and taumata are no doubt the same with the Motu tau and Fiji tamata ; and ta is the root of the words tamoli, tanaloe, tati(a, tanun, tamsar of the Banks' Islands and New Hebrides, and the ' Though orang is unknown in Melanesia, the second word in the well-known orang ufaii is common there for the ' forest.' 82 Melanesian Languages. Polynesian tamjata. The root ta does not often occur, but it is in Volow, and reversed in Motlav and Ureparapara. The Eotuma becomes, by the change common there, fa instead of ta. In the neighbourhood of the Gulf of Papua, New Guinea, man is in Port Moresby ta%na, Kei'epunu auna^ Teste Island, E. Cajje, and Heath Island tau. It is not easy or possible in most cases to explain the words compounded with ta. In Mota it is not unreasonable to suppose the word tanun to be the real man ; nun true, not the tamate the dead man. There is in the language the word tamaur live-man, opposed to tainate dead-man, a ghost ; tavine is a female, ta a woman ; tamatua is a full-grown man, the mature ta ; tamaragai an aged man, a trembling ta. No doubt the Fate and Sesake word tamoli is identical with the Mota tamaur. When a native says that he is a man, he means that he is a man and not a ghost, not that he is a man and not a beast. The intelligent agents in the world ai'e to his mind the men who are alive, and the ghosts, the men who are dead, the ta-fnaur and ta-mate of Mota, na ta-moli and nat-mas of the New Hebrides. When white men first appear to Melanesians they are taken for ghosts, dead men come back ; when white men ask the natives what they are, they proclaim themselves to be men not ghosts. 40. 3fat. — This is a bad word for the purpose and ought to have been left out. Mats are of very various kinds, and e^ch has its own name. No collection of words, however, can be entirely void of instruction. The Alite vau gives a word as a noun which alike in Florida, Mota, and Fiji means to weave. The Eotuma gives eap, which is the same as the Mota e;;a, but with the last syllable reversed, as we have seen in falian for talina, an ear. 41. Moon. — In Mr. Wallace's Vocabulary twenty- five out of thirty-three are forms of the Malay buJan, in the Melanesian list nineteen out of forty are forms of the same. The Malagasy volana is the same. The forms vary, as in other words, by the change of h, V, f, w, h, in the initial, and of I and r. There is a question whether we can ascertain the meaning of the word, which, from the use of vula as white in Fiji and ]\lota, and 2>ura (probably the same word) in Florida, may be white. Or the word meaning moon may be used in a secondary sense for white. Of the exceptions the Sula Island fasina is no doubt the Sesake masina. The distance geographically is immense, but both words are connected by the Polynesian masina of Samoa, mahina of Tonga, and the formation from the verb meaning to shine. i Notes oil the Vocabularies. 83 42. Mosquito. — In regard to tliis word there is a great difference between the Malay Archipelago and Melanesia, the name extremely common in the latter for the mosquito, namu, appearing only in Malay and Javanese nyamok. In the iSIelanesian list out of thirty-four words given twenty-seven are the same, and with no very great diversity of form. Samoan and Tongan also have namu. 43. Mother. — There is likely to be the same confusion here as in the case of ' Father,' between the common noun and the vocative. However, in the Malay Archipelago and in Melanesia alike, there is a good deal of agreement in the word tina or ina, which also is in Samoa, and in the Gulf of Papua, and, as jine, in the Marshall Islands. The chief interest of the word is not philological. It will be seen that in the Banks' Islands the word ve, veve, is most common ; and that in Gaua with that word in rave, in Whitsuntide in ratahi, in Vureas in retTie, in Torres Islands in reme, there is a prefix ra and re^. In the word and in the prefix there are the marks of the native customs in regard to marriage and of their history. In that j^art of Melanesia all the poj^ulation, without distinction of island or language, is divided into two sets for purposes of marriage. Each of these sets is called in Mota veve, a word that means division. If, then, this word veve is used for mother it is because the veve is looked upon as the parent, the division is not called veve because it may be figuratively called a mother. But if the set, the division, is properly the mother, and the word used for mother is properly the name of the set, it is evident that the individual woman who is the parent is in the second place ; the child is the child of the set, not hers, the women of the set are the mother, not she. Hence has come the use of the plural in speaking of a single mother, raveve, rave, retne (i.e. re tine), reme, ratahi. There was, in fact, a communal marriage, every woman on the one side was wife to every man on the other, and consequently every child had the women of the set into which it was born for its mother. The plural form of the word for mother, where it exists, is a surviving witness to this. In exact agreement with it a word in plural form, rasoai, in Mota describes a husband or wile. The members of one set were ra soai"^ to the others, males of one to females of the other respectively, and the plural form that was appropriate to that state of things, now long passed out of general ^ A common plural particle. ■■' The reduplicated form soasoai, member or part of an organic whole, shows how soai came to mean husband or wife. G 2 84 Melanesian Languages. recollection, remains to show what the state of things was. No such communal marriage exists or is remembered, but it is known among the natives that the words are plural and why they are so. 44. Mouth. — There is no kind of agreement concerning this word in the Malay Archipelago, nor, with one interesting ex- ception, does Mr. Wallace's list contain any words that are common elsewhere. This exception is hawa of South Celebes, which is haha of Batak, and haiua of Nias, Sumatra, and also vava of Malagasy. The same is waha of New Zealand ; and in Melanesia is vava of Espiritu Santo, ivaivagi of Lepers' Island in the New Hebrides, and waiva of Ulawa and Malanta in the Solomon Islands, loa of Duke of York. This word, then, is widely diffused, though not generally. Its presence in Sumatra and Celebes in one quarter, in New Zealand in another, in Madagascar, and in two different regions of Melanesia, makes it quite impos- sible to suppose it an importation anywhere from outside. There is a further interest in this word. In Mota the verb vava is to speak, which is the form the word for mouth has in Espiritu Santo. Words to be mentioned shortly hereafter will show that it is not unreasonable to suppose that this word has got into Mota in two forms, one in which it means to open the mouth, and one to speak. The word which in Aurora and the Banks' Islands is almost universal, vala, appears isolated in the Wano hara as meaning mouth. But wala, in Ulawa, is to speak. In the same way mana, the Maori mangai, a mouth in Florida, is also in that language speech, and the Mota verb manasag to tell out, has no doubt its stem the same. There are, therefore, three words in various parts of Melanesia which mean both mouth and speaking, vala, mana, and vava, and what one might hesitate in accepting as more than a chance resemblance about one woi'd becomes pretty certain when confirmed by analogy of others. In Melanesian languages, as was observed under the word ' Face,' there is a certain indistinctness in the naming of features. Thus the Fiji gusu (g = ng) is the mouth, in Eotuma michu, Motu iitu, and probably the woe of Ambrym, and no of Fagani, and forms part of Nengone tubenengoce the ' row of the mouth ; ' but nusui in Mota is the lip, and the same word in Maori ngutic is lip also. The notion at the root of both uses is no doubt that of a pro- jection, in which sense it is used in Mota of a point of land. In the same way the Bugotu livo is in neighbouring languages, and very generally, a tooth ; the Santa Cruz iiao is no doubt the Nolcs on the Vocabularies. 85 word nmjo, common as ' face ; ' the word also, which four times in the Banks' Islands means mouth, naregi, in another island of the same group means face. It should be observed that these !Melanesian words were not obtained by pointing at the feature, from which mistakes between mouth, teeth, lips, might easily ensue, but by asking natives acquainted with Mota the equiva- lents for the Mota word, and, in case of difficulty, explaining the precise name desired. The Alite voka is no doubt the Florida voJca to separate, to open ; in Mota mana is the fold of the skin between the arm and the breast. 45. Night. — Though it occurs very seldom in the Malay Archi- pelago no doubt the common Melanesian word honi, qon, is the characteristic name for night. It appears in Javanese and Salayer bungi, and probably in composition as po in three other places. In Polynesia it is po, in the Marshall Islands biing. In Melanesia it is in thirty places out of forty. In Nengoue though ridi is night, days, as we now count, but nights as natives always count them, are reckoned so many bune : in Savo ^>o in pogala to-morrow, is no doubt the same word. Among the excei^tional words the Ulawa and !Malanta roto, Wa?iO rodo, is in Florida rorodo blind ; the word is j)robably the Lepers' Island dodo cloud ; matadodo blind. The Mafoor in the North of New Guinea, rob, is no doubt the Florida ropo. 46. Nose. — The Malay idong, and Javanese irong with hiru, iru, iri, Hi, are the only words in Mr. Wallace's Vocabulary that agree. The Melanesian words do not agree with these. The common Melanesian word is the same as the Polynesian, ihu Maori, iso Samoan. The forms are various, but the identity of gusu (g = ng), nisu, ucu (c = dh), usu, isu, and ihu can hardly be doubted. The word means projection, and is applied to poi/its of land. If isu is the same as nisu, and usu - gtisic, uusu, then this word connects itself with the word mentioned under ' Mouth,' above, the Fiji gusu identical with Fate gusu, and in sound with Mota nusu. Nor is this at all improbable, for the same word may have come into use in some languages in different forms at different times and with the signification differently particularised \ If this is so, the Santa Cruz no, and Nifilole nolo belong to this root. ^ In the same way as, to take examples from our own language, we have yard and garden, captive and caitiff, guest and host, inch and ounce. The word ' snout ' might well be particularized to mean nose or mouth. 86 Melatiesian Languages. The word mama in Mota, with all the forms that surround it, means a beak. The Lepers' Island qanogi applies properly to the nostrils, Mota qauai gills. 47. Pig. — The Malay hahi, there can be little doubt, is the same with Sauguir and Salibabo hawi, and this may very well be hoh of Gilolo and Mysol. If so, the verj^ common Melanesian word ho, qo, qoe, connects with the Malay. This word runs through the Melane- sian Islands from AjdI ^;M^ to Vaturawa ho at the North of Gua- dalcanar, with remai'kably little change of form. Does this, then, argue a comparatively recent introduction of the animal from a common source 1 Hardly ; because hahi is not likely to turn into bo, though it may well be a form of the same root ; and the small places where hoh occurs are not such as could well be the origin of the pigs, and their name, which occupy the central islands of Melanesia. All these words more probably belong to one original root, and spread with the animal as men took pos- session of the islands where they now dwell. A word no doubt recently imported, and probably taking the place of the old word, is puaha, iwaka Maori, pucia Samoan, which appears in Nengone, and Rotuma, and as vuaka in Fiji. It is not likely that there were no pigs in Fiji before the w^ord vuaka was used there. It is more probable that the Tongans brought over their pigs, which were valued and called by the Tongan name, and the name of the newer and fashionable kind of pig superseded the old one. The local word holo in the Solomon Islands has probably the same sort of histoiy. This is parallel wath the substitution of kokok for toa in Mota, No. 26. 48. Rain. — The Malay hujan no doubt represents a word common to that Archipelago, Polynesia, and Melanesia ; it is the Maori and Samoan ua, the usa of the New Hebrides and Solomon Islands, uha and uhe of the same groups, uca of Fiji, xia of Api and Santa Cruz, uh of Lakon, even the of Ambrj^m. In Anaiteum the verb to rain is ehe, which may be the same. The Javanese and Batak is udan, which, by change of d to r, becomes the Malagasy orana. In the Banks' Islands there is a local word wena, weta, wen, wet, which may possibly be the same as the Marshall Islands tout. The San Cristoval rani is the same word with the JNIalay langit, Malagasy lanitra, and Maori rangi, which means wind and sky. 49. Rat. — There is no sort of agreement generally between the names for a rat in the ISIalay Archipelago and in Melanesia. The i Notes on the Voeabiilaries. 87 former differ very iiiucli among tliemselves, tlio latter very much agree, twenty-six out of forty being tlie same. The forms of tlie word in which they agree are very various, and witliout intermediate forms it would hardly be thought that cedo, wohow, and kuzi were the same. All may be clearly seen, however, by beginning with the Alota gasutve. The change of to to v, f, h, gives gasuve, gasufp., gasuhe in the Solomon Islands. The change of w to g and g to w, very common in those parts, accounts for the variety of the Banks' Islands words, gosog, gosug, wohow. The change of s to h gives gahuim, gohow, ivohoiv. The dropping of the initial makes asuhe, and there can be no doubt but that kulii, kusi, kuzi are the suwe of gasuwe. To account for the Anaiteum cedo requires the explana- tion that c is hard g and d = dh ; the word written gedho is not far from gosoio. There are two local and exceptional words in the Melanesian Vocabulary which are well worth notice ; garivi of the four Northern Islands of the New Hebrides and kalavo ' of Fiji. The first of these can hai'dly be other than karufei of Ceram. The second, kalavo, is the provincial Malagasy valavo, Dyak hlawow, Mangkasar balaivo, keluf of Mysol. These two words, exceptional in Melanesia, have, as we have seen in other instances, their kindred words in far distant regions, with which it is impossible that they can Imve had any recent or direct communication. 50. Bed. — The only word for red at all common in the Malay Archipelago is merah, which, however, only appears in Malay, and a few other languages. This word is not unknown in Melanesia in San Cristoval meramera and merameraga are red, mela is used also in Bugotu. In Mota, the red dawn of morning, or the red sky of evening, is called mera ; no doubt the same word. In Vaturawa it is yellow. The meaning of some of the Melanesian words is plain. In Anaiteum caj) is hot as well as red, and is the word used for fire. The Mota memea is from mea red earth ; lawlaw is flaming like a fire ; the Florida sisi, Vaturana chichi, Gao jigia are from the red hibiscus flower ; the Duke of York dara is blood. The reduplication and termination ga characteristic of adjec- tives are conspicuous. 51. Road. — One word is common in all the language area before us: the Malay _;«?«, Malagasy lala, Polynesian ara and ala, '^^z Marshall Islands ial, Melanesian hala, sala, tola, tara, hatha, sal, ' The old black Fiji rat is gaco, c = dh, the Mota gasuwe. 88 Melanesian Langtiages. hal, al. In the ISIalay Archipelago this word appears in twenty- cue out of thirty-three places, in Melanesia in twenty-seven out of forty. The forms are various, hut vaiy by plain changes. In Melanesia the word is often compounded with inata the eye or the middle of the road. The way is sala, the matesala is the path along which one goes. The second part of the comjDOund in halautu, sauiu, cannot be exj^lained. 52. Root. — The Malay akar is also in Celebes, Sala Islands, and Ceram, and is probably the same as waan in two other places. This word is different from the Malagasy and Poljniesian vody and take, the first of which is most likely the Mota vuti. In the Banks' Islands a word, the Mota form of which is gariu, is no doubt the Malay ahar, as is more plainly the Duke of York ahari. It is possible that the Araga garo, Alite kalokalo, may be the same as this. It is not an easy thing to get the correct equivalent for the English word; there is a confusion between the part of the stem underground, the root, and the fibres and roots. 53. Salt. — Between this and the next word ' Sea,' there is in these languages a certain confusion, because salt water from the sea is used for salt, and the sea is distinguished from water by being called salt. The word tasi is common to the Malay Archi- pelago and Melanesia in both senses : as in Amboyna, New Hebrides, and Solomon Islands ; using in Celebes, and asi in Solomon Islands. The Maori tai is sea, wai tai salt water. The most interesting consideration, however, concerning this word belongs to its use in local names. In Malay tasek, though neither salt nor sea, is a lake. In the Banks' Islands the lake in the middle of Santa Maria is the tas ; and it is hardly jiossible to separate this name from that of Itasy the great lake in ^ladagascar. In the same Banks' Islands, the use of the word tas is obselete in the sense of sea or salt, though they still tasig their food with salt water. But the side of the island at Mota where the surf breaks is Tasmaur, the lee side is Tasmate, the live and the dead sea. Tlie same expressions are in use for the weatlier and lee sides of islands in the New Hebrides, and in the Solomon Islands. In Madagascar at the S.E. of the island is Taimoro. A very common word in Melanesia meaning, 'salt,' and also 'salt water,' does not appear beyond it ; and within ]\relanesia is con- fined to the northern New Hebrides, Banks' Islands, Santa Cruz, and Solomon Islands — a continuous stretch of islands. This word Notes on ike Vocabularies. 89 is nawo, navo, nao, which is used very commonly indeed where tasi also is used, as, for example, in San Cristoval and Malanta, where navo is surf, while asi is salt. The word, then, is more commonly used than would appear from the Vocabulary. Another word, masima, is in Fiji and Duke of York. In Fiji there is also taci, meaning the sea. In Ceram there is masin, in Mafoor of New Guinea masen. Salt in Malagasy is sira, also fanasina. 54. Sea. — One of the words commonly used for ' sea ' has just been mentioned, which, however, only appears in ^Mr. Wallace's list as 'sea' in Ceram and Matabello, tasi, tahi; the Maori tai; in Melanesia tasi, sasi (as in Celebes and Bouru), tahi, asi, tas, tai. The Malay word laut is more common in the Malay Archipelago ; a word which never stands for sea in Melanesia, except in Nifilole as lo. The word, however, as lau, for t is only a tennination, is very common, almost universal, in Melanesian languages, in the sense of beachwards or seawards, or as in Fiji, the windward region. The Malagasy word for sea is riaka, and also rano masina, which is interpreted as ' holy water.' But the word masina can hardly be other than that mentioned above as masin and masen salt, in Ceram and Mafoor. The lake Alaotra, a being the preposition ' at,' would seem to be named from laut, as Itasy from tasi. Although the words tasi and lau are common alike in the Malayan, Polynesian, and Melanesian regions, yet there are a great number of words besides, quite different generally one from the other. There are, in fact, several distinct things to be named : I. the sea as salt water; 2. the sea within reefs, in lagoons, or shallow near the shore ; 3. the sea outside ; 4, the open sea, the Ocean. Words no doubt are given which apj)ly in one or the other of these significations. Thus in Fiji taci is the sea generally, the open sea is wasawasa ; in Lepers' Island wawa is the open sea, tald also the sea as salt water; in Whitsuntide, tahi is sea, waivana, the open distant sea ; in Ulawa, asi is sea, ahowa, the open sea ; in Fagani, asi and matawa ; in all these words, as in Sesake tasipua, Santa Cruz daopue, Alite matakua, wa forming part and probably the distinguishing part of the word. In Vatura?ia mao is shallow sea, horara the deep sea, and this is the meaning of horara, zorara, orara in that part of the Solomon Islands. This is also the distinction between the Polynesian tai 90 Melanesian Languages. and moana. In Marshall Islands the sea within the lagoon is lama-lo, the outer sea is lame-do, the sea generally is lojet. Here lo probably is lau, in the sense of seawards, and do is landwards, lok and dok being the common particles of direction ; and lama is the Banks' Islands word. 55. Skin. — In Mr. Wallace's Vocabulary eleven places in the Malay Archipelago have the word kulit, kurito, koli, urita, uliti, holit, which is also the Malagasy hoditra, in Betsileo holitra. In the Marshall Islands the word is gil. Tlie -same word continues to ajDpear at intervals in Melanesia as far as Fate in the New Hebrides. In the Solomon Islands, at Florida and neighbouring parts, it is guiguU, guli, in San Cristoval uriuri, in a dialect where g is drojiped. In Fiji it is kuli, in Eotuma uli. The Torres Islands, gilit is no doubt the same, and, by the common change from g to w, it becomes in Fate ivili ^. The word, therefore, has a vast stretch of extent with very little general variation in form. It is also in the Polynesian languages, the Maori kiri, the Samoan iliola. Another word occupies the Northern New Hebrides and Banks' Islands, the Mota viniu — a word not altogether local, for it appears in the Duke of York pin. Of exceptional words the Espiritu Santo tinina corresponds to tinyan of Amblaw, hlutini of Ceram. No other correspondence is shown between one Vocabulary and another. But there are two words worth noting. The Fagani gafo recalls the word gngavu used for cloth in the Banks' Islands and New Hebrides, the Maori kakahu, kaJiu. Another word for bark in New Zealand is hiako, which is the Samoan siapo the bark of the paper mulberry and the cloth made therefrom. In Mota, cloth is called siopa, and the origin of the word is ascribed by natives to a visit, some fifty years ago, of a party of Polynesians from Tonga, who were clothed in siojya, by transposition from siajyo ; so when Europeans appeared clothed, they used the word again. 56. Smoke. — There is no word which occupies many places in the Malay Archipelago Vocabulaiy, but there is one, the Malay asap, which represents a word very common in ^Melanesia, and in Polynesia also. The Maoi'i is au, Samoan asu, Tongan aim. In Melanesia the word is found from Ysabel to Fate in varjnng forms, aim, alw, ah, asua, asu, as, as, es, and belonging probably to these sasu, and rasu. The Malay word is probably compounded with * It is possible that the Mota wil, to peel, is the same word. Notes 0)1 tJie Vocabalai'ics. 9 1 api fire, and so we find in Lakona alimi av, in Torres Inland hiev. The Mafoor of New Guinea is aas. The exceptions in Melanesia are not numerous. The Cao gagahu is the Mota gagavu thick, clouded. 57. Soft. — The words in this list, as vni\\ adjectives generally, show the characteristic prefix ma, and the reduplication. It is the same with the Malay Archipelago words, which begin, many of them, with wa, mw, and da. The Mafoor mababoot shows the same formation. A root common through a very large part of the language area is lum. It appears in Malay lurnbut, in lumut and muruminto of Celebes, in lomo of Bouru, mulumu of Ceram, rum of My sol. The Malagasy malemy may be the same. In Melanesia it is common, in the northern New Hebrides, Banks' Islands, Fiji, Solomon Islands, up to Duke of York cjalom. Some of the exceptions in Melanesia are interesting. The Motu, of New Guinea, manoka is no doubt the Mota, of Banks' Islands, manoga. It may be very likely that their sense is identi- cal, though manoka is soft, and manoga is said of cooked food. The Sesake manukunuku may be the same word. The Anaiteum mulmul may be lum reversed. 58. Sjiear. — No doubt there are many kinds and shapes of spears, each with its own name. It is not surprising, therefore, that there should be a great variety of words. It is very remark- able, however, that in Mafoor at one end of New Guinea, Motu at another, and at Sesake far down in the New Hebrides, there is the common word io. This is not by accident. The Rotuma, which has a fancy for reversing syllables, may have turned io into oi. In the Banks' Islands spears are not fighting weapons, the names are merely descriptive : sar is to pierce, isar, after the Fiji manner of prefix i to an instrument, a piercer ; as is to stab, hence mafas, matah. The names are applied properly only to a sort of spear used in killing pigs on solemn occasions, and by analogy to foreign spears. 59. Spittle. — There is no word at all common in the Malay Archipelago. In Melanesia the word in Mota, anus, is found in fourteen places ; and this is the Samoan and Tongan anu. There is some variation in the Melanesian forms, anus, anuh, onusi, awusu, nusu, nisu, kamisu, misu, nisu, nih, nih. Though the noun in Fiji is different, the verb ' to spit ' is kanusi. In the Banks' Islands the word garmes is no doubt connected with garameai tongue. The Duke of York kara shows the word 92 Melanesiau Languages. at a considerable distance. In one instance a word belonging to the Malay Archipelago ajijoears in Melanesia, hulai of Bowru is evi- dently wura of Aurora, imvurai of Sesake. 60. 8tar. — This is an interesting word. In the Malay Archi- pelago there is no great measure of agreement ; hintang and hitnin must be looked u2:)on as kindred forms, t being sujiported by n ; hetol of Gilolo again is the same. The Malagasy kinlana is further away. The Polynesian forms show a simpler character than the Malay whetu, fetu. The Dayak betuch, though the signi- fication is different, is plainly the same word. The Marshall Islands iju is formed by the loss of the initial consonant and change of t. In Melanesia the same word occurs in twenty-five of forty places. The forms ai'e very various. With a termination it is vaitugu, vitugu, vitig ; with the change of the initial consonant hefu, he'u,Jigu ; by the change of t, mo-ijeuv, visiu, veji, 7iefu,Jigu ; by cutting off t in a way characteristic of Vanua Lava vi'. From hintang to vi is a long way, but the word is evidently the same throughout. The formation of the Fagani Jigu deserves notice. In that place the h of Wawo, three miles off, regularly turns to /, but g represents the same letter left out, perceptibly, with a gap in the sound, in Wawo. The Fagani (Ha'ani at Wa?io) ^\ov(\. jigu ought, then, to represent the Warto M'u, and in fact it represents heu. But it is very instructive to observe that the gap in the Wa?io word really means t not g, and has been filled up with g in the Fagani word under a misaj)j)rehension. It is plain that the Fagani and Wawo words are independent, because one comes from vitu, one from vetu. The interest lies in the filling up the gap with g in Fagani, because the gap in Wa?20 generally represents g, though sometimes it is in place of t. Whether all Fagani peojDle, or only the one who gave me the word, saj^gu, I cannot say ; but the mistake is interesting. It is too far to go back to an inter- change of the piimary tenues, as if Fagani Jigu came from Jiku, Wano heu from hetu, or both from a viku parallel to the common vitu. The Dayak betuch is used for the ' Sun,' but it is clearly the same word, and the original idea exjiressed can easily be conceived which would include sun and star. The Dayaks, who call the sun betuch, have the Malay bintang for a star. The two words, the same originally, have come to be particularised, as the Dayaks, needing a word for star distinct from that used for sun, borrowed Notes on the Vocabularies. 93 from the Malay. The anticjiiity of the use of hetuch as sun as well as star must be great. Tlie form of betuch is just parallel to that of vitig in the Melanesian languages. There is another word, also widely used, which has the same double signification of star and sun. The word maso is the sun in Espiritu Santo, a star in the New Hebrides and the Banks' Islands. In no single language does it signify both sun and star ; but it is sun in Espiritu Santo, star in Fate and Sesake, Ambrym, Lakona, and Vureas. In Mota it is used only for a conspicuous planet, maso maran the morning star. In Malagasy maso andro is the sun, the maso of day, and maso is an eye ^. But from the use of maso as the sun and as a star it may be concluded that maso is not originally an eye. Rather it is that the primitive idea expressed by maso is one under which both sun and star and eye can come — the notion of a disk or round. There is no metaphor of eye of day, the word maso is too old. 61. Sun. — This word shows much more variety and complica- tion in the Vocabularies. The Malay mata art shows very plainly the word mata, which, like maso above, is eye or round, and ari which is day. But there is in Amboyna and Ceram a word, the constituents of which seem the same, riamata. In Celebes it is mataalo, and in Salibabo alo is sun as in Melanesia, and matalon, Baju, is probably the same. WTiile there is alo, in common with the Melanesian languages, there is no appearance in the Malay Archipelago of the ra and la of Polynesia. In Micro- nesia, Marshall Islands has al, the same as alo. In Melanesia alo is much the most common word, supposing loa to be the same ; in Aurora and Merlav it is aloa, and often elo, in Ambrym yial. With this we have again in Api mata an eye, mat ni elo. The word does not extend further than from Fate in New Hebrides to Nifilole near Santa Cruz. In the Solomon Islands the word alio, which is also in the New Hebrides, is common. In the latter, in Lepers' Island, matan alio is used for the sun just up. The maso of Espiritu Santo has been men- tioned. In Fiji and San Cristoval we have siga and sina, mata ni siga more definite, for the sun. This is no doubt the word sina sina to shine, which appears also in the names fasina, masina, for the ' The Latin sol sun, is the same word with ^(ipios the Dog Star, and with the Irish suit eye. Curtius. 94 Melanesian Languages. moon. In dina of ^lotu, New Guinea, there is the dina Imri day break, of Malay. In mcike of Duke of York, k being hard g, we have perhaps magag, used for the moon in the Banks' Islands. 62, Sweet. — The interest of this word is grammatical ; as is usual with adjectives, there is no agreement in the Vocabularies. In the Malay Archijielago words, and in the Melanesian, may be seen the adjectival prefix of quality ma commonly occurx-iug. In the Melanesian words there is the characteristic reduplication, and the terminations, s, ca, a. The word local about the Banks' Islands is worth noting for the changes of its form. The root is tar, representing some effect on the mouth, reduplicated tartar in Aurora ; the change to d and n is seen in the other forms with sharper or thicker vowels. With redujilication, and the adjectival tennination s, the forms tetres, derderes, dodoros arise. In Gaua the causative prefix makes vadurus. Further than tliis there can be little doubt but that the Araga reterete is the same, it being so common to reverse the words, ret for ter. The difference in form between reterete and vadurus is great ; but these and the other forms of the same word are contained within a small area. From the root comes the Mota neremot, Merlav dermot ; mot is to cut or stop short, nere- mot is that which ners short in the mouth. Hence vadurus with the causative is that which makes the mouth durus. 63, Tongue. — In Mr, Wallace's list the Malay lidah with lilah, and dila must be considered the same ; and the Malagasy Ida belongs to them. This is only represented in Melanesia by the Eotuma alele, which may connect with the Polynesian alelo, arero. There is, however, another word more common in the Malay Archi- pelago which is abundant in Melanesia, me, ma. With what may be supposed to be suffixed pronouns it appears as maki, mahmo, maan, me, meem, &c. In Melanesia this word is pi'esent from Anaiteum man to Duke of York karame na tva. It is combined in the Banks' Islands and Duke of York with gar, kara, which in both places also has ap2)eared as spittle. The word lua, which is compounded with me in Eromango, Aurora, and Merlav, has the meaning of putting forth. 64. Tooth. — This is a remarkable word, because the two forms, one with I and the other with n, are so widely distributed that the change or distinction must be very ancient. There can be no doubt but that the ^Malagasy nifi/, Samoan nifo, Maori niho, are tlic Solomon Islands livo, Banks' Islands liwo. In the island of 1 Notes on the Vocabularies. 95 Malanta both forms are present, niho and livo, thougli at Alite, where they are fond of n for Z, they have the more common Melanesian form. Both forms appear in the Mahay Archii)ehigo ; nifoa in Matabello, hdif, kalifin in Mysol. Nor is there any reason why what is / in these words should not be li in nihi, and s-=li in nisi. The Malay gigi is exceptional. In Melanesia in one jilace liho becomes rilio ; liwo beoines loxuo in Vanua Lava and Anibrym. It has been mentioned that livo is tlie mouth in Bugotu. The Fiji hati is in the New Hebrides, Fate, and neighbourhood. G5. Tree. — This word is substituted for Mr. Wallace's word 'Wood,' for the reason that the Malay kayu wood, is undoubtedly the Batak hayu, hau, the Malagasy hazo tree, the Fate kasu and kau, and so all the many forms of the same word that mean primarily a tree and secondarily wood. Of Mr. Wallace's thirty-three words twenty-eight are forms of this, taking kayu to be a longer form of the word, which in its shortest form is at, ei. Of the forty Melanesian words thirty-seven are forms of the same word, alone or in combination, ranging from kasu to ie. The Mafoor of New Guinea is ai. The rakau of New Zealand, la'au of Samoa, contains the same word. If at the two extremities of the long geographical line which stretches from the Malay Peninsula to the Loyalty Islands we find words so different as kayu and ie, the statement that they are in fact the same may require some defence. But, if taking some more central position we find a word such as kai of Teor, it is not difiicult to follow the variations in the direction of greater fulness and complexity, or of slenderness and simplicity. When to the stem kai the terminations su, zu, ju, u, are added, we have kaju of Celebes, kayu Malay, hazo Malagasy, hayu Batak, gazu Gao, kasu Fate. From this to kau^ there is but little change, or hau as in Motu of New Guinea, or gau as in Es2)iritu Santo. It is the same whether a word is in the form kai, hai, wai, gai, or gae or kei. To drop the initial leaves ai as in Amboyna and Ceram, or in the Solomon Islands. And when a woi'd is so very commonly diffused there is less hesitation in admitting a variation such as ei in Mysol or ie in Nengone. It must be observed that in many words this is compounded with some other, as in Maori rakau, Santa Maria regal, the Mota tangae, the Duke of York diwai, San Cristoval hasie, Nengone ' In a dialect of Fiji, kai represents the Ban kau. 96 Melanesian Languages. sere-ie, Ambrym and Ceram liye, lyeAi. In the case of some of these the natives who use them are well aware that they are com- pound words. Thus in Mota thoI is a native orange, and properly describes the thorn ; tan mol is the trunk and body of the tree ; tan gae is the tree regarded in the same way, gae being tree, and tan the bulk of it. The Santa Maria people explain regai in the same way, re is the bulk, gai the tree. By this the Maori rakau is explained. The resemblance between two words evidently of this character is extraordinary, lyeii of Teluti in Ceram, and liye of Ambrym in the New Hebrides. ' Backbone ' is ' tree of the back,' hazondamosina in Malagasy, just as in Toba hau-tanggunmg. So in Mota the backbone is ga-togoi. 66. Water. — There is pi'obably no doubt that the Malay ayer is the word wai so common in the Malay Archipelago, universal perhaps in Polynesia, and common also in Melanesia. Out of thirty-two words given by Mr. Wallace twenty-three are forms of this word. It may be doubted whether the termination er is with- out meaning, seeing that it appears as I and li. The Ahtiago of Ceram, wai-im, is probably drinking watei", the Banks' Islands im to drink. In Melanesia wai appears also in composition, for noai of Fate and kuai of Alite can hardly be other than compounds ^. It will be seen that from Nengone to Malanta in the Solomon Islands the word is present ; but wai cannot be considered the characteristic Melanesian word. The Banks' Islands are entirely occupied by another word, pei, which itself has no other reiDresentative iji any word in these Vocabularies. In New Guinea, however, there is hey. The most interesting word in the Melanesian Vocabulary is the Motu of New Guinea rano, and Port Moresby lanu, because this is identical with the Malagasy rano, and the same with the Marshall Islands dren. The word is present also in Duke of York, though not standing for water generally ; danim is used for a river. In Fiji drano is a pond or pool of water, and ano is the same in Tonga. In Malay danau is a lake ; but in three languages of North Celebes rano is water. These are all isolated usages, and the word is a very good example of the way in which a word which belongs to the stock of languages generally maintains itself here * In Nengone kua nl hone is 'liis drink.' This can hardly be other than kii in Alite knai ; iii.ay be the Vaturawa ko. Notes on the Vocabularies. 97 and there in j)laces which can liave no recent communication witli one another. The Gaua liiou^ which may be Santa Cruz Anw, has a correspond- ing lewo in Sumatra. It is the same with the Mota ligiu fluid. Just as rano is water in ^Malagasy and a pond in Fiji, so tuxv is water in Lakona in the Banks' Islands, and tuwin is a pond in Torres Islands. These uses correspond, whether the distance which separates the varieties be a few miles or a third of the circumference of the globe. 67. Wldle. — Unlike most adjectives, this is an interesting word. In the Malay Archipelago twenty-four words out of thirty-three are the same as the Malay jyutxh ; and it should be observed that the prefix of quality is present in maputi, mopotito, maphutu, babut. This word in Malagasy is fotsy. In Melanesia it only appears in one place in Lepers' Island mavuti. The Maori ma does not seem to have any kindred elsewhei'e. In Melanesia there is no common word. In the Banks' Islands the word is local : elsewhere there is no word common to more than two or three languages. There are words, however, of much interest. The Fiji vulavula and Florida pura are probably the same, isolated in the Vocabulary, though vula is used as white in Mota, make vula a white make tree. But this word is not without representatives elsewhere : the Malagasy vola silver is probably the same; and in Gilolo wulan, in the Moluccas bulam, in Rolti and Solor near Timor fula and burang evidently correspond. These words suggest relationship with vula the ]^Ioon. Another ]\Ielanesian word, which is common also to the IMalay Archipelago, is the ^;zto of LejDers' Island, bitbit of Volow, which, with the prefix of quality, is mahida of Celebes. The word pita is used in Mota only of a light complexion. The Fiji siga sun, day, is the same word with Mota sina to shine ; the E. Fiji sigasigait and Maewo siwara white, are formed from these words. Several words, in fact, are thus common to different islands, in one in a primary, in another in a secondary signification. Tims voke is white in Esi^iritu Santo, and in the Banks' Islands %voke is a kind of nick-name for an albino ; ivedtved, tvewed, iveticet white in some of the Banks' Islands, are the Mota weneivene clean, the Fiji icede- wede ; in Torres Islands lul is white, in Mota fair ; rerea, rearea white, in the Solomon Islands, is at so great a distance as Fiji rea an albino. Thus words are, in fact, common to many languages, are in the common language of the area under consideration ; but, not H 98 Melanesian Languages. all lying on the surface of the language, are not seen till lower strata of speech are explored. 68. Wing. — There is no common or prevailing word in either Vocabulary, but there is not wanting a word common to both. The word most used in the New Hebrides and Banks' Islands, pajie, which is also the common name for a hand, is also a wing in the Malay Archipelago : opani (0 probably the article) and jmnidey in Celebes, paniii in Bouru, fanik in Teor. One of the words which occurs most frequently in Melanesia is akin to the Samoau apaau, the apaajxi of the Solomon Islands ; which is also probably in another form the gapugi of the Banks' Islands. A fluttering flight is ga2)aga2)a in Mota. It has been observed under 'Leaf that in Ulawa leaves are called apaapa ni ai wings of trees. The Fiji taba may be the same word as the Florida taha, which in that language means layers, taba ni vure people generation uj)on generation, or class above class. 69. Woma7i. — The word used for a wife is very often only woman, as that for a husband is man. There is also the distinction of woman and female, so that a word which means a woman in one language is ' female' in another. In very many words of Mr. Wallace's Vocabulary there is an agreement. In twenty-four out of thirty-three places wine, bine, fine, pin, hina, is found. This is the Polynesian ioahine, fafine, the tavine, ravine, laqavina, hoina, J'efene, vaivine, haini of Me- lanesia. The root appears in Eromanga sivin, and Ambrym vihin ; in Duke of York watvina, in New Britain vafini, is feminine. In Mafoor of New Guinea the word holds, bien woman ; in the Gulf of Papua waivine, babine, haine, sine, shine. Other words in the Melanesian Vocabulary are hard. It is not without meaning that in the Banks' Islands the words begin with the plural re, and that the Ureparapara retine woman, is the same almost as the Vureas retne mother. 70. Yellow. — Beyond the grammatical forms, the reduplication, and the adjectival terminations ga and r, there is nothing of interest in the ]\Ielanesian list of words ; and in the jNIalay Archi2:)elago list only the prefix ma. What interest there is, is of another sort. The Malay word kuning means the turmeric, the curcuma root, conspicuously yellow. Tliough the word is difterent, the notion is the same in Melanesia, where ano, out of which the adjective is made, is the turmeric. The Vaturawa mera is the word elsewhere, as iu Malav, used for red. Notes on the Vocabularies. 99 It may be well to add a few words concerning- the vocabu- lary of the Mota lang-uag-e ^^ith a view to meeting the question whether the stock of words in such languages as these is not scanty and deficient. Scanty it certainly is not, though in some ways it is very deficient. There are, as a matter of course, no names for objects which do not present themselves in the islands, but for everything there is to be seen there is a name, and for every particular action or way of doing- things. In this respect the ordinary vocabulary of a native is much fuller than that of a European, and a native language always suffers from European intercourse. An Englishman talks of 'shutting' a door or an eye or an um- brella; a Mota man uses t'lpag^, describing a downward dashing motion such as is used in striking the native shutter into its place, for shutting- a door ; he uses vataqav for the shutting of an eye, describing a closing over from above, and HI, to fold, for the shutting of an umbrella. To use tijjag for the closing of a door of European fashion is a necessary trans- ference, though in itself improper ; but natives will go on to use the word in imitation of Europeans where it becomes absurd. To carry, is used in English of any way of carrying' ; in Melanesian languages different words will always be used for carrying on the head, the shoulder, the back, in the arms, in the hand, or by two or more persons. Misuse of one of these terms wall often be most ridiculous. One who wishes to learn a native language should not be content with any native word which occurs as an equivalent to an English one ; he must find out what is the image pre- sented to the native mind by the native word, the particular thing or action it represents as in a picture, not the general class of things or actions which is in his own mind more vaguely conceived. Native languages, which are often spoken of disparagingly as deficient in general terms, are in this way fuller in vocabulary than the ordinary speech of Englishmen. Though abstract terms are not unknown in Melanesian lan- ' This is no doubt the Malay timpa to strike, used for forging iron. In Mota tipa is used for beating and breaking up stones. loo Melanesian Langttages. guages, such words are undoubtedly few, and they can hardly be expected to exist. At the same time no great difficulty has been found in expressing-, not in one word perhajis, but in a compound, the meaning of most English words, and such ideas as require words to express them in the translation of the Scriptures, at any rate in the Mota language. In making such translations nothing is to be more deprecated than the substitution of general for particular terms, or the turning of a metaphorical expression into dull prose because such a metaphor is not in native use. What can be more dismal than to translate, ' they fell by the edge of the sword ' as 'they died in war,' because natives have no swords? A true and natui'al metaphor will make itself at home among Melanesians, as images from the Hebrew Scriptures are in English. Missionary translations, sermons, and speaking are the ruin of native languages. With regard to the fulness of a Melanesian language I may give an illustration from my own experience of Mota. After some twelve years' acquaintance with the language, talking, teaching, and translating (with something of the effect above mentioned), and after having acquired more or less correctly a considerable Vocabulary of Mota words, I beg*an to buy words that I did not know at the rate of a shilling a hundred from the scholars at Norfolk Island. I left off when lists of three thousand words unknown to me had come in. It is certain that elder natives living at Mota use many words hardly known to those who have gone away from their own island as boys, and that the boys had by no means exhausted their stock. I calculate therefore that there were probably as many words still to come as would bring iip my Vocabulary to at least six thousand words. Of these many of course are compound and derivative, but they are distinct words. This concerns a small island with less than a thousand inhabitants, with whom European intercourse began within the memory of living- men. 111. SHORT COMrATlATIVE GRAMMAR OF THE MELANESIAN LANGUAGES. A COMPREHEXSIVE view of the principal grammatical forms of the Melanesia!! lang-uages makes it easy to compare them among- themselves, and to judg-e of their common character and relationship to one another ; and at the same time it supplies a convenient means of comparing" these languages with others to which they may be thought to be allied. The forms here brought together represent the Melanesian lan- guages generally which have their place between New Guinea and Polynesia. For the purpose of comparing these with the forms of the Oceanic languages generally, examples are added from Malay, Malagasy, the Maori of New Zealand, and, in part, the language of the Marshall Islands ; lan- guages which may well represent Indonesia, Polynesia, and Micronesia ^. These languages, all of them, are destitute of Inflexions, \ and thii^ives them- a common character. There are, there- fore, no Declensions or Conjugations ; there are no Cases, no Genders, and, excepting Pronouns, there is no Number or ' For this I have used Crawfurd's Grammar and Dictionary of the Malay Language ; Maxwell's Manual of the Malay Language, Triibner, i S82 ; Parker's Concise Grammar of the Malagasy Language, Triibner, 1883; Outlines of a Grammar of the Malagasy Language, by H. N. Van der Tuuk ; Grammaire Malgache, par Mari-e de Marin, Paris, 1876 ; Malagasy Grammar, by E. Baker, Mauritius, 1845 ; Maunsell's Grammar of the New Zealand Language, Mel- bourne, 1882 ; Shortland's How to learn Maori, Auckland, 1883; Archdeacon Leonard Williams' Grammar and Dictionary of the New Zealand Language ; Pratt's Granmiar and Dictionary of the Samoan Language, Triibner, 1878; Beitrag zur Sprache der Marshall Inseln, von Franz Hemsheim, Leipzig, I02 Melancsian Languages. Person. Since then these grammatical forms do not exist, it is unreasonable and undesirable to speak of them as if exist- ing'. A word in a sentence of one of these languages may be the Object, but there is no Case, and the word is not in the Objective Case ; if the Genitive relation is expressed by simple juxtaposition, or by the use of a Preposition, there is no word in the Genitive Case. Corresponding with the absence of Inflexion there is an absence of those variations in the form of words which may distinguish the Parts of Speech. It is not that there is a complete absence of such special forms of Verb or Noun; but that the same word, without any change of form, may be in use as almost any of the Parts of Speech. The use of the ^vord, not its form, commonly declares its character. ' Many Malay words must be treated as now substantive, now ad- jective, now verb, according to the position they occupj^ in the sentence ^.' This being the case it is evidently wrong to speak of a Noun as derived from a Verb, while the form is unchanged, or even a Verb from an exclamation. For con- venience words must be distributed into Parts of Speech ; but it must be understood that nothing, commonly, in the form of the word shows what it is. By way of example from a Melancsian language : in Mota siwo, the Maori iho, is ' down,' and in most common usage would be an Adverb ; but with, a verbal particle it becomes a verb, and wdth a preposition indicating place, i siwo, it is a noun ; Ke ! is an Exclamation, and yet it is used as a Verb ^. In these cases it may be fairly conceived that the words came into existence, the first as rather an Adverb, the second as a mere cry, and that the use as Substantive or Verb is posterior. But yet, as the words undergo no change in form, it is merely their use that distinguishes in the one case the Noun from the Adverb, in the other the Verb from the Exclamation. {^ In whatever way a word has come into existence, when once ' Maxwell, Mamial of the Malay Language. * Ni me ke apena he ke'd at it, as we say that one poohpoohp a thing. Short Compai-alii'c C raininar. lo;, it has come into existence it may be used as alnu)si any Part of S])eecli. In this matter there can bo no doubt but that the Melanesiau lang-uao-cs and those of the Pacific and Indian Oceans oi-enerall}^ are at one. It is highly probable that words g-enerallj'^ are in the native mind names or Nouns. The thing-, the action, or the state, receives its name. Words thus are Nouns or Verbs, and they receive discriminating marks, Articles or Verbal Particles in these languages, according to their use ; but there is no such distinction in the native mind between the thing and the action, between the visible object and the visible act, as to force them to think the name ol' an action a different sort of word from the name of a thing. Nevertheless, in all the languages under consideration, a word used to name an action or a state has a special particle attached to it marking that use, making in fact grammatically a Verb ; and in many of these languages the presence of Verbal Substantives shows that the abstract idea of the action or condition has required, and has found, a name. From words thus originally Nouns or Verbs, the Adverbs, Prepositions, Adjectives, possibly even Conjunctions, in com- mon use as such, have proceeded. Some words in Melanesiau languages which must needs in their use be called Pre- positions, still are in use also as simple Nouns. It is e\adent in many cases that what must be called Adverbs are merely Nouns. The Mota vea, Duke of York wai^ (Maori Iiea,) though they can hardly be translated except as Adverbs, ' where,' are in grammatical use Nouns. In INIota pe is in use as a Preposition ; in Lepers' Island it does the same work as a Preposition, but always in full form as a Noun. But if all Avords were in their origin names, there is a class of vocables in the ISIelanesian languages which certainl}' are not now the names either of objects or actions. These are the Particles which point in one direction or another, the demonstrative directive particles with which language itself gesticulates. These may be found separate as demon- strative particles, and probably as the simplest Prepositions ; I04 Melanesian Languages. ])ut tliey are found combined in Pronouns, in Adverbs of Place, and therefore of Time, and in Articles, If they are frag-ments of old nouns they are now nothing- but fragments of that which has been lost ; they name nothing, they only point. These cannot, like ordinary words, become, as the speaker is pleased to use them. Nouns or Verbs ; they never can have an Article or a Verbal Particle prefixed. It may be thought that the presence of these Particles, if they be fragments, shows that the language in which they are present is not in its primitive condition. At least the use of directive demonstrative particles, not im- bedded in words, but inserted continually in phrase or sentence, is the use of people who have visible in their mind's eye the actions and the things of which they speak, — a simple primitive condition of mankind. In this condition it is not only with such particles, but with Adverbs also, that language will be continually pointing to this and that, here and there, up and down, seawards and landwards^. Whether the directive and demonstrative words employed are plainly Adverbs and Pronouns in which the demonstra- tive particles have been combined with some other root, or whether the particles themselves simply are used, it is by no means easy in the Melanesian languages exactly to distinguish the place or the direction indicated. Nor for the present purpose is it necessary. The simplest particles represented by k and 11 may in some two languages point in opposite directions ; what in one language points here, in another points there : but both point and direct the mind as the linger might the eye ; both are demonstrative, and can fairly be classed together. It may be said again that the variety of meaning in these Particles, while the character- istic demonstrative force remains the same, shows rather the antiquity of their place in the various languages in which they are found. If the Mota of the Banks' Islands has ma ' ' A Lifu, comme en Polynusie, la direction vers I'intdrieur des terres ou de I'intdrieur vers la niei- joue un grand role dans la langage.' — Notes sur la langiie de Lifu, par le P. A. C. Short Comparative Gra))imar. 105 and at hither and thither, sage and siwo up and down, as the Maori of New Zealand uses mai and aiu^ alee and i/io, the close similarity of form and meaning- does not ar<^ue the remoteness of the source from which both have received the words. But if ko in the Banks' Islands points somewhere near, and ko in Santa Cruz points afar; if iiie in Mota is that, and ini in Florida, this ; while ne, na, is a general and vag-ue demonstrative in Ambrym of the New Hebrides, and Malanta and San Cristoval of the Solomon Islands ; then it appears as if widely separate lang-uag-es had received their common word, in a g-eneral not yet particularized sense, from some ancient remote original. The Demonstrative particles in the Melanesian languages may be found in Pronouns, Adverbs, and Articles, answering generally to the English (i) ' this' and (2) ' that ;' (1) 'here' and (2) ' there,' and the definite article ' the.' In the follow- ing table these are given in the simplest form, either as distinct particles or as combined. The geographical order in which the Islands to which the languages belong are arranged is that which begins with the Loyalty Islands at the ex- tremity of the Melanesian chain, and follows on to the North and West towards New Guinea and the Archipelago of Indonesia. 1. Demonstrative Particles. Loyalty Islands. Nengone o, ko, no, le, Lifu la, ke. New Hebrides, Anaiteum Fate .... Sesake .... Ambrym . . . Espiritu Santo . Pentecost, Raga Lepers' Island, Oba Aurora, INIaewo 1 ki, 2 ko. wa, na, ga,. wa, wo, na, c/a, se. He, Ha, ge, le, li. ne, na, ka. ke, ko. I naha,, 2 nehi. I ka, 2 la. Banks' Islands. Merlav i ke, 2 we. Gaua I kere, 2 keren ; i kose, 2 kosen. io6 Melanesian Languages. Lakon i og, 2 rek. Vanua Lava . . . . i le, 2 no ; i ko, 2 lo ; i es, 2 ne. Mota I ke, 2 ne. lo. Motlav I ke, 2 nen ; i igoh, 2 hanen, Volow I iges, 2 ena, wo. Ureparapara . . . i ke, 2 ne. Torres Islands . . i ke, 2 na. Fiji i qo, 2 qori ; i eke, 2 kikea. Santa Cruz i lo, la, 2 ko, ka. Nifilole I keli, 2 ela, kala ; i emi, 2 e»a ; i la, 2 li. Solomon Islands. Ulawa I aka, 2 kala ; i ga, 2 la. we, ho. Wawo I nani, 2 nasi ; i ini, 2 esi. Fagani i «a, 2 ni. Saa iiQ. VaturaMa . . . . i wene, 2 nuMu. Florida i eni, ini, 2 keri, iani, ya. Savo lo, la. Bugotu I iaani, 2 iaweni. Gao I ani, 2 igno. Duke of York . . . i kumi, 2 kuma ; i kuri, 2 kura. The g"eneral result of the bringdng- together these Demon- stratives is to show that in the thirty-two places represented some form of a particle of which k is the characteristic occurs in twenty-one. Some form with n occurs in twenty-two. In ten places a form with / occurs ; in five places a form with 8. The forms with k and n are very g-enerally, almost uni- formly, distributed : those with / appear in each group, except Fiji. Those with .?, or its equivalent //, though so much more rare, isolated, and distant one from the other, are the more interesting, because it is impossible to suppose that they have been communicated directly from one of these groups to another. Attention must again be called to the fact that there is no fixed meaning to the j^articlcs with /?•, w, or / ; they jwint, direct the view, demonstrate, everywhere, but generally ; and when they particularize, their particular force is local. In the Banks' Islands generally k points to 'this ' or 'here;* Short Comparative Grammar ro7 in Santa Cruz to a more distant object or ])lacc : tlic rarer s in the Banks' Islands points near, in San Cristoval afar. Bring-ino' into compai-ison with these jNIelanesian demon- stratives Pronouns and Adverbs of Place of Polynesian, Malayan, and Micronesian lang-uag-es, we find as follows : — Malay, ini this, Hu tliat, s'lni here, sifn there. Malagasy, ify this, my that, itsy that near, iry that afar, aty here, ary, any there. Maori, ne'i this near me, na that near you, ra that afar. Marshall Islands, kein this. In these the particle na, ne, ?ii, so common in jNIelanesia, is conspicuous. No form with k appears ; and, unless r has taken the place of it, no I. In Maori and in Malag-asy rl and ra point afar ; which may very well be li and la of Melanesia. But the demonstrative Pronouns kere of Gaua in the Banks' Islands, keri of Florida, and rek of Lakona, give something- more exactly resembling, being probably compounded of ke and re or ri. 2. Articles. The Definite Article is in itself a kind of demonstrative, and it is natural to expect, in these languages as in others, a likeness in the form of the Articles to that of the demon- strative particles. The following table, which gives a view of the definite Articles in use in Melanesia, will show that this likeness to a considerable extent exists : — Loyalty Islands. Nengone . . . re. New Hebrides. Anaiteum n-, in Fate . . . n-, in Sesake . . . na. Pama . . . a, 0. Espiritu Santo na, a. Oba ... na, a. Maewo . . na, a. Banks' Islands. Merlav na. Gaiia . . . na, u. Lakon . . . en. Vanua Lava . n-, na, o. Mota . . . na, o. Motlav . . n-. Volow . . . n-. Ureparapara n-. Torres Islands n-. Fiji na, a, ko, o Santa Crcz . . na, te. Nifilole . . n-. io8 Melanesian Languages. Solomon Islands. Ulawa . . . na. Florida na. Wawo . . . na, e. Savo . . . lo Fagani . . na, a. Biigotu . . na. Saa .... wa. ffao. . . . na. Vaturawa . . na. Duke of York na, a In these the predominance of na cannot fail to be observed, and it can hardly be doubted but that it is the Demonstrative particle so conspicvious in Pronouns and Adverbs. In some languages it coalesces with the Noun, and is written with it in one word, as in Anaiteum and Fate, where it has not even been recognised as an Article at alP. In some languages it shifts its vowel according to the first vowel of the noun which follows, as in Motlav, na tar., ne tenge, ni til, no to, nu Jmg : or, as in the same language, it parts with a vowel altogether before a word which begins with one. But almost throughout Melanesia some form of na appears. In Nengone re, in Santa Cruz te, are probably borrowed from the Polynesian settlements close by ; but in Nengone re is always accompanied by the Demonstrative o. In Savo lo is used certainly as an Article, but is plainly the Demonstrative. In Fiji alone ho appears, but o is probably the same. The Articles of the Oceanic languages which have been brought in for comparison are : — Malagasy, ny. Maori, te definite, he indefinite, nga Plural. In Samoan re is in the definite Ai-ticle. In Malay no Article is used. In Ambrym and Araga, of the New Hebrides, and in Lifu, of the Loyalty Islands, no Article is found. In the Malagasy ny there is no difficulty in recognising the particle that appears in my ' that,' in that language, and in so many of the Melanesian Articles. 3. Perso7ial Articles. These stand altogether apart from Demonstrative Particles ; but they are so common, though not universal, in Melanesia, ' ' The Rev. J. Copeland, an accomplished linguist in the New Hebrides, Bays, "Inthe Aneityumese language all the Nouns, with scarcely an exception, begin with in or n." ' — ^Dr. Steel's New Hebrides. Short Comparative Grammar. 109 and so characteristic, that they require and deserve observa- tion. They are called Personal Articles, because they are with Proper Names of persons what Articles, definite or indefinite, are with common Nouns. They accompany the name, not in any way qualifyin<^ it, except as poinlinia veivale if you mean ten houses, though this is often said by foreigners. If you mean ten groups of houses, then the phraseology is correct.' — Rev. L. Fison. I50 Melanesiari Languages. 18. Prepositions. A A-iew of the Prepositions of the Melanesia!! lang-uages may be given here, because many of them are in fact Nouns, and some of them are Nouns in form by construction with a suffixed Personal Pronoun. Prepositions may be divided into those which are simple particles, and those which can be recognised as other parts of speech. These latter again may be seen to be either Nouns or Verbs. The Prepositions can be classified as Prepositions of Place, of Motion to and from, of Relation, Dative, Genitive, and Instrumental. In many cases it is not possible perhaps to separate a Dative Prej^osition from one of Motion ; but in some languages the distinction is so clear that it is right to make it conspicuous. ( I ) Prepositions that are simple Particles thus classified are : — Loyalty Islands. Place. Motion to. Motion from. Relation. Dative. Geni- tive. Instru- mental, Nengone i bi du, so ni Lifu e koi e, i, > New Hebrides. AnaHeum u, an u, an Fate e, to toll gor, me ni ki Sesake a, e ki ki 90V0 ni Ambrym a tene me, pe, ta ne Espiritu Santo a sur tan sura, goro sur, ne ni ni Araga a, la nin huri huri gin Lepers' Island a, lo den ta, me, huri, goro lawe gi Maewo a, le dani Ban be, ta, me, goro Ks' Islands, sur gi Merlav a, i, le sur dan mi, ta, be, gor min, sur ni, gi Gaua a, i, le den ta, be, ag, gor min, sir ni Lakona a ten to, mi, at mun, uh men Vanua Lava a, e, lo, le nen, 'en ta, me, gor mi, sir mun, min Mota a i, sur nan ta, ma, goro mun mun, nia Motlav a, 1- den be, gor mi, hir mi Volow a, 1- dean ma, ta, gor bev, hir me Ureparapara a, le, re, mi den to, ta, be, gor hiv mi, ne Torres Islands a, li den te,pi, hi, mi, goro hiv mi Fiji e, i ki ni e Santa Cruz nia ma Nifilole »a na ffo Short Comparative Grammar. i=;i Solomon Islands. Place. Motion to. Motion from. Relation. Dative. Geni- tice- InstrU' mental. Ulawa man mai, ana mnni, suli ni ana Wa»o i, nai tai.bani horo tana, bei, suri i ini Fagani taHa ni gini Saa ana, pe huni, suli i ana VaturaMa tani ni, na hini Florida ni nia Bugotu i, kori ni, i nia (?ao hara na Duke of York a, i, u ko, taka kup na ma. Among" these Prepositions are some which there will after- wards appear reason to believe are not simple words con- cerning which no explanation can be g-iven ; some v/hich may be taken as really Nouns, thoug'h in the languag-es in which they occur they are simple Prepositions ; such, for example, as ma and ta. Since Prepositions may generally be taken as a comparatively late product of speech, and generally as having their origin in Nouns, it is only natural that, in bring- ing together the Prepositions of cognate languages, we should not only find many of them the same, but find them in different stages of development. Suppose a common stock and origin of the Melanesian languages, it would be natural that in one language ^a or ma should be still in sjTitax a Noun, while in another it is a simple Preposition. It is in- conceivable that a word imported and borrowed from a language in which it is a mere Preposition should have been tm'ned by the borrowers into a Noun. (2) The Prepositions which are plainly Nouns in the languages in which they are in use as Prepositions are as follow : — Locative. Relative. Dative. Instrumental . ri hne, ae, we, ba hne 1. Nengone 2. Ambrym 3. Lepers' Island 4. Gaua 5. Vanua Lava 6. Mota 7. Motlav 8. Ureparapara g. Santa Cruz ra, lo lo be, me gi, n i pe pe be ba Locative. Belative. kone ta ta ta ta ta, iia 152 Melanesian Languages. Dative. Instrumental. 10. Vaturawa 11. Florida 12. Bugotu 13. Gao 14. Duke of York It is here taken as a proof that a word is used as a Noun, that it takes a Personal Pronoun suffixed, because such a construction is in fact that of two Substantives, the second of which stands in genitive relation to the first ^. It is a proof also if one of the words, otherwise used as a Preposition, takes a simple Preposition before it. For example, pe in Mota is used as a Preposition, and according-ly appears in the list above. But the form apena, in which pe is unquestionably the stem, shows a suffixed Pronoun na, and also the Preposi- tion a prefixed. The word is not written a pena because ape is commonly used as a Preposition ; and this is itself compound, consisting- of a the true Preposition, and pe^ which is a Noun. The translation of ^^ cannot be given perhaps in English, but 'by' may be taken as sufficient to show the formation of apena. The Preposition a is ' at,' the suffix na is ' its ;' the whole word, therefore, is ' at its by,' ' thereby.' The compound Preposition ape is thus intelligible as ' at by,' ' by,' being made for the occasion into a Noun. To go through these Prepositions, which are really Nouns, in the languages to which they belong : — 1. Nengone. — ri appears as a true Preposition, bo7ie ci sere ripa-via he stands at the door ; and is shown to be a Noun in the sentence Into rine piiha within the box, when lino is a Preposition, and ne is the third Person Pronoun suffixed to ri, in the inside the box. hie, with the suffixed first Personal Pronoun hnegn, is 'by me,' in an instrumental sense, and ' with me ' in the relation of place. se is similar, ci hie seyo he lives with me. ' In the same way some Prepositions in Mota may be seen to be Nouns by the absence of the Article: ape nayok before my face, not ape nanagok; here a is the true Prefosition, j?e the first of two Substantives, of which nagok is the second. Short Comparative Grammar. 153 7<.' r ga, ra, ta Ureparapara a, ra Torres Islands ga Fiji a, ta, Solomon Islands. Ulawa a Wa«o 'a Fagani ga Saa 'a Vaturawa ha Florida ga Savo sua? Bugotu ga Duke of York ina. g From this list several languages of those which are under consideration are absent — notably those of the Loyalty Islands, Ambrj-m, and Santa Cruz ; while that of Savo is an exception. There is a certain significance in this, inasmuch as it is pre- cisely in those languages that such archaic characteristics have been observed as the use of Nouns for Prepositions, and a less full form of Pronouns. It may certainly be thought that the formation of Adjectives by suffixing a termination, w^hether a mere particle suffixed or a word ^^■ith a sense of its own, to a Noun, would come comparatively late in the history of a language. It cannot escape observation that the Adjectival termina- tion is almost everywhere ga, g, a, or gi ; ga being emlcntly represented by g and a. There is no language which uses 1 68 Illelauesian Languages. this termination so boldly as that of Florida, and so well offers illustration of the common use. The natives can make an Adjective at will by suffixing" r/a to a Noun or Adverb, so much so that there are no fixed words with this termination ; but it can be applied to any words except to Verbs, as in English 'y,' 'ly/ 'ish,' or 'some' are suffixed. Thus a room lig-hted with candles, hiihi, is said to be huluga, as we might say ' candlesome ' as well as ' lightsome ; ' a collar with long points is kuliga, from kuli an ear ; a deep well is horuga from the Adverb horn down. In other languages, as e.g. in Mota, there may be Adjectives ending in ga which are evidently words of this kind, though the nouns to which the ga is suffixed are not in use, such as agaga white, iurturuga blue, faniuiga straig-ht. The other terminations, such as sa, ra^ fa, U, show no dif- ference in signification. In Fiji idouloa is maggoty, from nlo, dukadukali dirt}^ from cluka dirt, dregadregata gluey, from drega glue. In the Banks' Islands ligligira fluid from ligm a fluid ; sasarifa even, from sar to match. In Duke of York kibagina is white like lime, from kibag lime, rumaina full of houses, from ruma a house. In looking at other Ocean languag^es for comparison we see that in Malay an Adjective is merely a qualif}dng word with- out form or character of its own. Nor do Vocabularies of words from the Indian Archipelago show any Adjectival ter- mination. The Malagasy equally fails us. In the Polyne- sian languages, however, there is something for comparison and illustration. ' Maori Adjectives have no peculiar or appropriate form ;' but in Samoan the addition of a to a Noun makes an Adjective, as 'eleelea dirty, from ^ele'ele ; faUifatua stony, ixoxafatu a stone ^ It is at once evident that this is the termination ga, so common in Melanesia, and it deserves notice that the Samoan termina- tion is not '«, as if k had been dropped, as in Wawo or Saa the break is heard where g has disappeared. This is one of those cases in which the Melanesian might not unreasonably be * In Tongan gele mud, gelea muddy. Short Comparative Grammar, 169 thought to have borrowed from the Polynesian. But the Mehmcsian has ga^ ra, sa, ia, li, as Adjectival terminations, of -which the consonant is evidently the formative part, since y and r are used without a vowel ; and the Poly- nesian has only a, and that in not all Polynesian lang-uages. It is not to be supposed that ]\Ielanesians borrowed f/a or g from a. There are words which have been called compound Adjectives, two Nouns, in fact, in juxtaposition, of which the second qualifies the first, which, as there is nothing- in the compound form dif- ferent from an ordinary compound Noun, are really not more Adjectives than simple Nouns are. In English such compounds may be classed as Adjectives, but it is ver}'' undesirable to do so here. The same may be said of words compoimded in Fiji with the prefix dan, no doubt the same word as the Torres Islands to, to-wuwuh fond of beating, and the same sort of word with the Mota man, man-vus fond of beating, and with Duke of York tara, taradono full of leaves. These, whether compounded of two Substantives, Verb and Substantive, or two Verbs even, make one word, and that a Noun used to qualify. The common causative prefix vaJca gives in Fiji a sense not commonly given in other Melanesian languages, vaka vale 'having a house.' In YsLtefaka is used to make an Adverb, but generally the prefix belongs to Verbs. With Verbs also it will be better to connect the prefix of condition ma, though the word resulting from the prefixing of it is often only an Adjective. This prefix is very common in Melanesia, and in Polynesia also ^ ; as in Mota save to tear, masare torn, in Sa- moan lilirji to pour, the Mota lixs., malij'i spilt. Similar pre- fixes are ka and ta. Though these words often have the appearance of Participles, ' torn,' ' sj)ilt,' &c., yet they are not always formed from Verbs, mavinv'm thin, matoltol thick. The Malagasy words with this prefix ma arc classed as Adjectives. Many roots in that language form Adjectives of quality by 1 Compare the servile m in Semitic languages, prefixed to the verbal root to make Participles. 170 Melanesiaii Langtiages. prefixing- ma, loto dirt, maloto dirty ^. Compare Mota nara blood, manaranara bloody. See also the Adjectives in the Vocabularies. 21. Terls. The two broad divisions of Melanesian speech are Nouns and Verbs, the Names of Thing's and the Names of Actions or Conditions. In Melanesian languages a word is marked as a Verb by its being used with a preceding Particle, the office of which is to declare, very often with something of Tense and Mood, that it is an action or condition that is spoken of, and not a thing. It is true that in all the languages it is possible, in some it is common, to use a Verb without a Par- ticle ; but all these languages use Particles with Verbs. Besides these Particles used with Verbs, and in these lan- guages generally written distinct from the Verbs, there are Prefixes and Suffixes written with the Verbs which make a change in their signification. It will be better to leave these till later and deal first with that which is purely Verbal ; the addition of which marks certainly the presence of a Verb. 22. Verbal Particles. Verbal Particles always precede the Verb. It makes no difterence whether they are written in one with the Verb, or separate, except in so far as their separation keeps the Verb itself more clearly in view^^. In Melanesian languages the Particle is written apart, except where the contracting ten- dencies of the languages, as in Motlav, make it impossible. There are Particles also which follow Verbs, but these are of the character of Adverbs. The Verbal Particles themselves cannot be called anything but Particles. The Personal Pronouns in many languages coalesce with them, or influence their form, or accompany them, even sometimes replace them, and for this reason in ' Parker's and Baker's Grammars. * If Particles of this kind were written separate there would not be so many cases of Infix discovered. SJiort Comparative Grammar. i 7 1 some lanofuages they may apj^ear to be forms of Pronouns. Bishop Patteson at one time classed some of them with Pro- nouns. But, when taken as a whole, includino: the languag-es in which the Particle in no way varies with the Person or the Number, and considering- the universal use of them in the Melanesian, Polynesian, and Malag-asy lang-uages, and their presence in ISIicronesia, it will be seen to be impossible to class them as Pronouns. Much less can they be called Substantive Verbs, as has been done in the New Hebrides ^ ; any extended \dew of the Melanesian languages, even without including others, must be thought to make it impossible. In every Melanesian language here considered these Par- ticles are, more or less, in use. There is one marked division according to which they may be classified, viz. those which are invariable in form in each Number and Person, and those which change. These latter again have to be divided into those that change in accordance with the characteristic vowel of the Pronoun, and those that change inasmuch as they coalesce with some short form of the Pronoun. It must be particularly noted that the Particle of the third Person singular neither shows a vowel in sympathy with that of the third person Pro- noun, nor coalesces with it. Before proceeding ftirther it will be well to give examples of these three kinds of Verbal Particles. 1. Invariahle. Mota ^oe, Fiji sa, the most common Particles, are the same in all Persons and in each Number ; nan we valago I run, Mota, an sa lalco I go, Fiji. The Verb, in the same Tense, will be throughout roe valago, sa laho. 2. Tariahle by change of Towel. In Florida and the neigh- bouring islands this change is most regular : — Singular. Plural. 1. iV^rtM^M ios« I speak. \.\ •^ ' vigami fat bosa, excl. 2. Igoe to bosa thou speakest. 2. Igamu tau bosa. 3. Agaia te bosa he speaks. 3. Agaira iara bosa. Here what may be taken as the characteristic Particle is fe, * Eev. J. Copeland, in Dr. Steel's New Hebrides, 'The Verb "to be," as in English, conjugates the verbs through all their moods and tenses.' 172 Melanesian Langtiages. the vowel of which is not that of the Pronoun. In the other Persons it is plain that the vowel with t follows that of the Pronoun, Here also two observations must be made, (i) The presence of t in every form of the Particle, and never, except once, in the Pronoun, shows that the Particle is not a Pronoun. (2) When the character of the Personal Pronoun has been conveyed to the Particle by its characteristic vowel, there is no need for the Pronoun to be expressed at all. In Florida tu bosa, fa hosa, is ' I speak,' ' we speak,' without inau or igita. In Mota or Fiji, where the Particle is invariable, this could never be ; we vava, sa vosa, is the verb ' to speak,' but a word only, not a sentence ; there is no subject indicated. 3. Variahle hy coalescing with the Pronoun. The Northern Islands of the New Hebrides g-ive examples, as Lepers' Island : — Singular, Plural, , i\T J T -i . \Dam toga, incl. 1. jyom toga I sit. I.-^ ^ ' [Gam toga, excl. 2. Gom toga thou sittest. 2. Mini toga 3. Mo toga he sits. 3. Eajn toga. In this mo in the third Person contains nothing of the third Personal Pronoun ne, and is the representative Particle. In the first Person Singular no is a short form of neto, in the second go of in/go ; in the Plural da represents gide (of which de is the pronominal stem), ga is gamai, mi is the pronominal stem in gimi, ra is the pronominal stem in n.ere : the Particle is -m.. Since the Pronoun is in fact, as in Florida, present in these forms of the Particle, nom toga is a sentence without requiring* another Pronoun. The avoidance of a Pronoun in the third Singular is remarkable : the Particle is, in fact, w, and the natives will not assume e or a from the Pronoun, for the beginning, but fill up the end with 0. From the manner in which the forms of the Particles vary, when they do change, it may be conchided that tlie invariable Particle is of the older use. The languages, it may be observed, in which variation occurs are comparatively few, when the whole area of the languages is regarded. In some of those Short Co7npai''ative Gravimai''. 173 languages where the Pronoun coalesces with the Particle we shall see that it does not do so with all the Particles. In some languag-es also the Particles here given cannot he ex- actly ascertained, though enough is shown to prove that Verbal Particles are in use. The extent of the use of these Particles varies in different languages, and this too is not easy to ascertain. In the languages of which the writer has any considerable knowledge rules can be made out according to which the Particles are dispensed with ; but concerning many others it must suffice for the present to know that the Verb is used with these Particles. As it is by means of these Particles that a word expresses itself as a Verb, so it is by means of these that the Verb exercises its power of expressing Tense and Mood. It is by no means common in Melanesia that Particles should be used to express ^Mood ; such is found to be the case only, in fact, in the Banks' Islands and Fiji. The use of them to express Tense is much more common, and yet there are some which depend entirely upon Adverbs to declare the time. jSTor must it be understood that the expression of Tense is very definite. For the Present Tense it may be said that there is no Par- ticle ; what is used is indefinite in regard to time. If nothing is added to fix the point of time, it may be assumed to be the Present ; when the point of time has been fixed by an Adverb, or by a definitely temporal Particle, then the Verb is carried on by the indefinite Particle without the expression of Tense, but ^^•ith the understanding that the Past or the Future is in view. This is the case not only in narrative, but even when only two Verbs are connected by a conjunction ; ' shall eat and drink ' in Mota is te gana wa we ima, te being the Future Particle and loe the Indefinite. In the Banks' Islands, jSTorthern New Hebrides, and at Nengone there is a Particle which expresses continuance, t'l, ji, ci, giving the sense of continued action or condition, which is also used in narra- tive. In the following table, which exhibits a view of Verbal Particles in use in JNIelanesia, those in which the vowel 174 Melanesian Languages. varies in agreement with that of the Personal Pronoun are marked ^ ; those with which the Personal Pronoun coalesces are marked | ; a third kind, not yet described, are marked §. These last occur in languages in which the vowel of a Particle shifts to match the first of the Verb to Avhich the Particle is prefixed. In Motlav, as the Article shifts its vowel in this way, so do the Verbal Particles. But these Particles are substantially the same as the Invariable Particles, and must be by no means confused with those in which the change depends upon the Pronoun. For this reason they were not mentioned in the division established between those Particles which change ^^dth the Person and Number, and those that do not ^. Table of Verbal Particles. Loyalty Islands. Of Modal, Indefinite. Past. Future. Of Contin- uance. Nengone Lifu me a, e, ka ha, hna na CO tea ci New Hebrides. Anaiteum a, i Fate e, u, te Sesake *e, u, t Ambrym f ma, a, e, te, ve Espiritu Santo mo, mu, ga Araga tma, me, tn i, tvi Lepers' Island fmo, ga fna, fvi, i Maewo u, mo ta ni, fn ti Banks' Islands. Merlav nu ma sa ti mi Gaua ve me te to, qe Lakona e, ft en te Vanua Lava ga, ge, §g - me, §m- te, §k- ti Mota we me te ti ta, qe Motlav -k, §n- §m- §t- pe, mu Volow §n- §m- §t- * For example, in Motlav the Particle n- with shifting vowel is no before goh, ne before we ; the Particle t- is ta before rai\ in every Person and both Numbers : whereas ta in Florida has the vowel without any reference to the following Verb, but \\holly to the Person of the Pronoun. Short Comparative Gramvtar. 175 Indefinite. Past. Future. Of Contin- uance. Ureparapara §k- §111- te, ji ji Torres Islands na, ve §m- te Fiji e, sa a, ka na Santa Cruz ka, ti na Nifilole ki na Solomon Islands. XJlawa *a Wa«o *a Fagani *a Saa *e, *ke Vaturawa *e *ke Florida *e, te *ke Save na Bugotu *e, ke (?ao e, te, we Duke of York i a n. Modal. mo, me This table does not perhaps exhibit all the Particles which may be found in use, nor does it possibly show them all cor- rectly. But it is enoug-h to show that the use of Particles with Verbs is characteristic of the INIelanesian lang-uag-es as a whole. To observe carefully so characteristic a way of forming or using a Verb is evidently important. For details the Grammars of the several languages must be consulted. It is plain here that, with great diversity in the particular Par- ticles employed, there is a remarkable agreement in the use of them. 23. It becomes, in the next place, necessary to look to the other Ocean languages for comparison. Among these the Malay is disappointing. According to Crawfurd the Verbal Particle ma is in very frequent use prefixed to Verbs, and is a sign to distinguish a Verb ; so much so that the use of ma makes the difference between Verbs and other Parts of speech^. 1 In Maxwell's Manual of the Malay Language, the particle ma is not mentioned ; the ' inseparable prefix me (in aU its various forms meng, men, menif) is the mark of a Verb which expresses an action.' 176 Melanesian Languages. Yet no prefixed Particle avails to g-ive Tense or Mood to a Verb. The practice also of writing the Particle ma with the Verb adds to the difficulty ; makan is given as to eat and mimcm to drink, words which the Fiji kana, Mota gana, to eat, imi Florida, nn Mota, to drink, clearly show to exist as Verbs with- out the Prefix. It is therefore not easy to determine whether the Verbal Particles of Melanesia are present in IVIalay. In Malagas)/, though a certain obscurity belongs to the practice of writing the Particle in one with the Verb, Verbal Particles appear which change according to the tense : mijeri/ alio I think, n'ljery alio I thought, k'ljery alio I shall think. By writing nnjery it becomes if not necessary at least natural to say that to form the past vii is changed to ni ; for m'ljerij is taken as the Verb, jerij as the root. By writing mi jery separately ml is shown as the Verbal Particle. The Prefix or Verbal Particle may be ma, na, ha, as mahay (know), nahay, haJiay, or mo, no, ho. The main point of comparison is the common use in INIalagasy and the INIelanesian languages of Particles prefixed to Verbs which change to mark the tense. As in the Melanesian languages these Particles are used when a prefix, reciprocal, causative, conditional, is taken before the Verb : mankafia is to love, mampankatia to cause to love, the causative prefix is said to be intercalated, ma remains before the causative. In Mota tape is to love, with the Verbal Par- ticle we tape, vatape to cause to love, and this with the Verbal Particle loe vatape ; loe corresponds to ma, va to vipa, the true Verb is tape and katia. In this there is the double cor- respondence of the Verbal Particle and the causative prefix. In the Philippine languages the prefix m changes into n to mark the past time. In Maori ' the Verbal Particles are words which have no meaning in themselves, but which, prefixed to a word, endue it with the qualities of a Verb ^.' The Particles are e, i, ka, k/a, kua. Of these ka 'is independent of time ; merely giving to the word to which it is prefixed the force of a Verb ; e is ' Dr. Maunsell's Grainiiiar of tlie New Zoalniul TjaiiLTiiaij'e. Short Comparative Grammar. i 7 7 used chiefly with the future, and accompanied with the suffix ana forms a present imperfect : / denotes the past indefinite ^' The ag-reement in the use of Verbal Particles with the !Mela- nesian lang-uages is complete. The Samoan Particles e, te, na^ sa, marking" the tenses of Verbs, are of the same character. In the !Micronesian languag-e of jNIarshall Islands the par- ticle e before Adjectives is no doubt the same ; e lap large, corresponds to the Mota we lava, in verbal and in gram- matical form. It has been already said that the verbal form is used \\ith most Adjectives in Melanesian languages. 23. Verhal Suffixes. To continue the consideration of the forms of words which make them or show them to be Verbs, the terminations — the Suffixes, not the Prefixes — must be taken in hand. The Prefixes to Verbs are common to them and other parts of speech, so far at least as that words with these Prefixes are not always used as Verbs. In the case of the Suffixes it is not so ; the Verbal Suffix marks a word as a Verb. It needs not to be said that all Verbs have not Suffixes ; a Suffix is added to a Verb to chang-e in some way its signifi- cation. It may very well be that a Verb with a suffixed termination may be found in a language in which the A erb without the Suffix is not found at present, but the stem is a Verb, and the sig-nification of the Suffix will be felt in the meaning- of the word. The suffixing of terminations to Verbs is practised to a great extent in Fiji ; and the system according to which it is practised is set forth at great length in Hazlewood's Gram- mar, It is not diflicult, however, to put it more briefly. I. An intransitive Verb receiving one of the Suffixes becomes transitive. 2. A transitive Verb with one of these Suffixes has its action determined upon some definite object. Thus (1) moce to sleep, mocera to sleep upon, lako to go, lakova to go in. Intransitive Verbs thus become transitive, the Suffix ' Shortland : How to learn Maori. N 178 Melanesian Languages. conveying" the action on the object, as a Preposition does in English. (2) Transitive Verbs are determined upon definite objects, which therefore have the definite Article ; caka to work, caha ^oere to work garden generally, cakava na tvere to w^ork a garden, the garden, some garden. If in the first case the Suffix appears to be equivalent to a Preposition, in the latter case it is seen to be not so. \nform the Fiji Suffixes are divided into two classes ; it being understood that the language does not close a syllable. The one class consists of a suffix of a, ca, ga, ka^ ma, na, ra, ta, va, wa, ya ; that is, of almost any simple consonant with an accompanying vowel. The other, of caka, kaka, laka^ maka, raka, taka, vaka, waka, yaka ; that is, of almost any simple consonant with aka. It is important to observe that Verbs take these termina- tions indifierently ; that is to say, there can be no rule found to determine what termination a Verb vriSS. take, and no par- ticular sense can be assigned to any termination. Hence it must be concluded that some sense of the fitness of some Suffix to a Verb, in sound perhaps, has fixed the native habit of using that Suffix with that Verb ; and further, that it is in vain to seek for a special meaning in each Suffix. A larger comparison, however, than can be made in one language is necessary to establish these conclusions. These Verbal Suffixes are as largely employed in the Banks' Islands as in Fiji, and they are in fact the same ; but they hardly appear to be used with so much exactness of definition. To take Mota as an example ; the way of using the Suffixes is the same in making intransitive Verbs transi- tive, and determining the action of transitive Verbs upon an object. In form there is only the difference which belongs to the character of the language as allowing a close syllable. The one class of terminations consists of a Consonant, almost any Consonant, but most commonly g ; g, n, n, r, s, I, v ; the second class of ag, and ag with almost any Consonant ; gag, lag, viag, nag, nag, rag, sag, fag, tag. It is equally impossible to say in this language that any Short Comparative Grammar. \ 79 one of these terminations has a sense of its own, wliich may be supposed to be derived from some orig-inal word now be- come a Suffix. Attempts have been made. Mr. Ilazlewood attempts to make the Fiji suffix m, which also means ' g"0 ' in Mota, applied by rule to Verbs of motion ; but he confesses that 'many other words besides those of motion take va\ adding-, ' but for these perhaps there is no rule.' Many Verbs in Mota with the Suffix v can be made to show a sense of motion, but as soon as others are adduced with the same Suffix which have no sense of motion, it is apparent that what sense of motion there is does not lie in the Suffix. If vanov to put, sogov to give freely, may seem to contain a sense of motion, tanov to touch, sarav to rub, vafaqav to shut down, have not ^. INIoreover a Suffix which in one lang-uag-e is used with a Verb in one sense, in another neighbouring' language is used in another sense. Thus in Mota roviotag is to hear ; the same stem and Suffix is in Fiji rogotaka, which means to tell, report : the stem rono, rogo, is to hear as an intransitive Verb ; tag in the one language gives the signification of listening to something, taJca in the other gives that of making hear. It is a proof of the same kind when the same taka in Fiji can with one Verb be represented by the Preposition ' in,' with another Verb by 'with;' sokotaka na waga sail in a canoe, kabaiaka na matau cKmb up with an axe. Or a Verb without difference of meaning takes two Suffixes, as Mota saromag and sarovag mean equally to sheath. The follo-^-ing table will show how characteristic these Suffixes are of the Melanesian languages ; the Santa Cruz * The Verb tano is to touch, and needs a Preposition after it to connect it with an object, tano ape touch at or to ; but the suffix v enables the Verb to work directly upon the object without a Preposition, the form tanov describes not mere touching, but touching at work on something. There is no more motion in one than in the other. So taqa is to incline downwards, with a certain idea of motion possibly in the ' downwards ;' va is the causative prefix, v the suffix makes the Verb mean to make something incline downwards ; whatever motion is in the word was there before the suffix was added. Sarav is to rub something definite; there is movement indeed in the act of rubbing, but generally, not with regard to one definite object. N 2 I So Melanesian Lano-uazes. group alone is not represented. The two classes of Suffix are kept distinct, though there is no difference of use or meaning between them. Melanesian Verbal Suffixes. Loyalty Islands. Consonantal. Si/Uahic. Nengone ne, ni, ti Lifu n .. New Hebrides. Anaiteum frai, raig, jai, jaig, naig, L taig, haig Fate si, ki reki, naki Sesake ti, i, vi, ki .. Ambrym ta Espiritu Santo V tag Araga i tai, mai, rai Lepers' Island hi, si tagi Maewo si, gi va, ragi, nagi. Banks' Islands. Merlav n, r, t va, ra, la, na Gaua g, n, V, r, t, s, n fvag, sag, tag, lag, mag, I rag, gag Lakon g, V sag, vag Vanua Lava g, n, r, n, t te, se, re, ve, me, teg, leg- Mota V, r, t, s, n, n, g fag, gag, lag, mag, nag, wag, L rag, sag, tag, vag Motlav g heg, teg, geg, veg Volow g. 1- via, nia,, hea, tea, rea Ureparapara n, V te, sa, ra» Torres Islands g. J, t te, ge. Fiji. a, ca, ga, ka, ma. na, ra, caka, kaka, laka, maka, raka, ta, va, wa, ya taka, vaka, waka, yaka. Solomon Islands, Ulawa si ., Wa??o si, ri, hi, ni Fagani si vagi Saa si, hi, ni hai Vaturana ni, si, li, hi, mi, vi vahi Florida fti, li, vi, hi, ni, I si, vi, «i mi 1 vagi, lagi, hagi, sagi, pagi 8avo li Bugotii ni, \ i, hi, ti, ri hagi, lagi, vagi Gao ni Duke of York i tai, pai, ruai, uai. Short Comparative Grammar. i8r For details reference must be made to the Grammars of the particular lang'uag-es ; the mere possibility of presenting- a comparative view of the Suffixes shows plainly the general use of them to be characteristic of Melanesia ^. Suffixes of this character are looked for in vain in Malay ; that is to say, the causative suffixes i and kan cannot be sup- posed identical with the Melanesian terminations shown above. Bitt in Javanese there is the Suffix ahe, and another /', which are added to the verb ^vith n or h ((peKKvaTuov, making" ^-ake, or tii, which correspond to the Fiji /caka, ]Mota ffa^, and Fiji na, ]\Iota n, Florida ni. Thus ' Javanese ngomhenl to give to drink (to a person), nffombekkake to give to drink (milk) ^.' If there be no other Suffix to Verbs in use in Javanese, the presence of these two suffices to encourage the student of Melanesian languages with the sight of a kindred form. The Suffix ake is not in separate use in Javanese as a Preposition. In Malagasji no Suffix of this kind appears. In Maori also there is no suffix to the Verb. But in Samoan the termina- tion ia^i or sal adds the sense of ' with ' to the Verb ; viomoe to run, mo'etaH to run with a thing, a'au to swim, 'ansa''i to swim with a thing. This is evidently the same suffix with the Fiji caka, taka, and in other Melanesian languages tag, sag, tagi, sagi. Besides it is said that ' the suffix ai (inter- posing a consonant when euphony requires) makes the mean- ing emphatic^.' These Suffixes again, with a consonant indiffiirently taken up, are no doubt the same as the Melane- sian. If it be true that they merely give emphasis, they must have lost the significance properly belonging to them, and witnessed by fa I and sa' i which work as with INIelane- sian Verbs. If such Suffixes are found in other Polynesian languages they certainly do not play the important part they do in Melanesia. In Tongan, which is nearest to Fiji, these Suffixes are not apparent. ' In Motu of New Guinea rano is water, rauosa to bale out water. ^ I am indebted for this illustration to Dr. Rost. ^ Pratt's Samoan Grammar. Verbs with the Ecciprocal Prefix have a'/, a^if'mai, nai as well, and consonantal suffixes^, ni, si. i82 Melanesian Lano^uai^es. ■2>'""*^>' In the Marshall Islands language it is plain that the Suffix hake is present ; wia is to buy, %via kake to sell, that is, to make a deal of something. However little Suffixes in these forms may be in use in the Polynesian languages, the terminations of Passive Verbs and Verbal Nouns in those languages resemble them in one particular so much that something may be learnt from them. The Passive Verb is made in Maori by adding to the Active the termination a or ia, either alone or with the consonants h, k, m, n, ng, r, t, that is, almost any consonant ; and these indifferently as regards signification. The Passive is made by any form of the Suffix ; all have equal signification, but custom confines the Verb to its own Passive termination. The same thing happens in the case of the Verbal Nouns ; the Suffix is nga or anga with the consonants h, k, m, r, t, icpeXKvaTLKoi', the Suffix with any one consonant having the same meaning as with every other. It may possibly be that the Maori Passive has arisen from the impersonal use of Verbs with the transitive termination ki, ki, mi, &c. ; as the nearest approach to a Passive Verb in the Melanesian lang'uages is an impersonal Active one, it may be that the forms of Verbal Nouns and of Transitive Verbs have the same origin : but the indiscriminate use of most of the consonants in Polynesian Passives and Verbal Nouns, where all must have one signifi- cation, where kia cannot mean anything different from mia, or hanga from tanga, supplies a ground for arguing that the Suffixes of Melanesian Verbs are equally destitute of meanings of their own. It points to these Suffixes not being originally independent words, Prepositions or others, come down to the position of Suffixes, but terminations, by which the language has contrived to make the Verb express itself in a way that was desired. Why should not a living lang'uage contrive terminations to supply its needs ? 24. Frefixes to Verhs. These Prefixes are not entirely and exclusively Verbal, they are applied to other Parts of Speech. Yet they show their force Short Comparative Grammar. J83 best when applied to Verbs, and it must be remembered that words used as Verbs can never be taken as nothin<^ else but Verbs, whether with or without a Prefix. It will be observed that those Particles which precede Verbs as belong- ing" to them strictl}^ as Verbs, and which are capable of marking Tense and ]\Iood, the Verbal Particles, are not included among these Prefixes, and are indeed written separately from the Verb in order to avoid being confused wath these. If such Verbal Particles were written in one with the Verbs, these Prefixes would have the appearance of Infixes. The Prefixes applied to Verbs come under four principal heads ; those of Causation, Reciprocity, Condition, and Spontaneity. The first is when a Verb comes to signify the making to do or be ; the second when a double action, one upon another, or of many on one another is indicated ; the third when a thing is shewn to be in or to have arrived at a certain condition ; the fourth when that condition has come about of itself. The two latter might well have formed one class, but that the last is somewhat remarkable. Table of Prefixes. Loyalty Islands, Causaiive. Reciprocal. Condition. Spontaneity. Nengone a e .. Lifu a New Hebrides. Anaiteum ua Fate baka ma, ta Sesake va, vaka ma, da Ambryin ma • • Espiritu Santo va, vaga ua • ' Araga va vei ma, ta .. Lepers' Island vaga vui ma tama Maewo vaga ma, mo Banks' Islands. tava. Merlav va var ma, ta Gaua va ver ma, ta tava Lakona va va' ma, ta tav 1 84 Melanesia n L anguages. Banks' Islands {continued). Causative. Reciprocal. Condition. Spontaneity. Vanua Lava va, ve ver me, ta, 'a tav, 'av, tamo Mota va, vaga var ma, ta tava Motlav va ver m-, t- tav Volow va vear m-, t- TJreparapara V- ver m-, t- tava Torres Islands va ver ma, ta teraor Fiji vaka vei ka, ta, ra Santa Cruz va .. .. Nifilole vva ., ., Solomon Islands. ITlawa haa Lai ma Wa«o haa hei ma .. Fagani faga fai ma ava Saa haa he ma taka Vaturajia va vei ta tapa Florida va vei ta tapa Savo au .. Bugotu va vei ta .. Gao fa .. .. Duke of York ^ wa we ma, ta 1. It will be convenient to take each Class of Prefixes separately. The Causative is almost universally va, alone or with a second syllable, Jca,ga. The Loyalty Islands have no v, and use a for the causative. Duke of York, having- no v, use wa. The Anaiteum na is equivalent to wa. The Savo au alone seems distinct. This Causative Prefix is plainly the toliaka, faka, aka,fa'a, &c., of the Polynesian lang-uag-es, in which the simpler form va does not seem to occur. In the Melanesian languag-es, except in Fiji and Fate, where it makes a kind of Adverb, this Prefix is purely causative, for in this sense it must be taken when it makes the multiplicative of Numerals. This Prefix seems unknown in Malaij. In MaJagasij the chano-es of letters /", p, m, make it difficult to ascertain its presence. It is perfectly plain, however, ihvii faJiatelo is the same word with Fiji vaka/obi, INIota var/afol, thoug-h in the Melanesian languag-es it is 'three times,' not 'third.' In Batak of Sumatra the same word, comjioundcd of the Prefix and Numeral, appears paltaiolu. There can be little doubt ' ]\Iotu of New Guinea sialm hot, vasia/iu 'hot water,' no doubt meaning heated, the causative va. Short Comparative Grammar. 1S5 but that the Malagasy Prefix maha is the same, which is said to form Potential Verbs, maJiarcv/, 'pouvoir vaincre.' What is called also by the same Grammarian ^ the Causative Inter- calary appears to be the causative particle /a ; mandeha to go, mavipandeha to cause to go, manao to do, mampanao to make to do ; viiditra to enter, manqriditra to cause to enter ; miboaka to go out, mampihoaha to cause to go out. To call the Particle intercalary misleads, for the Verb is nao^ delta, dira, hoaha, as is shown by the change of the Verbal Particle from ma, or mi, to ha, na, hi, ni with the change of Tense. The Malagasy Verb with the causative Prefix, like the Melanesian, takes the Verbal Particle before the Prefix. In Araga, where the ^^erbal Particle is ma and the Causative Prefix va, an example shows a complete likeness to the IMalagasy ; rahu to live, rarahu make to live, ma varalin makes to live. To write the Verbal Particle separate from the Verb prevents the misconception conveyed in the word ' intercalary.' In the Malagasy words above, 11 in mandeha belongs to d, not to ma, and the causative Prefix appears as mpa, mpi, for pa, pi, fa, fi, in accordance with the use of the language ^. The form va, fa, pa undoubtedly appears to be the original particle, to which kn, (ja, ha has been attached. This may perhaps be the Verbal Particle ka, ga, which is used in several languages. 2. The Reciprocal Prefixes of the Melanesian languages here given may be seen to be two, represented by vei and var ; the latter, with no material variation, in the Banks' Islands only, the former as vei, vni, hei, hai, fai, we and e, extending from Duke of York to the Loyalty Islands. It is plain then that vei is the more characteristic Prefix. The * Marre de ^larin. ' All words and even phrases are capable of assuming maha to cause to be.' — Baker's Malagasy Grammar. ^ It is presumptuous to offer a view of a ^lalagasy Verb which perhaps is not given in any one JNIalagasy Grammar. But writers of Malagasy Gram- mars are by no means agreed among themselves ; and it is certain that the true account of Malagasy formations will not be got without going outside the language, and comparing many others of the same family, among them the Melanesian, 1 86 Melanesian Languages. meaning is altog-ether one, and simply that of reciprocity, the action of one upon another, of two or many persons or things in relation to one another. In Fiji the Prefix applies to Nouns as well as to Verbs, and wdth Verbs is used when reciprocity is not altogether in view. This use in Fiji is useful as showing what is the notion that lies at the bottom, and rules every application of the Prefix. This notion is evidently that of relation of one to another. In this sense, as has been shown, it comes to be a Prefix of Plurality, veivale houses, not scattered singly, but standing grouped in relation to one another. The use of the Prefix in ve'i keve to nurse, to carry in the arms^ is thus intellig'ible though there is no reciprocity : and vei moku, literally meaning to strike one another, reasonably comes to mean to fight, as vei tofogoni, to spear one another, is to fight in Florida. The form in use in Lepers' Island departs rather widely from the type, but appears to be the same. The Prefix vnr of the Banks' Islands is interesting on account of its likeness to, if one may not say identity with, the Malay prefix har. To fight in Mota is varvus, beating one another, and is in Malay har-kalaJd, But bar in Malay is not a Prefix of reciprocity ; it is described as the mark of a Verb which expresses a state or condition ; a state of cor- relation perhaps with something. It is easy to comprehend how the general sense of mutual relation belongjing to vei in Fiji is particularised to plurality on one side, and reciprocity on another. It might well be that in some language vei should be found only as a plural sign, as there are many in which it is only a mark of reciprocity. We have in Fiji the explanation of both uses. So if in Malay we have bar a Prefix to Verbs expressing state and condition, and var in the Banks' Islands expressing reciprocity, we may well take the words to be the same, and suppose an original meaning, which on the one side has passed from a sense of relation of one to another into general correlation, and on the other side has been particularised to reciprocity. Malay words like harkalaiti to fight, baresiri to be married, barlamu to meet, seem Short Comparative Gravimar. 187 to lie half way: hartainu $a orang in Malay is 'to meet a man,' varnina tanun in Mota. In Malagasy there is a Prefix voa or vol which is called Passive. It may be that this is the Lepers' Island via, as ]\Talay har is Mota var. But there is in Malagasy what M. Marre de Marin calls the Reciprocal intercalary^ which, if treated as we venture to treat the causative Prefix, seems to answer to the Prefix vei. If after the Verbal Particle this Par- ticle of Reciprocity is added, of which f is the characteristic letter, we have a form of Verb which exactly corresponds to a Melanesian Verb with the Prefix vei. Thus mankatia to love, m'>fankatia love one another, corresponds in form to an Araga Verb with the Verbal Particle ma ; ma tape is ' loves,' ma vei tape ' love one another,' the parts correspond. If the causative Prefix is added also before the Reciprocal, we have three Particles before the Verb, the Verbal, the Causative, the Reciprocal, as in the INIalagasy mamjoifankatahotra make to fear one another ; a word to which the Araga Verb above adduced part for part corresponds, ma veivarahu, except that in the Melanesian word the order is Verbal Particle, Reci- procal Prefix, Causative Prefix, Verb ; the Malagasy ' do make mutually to fear,' the Melanesian ' do mutually make to live.' Among the Polynesian languages a Prefix of Reciprocity does not appear in Maori ; but it does in Samoan in the form of/i?, the same of course as vei. This particle also serves to make a plm-al ; not, as in Fiji, of Nouns, but of some Verbs, 3. The Prefixes of Condition ma, ta, are again almost imi- versal in the Melanesian languages. In Fiji ma is not counted one of these Prefixes, though the Dictionary shows many Adjectives with this beginning : ka and ra are not found in other Melanesian languages. There is no difference, however, in meaning, except that ia in most of the languages, more than ma, signifies that a thing has come into the condition the word describes, of itself, and not by some known cause from without. In Fiji ta, ka, ra are called Passive i88 Melanesian Languages. Prefixes, but clearly improperly if ' they imply that the thing" has become so of itself.' These Prefixes are not only applied to Verbs, and the word compounded with them would be ordinarily translated in Eng-lish by an Adjective or a Participle. The word to which ma is prefixed may not now perhaps be used in the lang-uag-e in which the compound occurs, and thus many Adjectives and Participial forms beginning- with ma cannot be resolved into their component parts, about which nevertheless there can 1)6 little doubt but that they are words with this Prefix of Condition. Examples : — in Fiji dola to open, tadola open ; voro to break, kavoro broken ; gvtu to cut off^', ragntu cut off". In Mota, ^vora asunder, mawora broken, Motlav moivor ; 2^^P^ (the same word with the Malay papau a board, Maori papa, but not used as a substantive in Mota), tajjfaimjm slab-shaped, with reduplication. In Lepers' Island hare to tear, ma/iare torn ; Arag-a mahera torn, dawaga come open, broken. In the Solomon Islands, Wa«o mahari torn; Saa oi to break, maoi broken ; Florida h'llu to pull out, tab'ilu come out of itself as a plug. Duke of York pala to unloose, fajiala get adrift, as a canoe. These Prefixes do not appear in Mala?/, but in the Malay Archipelago the Vocabularies of Mr. Wallace show that they are present in Adjectives. Thus jahat ' bad ' in INIalay is rahat in Matabcllo and Baju, hat, sat being Melanesian forms. In the words for ' cold ' several begin with via, some with da. The Malay panas hot, is mofanat in Celebes. The Banks' Islands sawsaiv is Celebes dasaho hot. In Malagasy ' many roots form an Adjective of the quality by prefixing ma ; loto dirt, iiialoto dirty ^.' In the Polynesian languages ma is present. The Jllaori Grammars do not acknowledge it, but it is conspicuous in the Dictionary ; //ora to spread out, Mota tvora, ma/wra, an adjective or participle, spread out, Mota matvora ; //ore to peel, maliore peeled. The Malay panas is acknowledged to be the ]\Iaori viahana. In * Baker's Malag-asy Grammar. Short Comparative Grammar. 189 Samoa)i it is said that ^ma prefixed to an active Verb makes it neuter ; as sasa'a to spill, masa'a spilt, liligi to pour, maligi spilt. The Dictionary shows many Adjectives evidently made in the same way. 4. The Prefix which signifies spontaneous condition — the state into which a thino" has come of itself — is probably a compound one, for we have seen that fa has in some languages something of that meaning. An example from INTota will explain it : to untie a ro})0 is to nl it, but a rope that has not been untied by anybody, has come untied by itself, me tavanl. The same is the case when the Prefix is not applied to a Verb : raka in Mota is ' up,' tavaraka is to get up, not to be raised, to get up of oneself. Thus also the Florida ti(gurii, to stand, becomes tapatugura to stand up. This prefix, contain- ing probably ta^ would hardly deserve notice, w^ere it not that it occurs with remarkable similarity of form and signification in Malagasy. There the difference between the Prefix voa and iafa is said to be that between a transitive and an in- transitive Verb : voa leniika izy it is sunk, i.e. by some one ; tafa leniika izy it is sunk, i.e. of its own accord^. The re- semblance between this Malagasy tafa and the Banks' Islands tava is so complete in form and signification, and this in a fine point of meaning, that, considering the space of Ocean that separates the languages, it is a matter of astonishment that it should exist. It is impossible that it should be ac- cidental ; it could not be introduced by IVIalays or Polynesians who have it not ; it must have sur\'ived no one can tell what vicissitudes and changes, in a course of years which no one can number, and presents itself, like a rare species of j)lant or flower in isolated and widely separated localities, a living and certain proof of common origin and kindred. 25. Reduplication of Verbs. It is possible to reduplicate either by repeating the whole or part of a word : and it is obvious that the way in which * Antananarivo Annual, 1876. igo Melanesian Languages. a part of a word can be reduplicated must vary according- to the syllabic character of the word. Languages which close a syllable with a consonant can repeat a syllable in a way impossible to languages which end every syllable with a vowel. Hence the Melanesian languages with open syllables reduplicate either the first syllable or syllables without change, or, if a change is made, take at any rate the whole syllable. Languages which have close syllables take for reduplication either the first syllable or syllables, or take with that a consonant belonging to a further syllable. Thus the Florida sopou to sit, can be reduplicated soposopou or sosoj)07i; while the Mota puie to sit, can be reduplicated ;;?/i!e- piite, impute^ or indpnte, in the last form the consonant of the last syllable being borrowed and reduplicated with the first. Nor is this the case with words when the consonant may seem to belong to the root of the word, as pmt might be thought the root of pute ; but tira is to stand, neuter, vaiira to stand, active, ra being the causative, which is reduplicated vatvatira. Changes in the form of a reduplicated syllable made in Melanesian languages are two. (i) In Florida, Bugotu, and thereabouts, when two syllables are taken for reduplication the consonant between them is generally dropped ; thus varono, to hear, is Tedu-plicaied vaovarono not varovarono, lahu to promise, lanhalin. This makes no difference in sense. (2) In Santa Cruz and Sesake the first consonant of the reduplicated word often changes into another akin to it, tahidabu to fight, Santa Cruz ; ganihcni to eat, qosiwosi to work, gnvahtva to fly, Sesake. In Lepers' Island not the consonant but the vowel changes ; galegcle reduplication of gale to lie. Reduplication in Malagasy also sometimes alters the first consonant of the root, mivadibadika, mizavajavatra^ vrifaopaoka. This is of course what we have ourselves in g'ood English in words like hurlyburly. It should be observed, as concerns for7n of reduplication, that though Prefixes, causative and other, are reduplicated witli the Verb, the Verbal Particles never are. This is the Short Comparative Grammar. 191 case also in Malagas}^, as in the examples jnst jGfiven ; and where, as in that lang-uag-e, it is customary to write the Particles in one \nth the Verb, it is a useful observation to make. With regard to the meaning of reduplication in Verbs, it has been mentioned that in Fiji and Samoan it is used with a sense of pliu-ality, and so makes what is, improperly, called a Plural Verb. Commonly, however, reduplication signifies repetition, or continuance, or emphasizes the meaning of the Verb. Reduplication of the whole word, or two syllables of it, rather conveys the idea of repetition ; reduplication of the first syllable gives rather the sense of prolongation of the act : and this may be done at pleasure by repeating over and over again the first syllable, pjiipipiva go on speaking, Santa Cruz, jmjji/jjupufe go on sitting, Mota, or by prolonged pronuncia- tion without repetition, as in Nengone. Reduplication with a close syllable rather intensifies the meaning of the word. As an example of each form the ^lotajnite, to sit, will suffice, j)2ifepufe to sit from time to time, pupnte keep on sitting-, putpute sit down closely. 26. Passive Verhs, In none of the ISIelanesian languages here compared, with the doubtful exception of Fiji ^, is there any Passive form of the Verb. It by no means follows because a Passive Verb in English is translated in a certain way in a Melanesian lan- guage that the Melanesian form is that of a Passive Verb ; nor because a Melanesian form is best translated by an English Passive that it is a Passive form. For this reason the prefixes of condition ma, fa, ka, may be at once dismissed as having no claim to make a Passive Verb. It may be said that what nearest approaches a Passive Verb is an Active Verb used impersonally. To build a house in Mota is we iaur z'ma, to say ' the house is built ' * ]\Tr. Fison says, ' I doubt whether there be a true Passive. No Fijian would use Hazlewood's example.' 192 Melanesian Languages, the expression is ivaa me taiir vefa : the Verb undergoes no change, yet the sense undoubtedly is Passive, that the house has been built. It cannot be denied that ima is the subject of the Verb tam\ if grammatical construction should be pressed ; but me taur veta iraa may equally be said, in which ivaa wovild appear to be the object of the active Verb tanr. The truth appears to be that strict construction, according to our Grammar, is not to be sought ; the Verb is impersonal, has no subject or object, and the Verb and Substantive simply combine to show the house and the building of it, and to make a statement. From such a way of conveying the notion which would be couched in a Passive sentence where Passive Verbs exist, may have arisen the Passive Verbs of the Poly- nesian languages. In Florida to express the Passive they put the active Verbs into the third person Plural, as we say ' they are building a house,' without reference to any particular persons. For ' the house is built ' they say, tara pltiia taa ua vale they have built the house. M. Marre de Marin maintains that the Malagasy Verb with affix, in its radical state, indicates a Passive, and that the various prefixes make the Verb active, neuter, causative, or reciprocal ^. The truth probably is that in these languages the Verb is originally the name of an action without any regard to the agent or the patient, and is neither Active nor Passive, until, in the advance and cultivation of speech, affixes come into use to give a positively active or passive form. ' ' On ne saurait trop insister sur ce fait si curieux et qui est I'line des assises fondamentales des gramniaires malgache, lualayse et javanaise.' IV. PHONOLOGY OF THE MELANESIAN LANGUAGES. I. Alphabet. The Melanesian languages have of course been written and printed in the Roman Alphabet. As regards the Vowels there has been little room for diversity of practice, no attempt having been made to use them in the English way ^. As regards the Consonants there is a good deal of diversity, because four missionary bodies have been engaged in reducing the native languages into print without any concert or agree- ment ; the Wesleyans in Fiji, the London Mission in the Loyalty Islands, the Presbyterians in the Southern New Hebrides, the INIelanesian Mission in the Northern New Hebrides, Banks' Islands, Santa Cruz, and Solomon Islands. There are many Consonants about which there is little room for difference ; a dental tenuis will be written t, a guttural tenuis k ; m, n, r, 1, s, h, represent sounds about which there can be in a general way very little question. But this is only in a general way. One will use t where another will use d. It is difficult to determine sometimes whether a sound is what in English would be k or g ; natives themselves are not certain about r and 1 : it is a question whether the sound made in some locality is really an aspirate which may be written h, or ought not rather to be represented by f. There is much dif- ficulty in settling the orthography of any one language or dialect ; and if it be settled in one the question arises whether the letter printed should vary with the change of sound be- longing to neighbouring dialects. Suppose, for instance, that ^ In Fate, however, oit is wrongly used for au. O 194 Melanesian Languages. in some island the people of two or three villages use ng-g- or ng"k where all the rest of the population use k, is it desirable to mark their nasalization in letters, or is it enoug-h to use k in printing- and let them give it a nasal sound if thej please ? If the language is printed for the benefit of foreign scholars the system of orthography should no doubt be strictly phonetic, each symbol representing one distinct sound ; but if the printing is for the use of natives, it is better to have one typical symbol, and then dialectical varieties of sounds will be represented by a single character the value of which will vary in each dialect. Again, when the Consonant as sounded by the natives differs from the same Consonant when sounded in English, is it necessary or desirable to mark the difference by diacritical points or some such contrivance ? In no Melanesian language is the dental tenuis the English t, yet it is the hard dental check of the natives ; it is t, though not our t. There must be taken into consideration the fact that generally what is printed in one of these languages is printed for the use of natives, and very often has to be done with only the suppKes of an ordinary fount of tj'pe. It is moreover very desirable to make reading and writing as easy as possible to the natives for whose benefit the art is introduced. To take the case of Fiji ; the natives cannot close a syllable with a consonant, and they cannot say d without the sound of n preceding. A word sounds enda^ but if it be so written the native scholar will naturally insert a vowel between n and d and turn the word into enada ; a word which sounds wa^igha would puzzle them altogether with its three consonants if so written. In Fiji as it is printed the first word is eda, the second waqa ; every Fijian child who learns his letters learns d to represent nd, and q to be either ngg or ngk, calls them nda and ngga. The problem, then, is a difficult and complicated one. If a language be written scientifically for Europeans it may be done accurately but laboriously, and will bo most inconvenient to the natives. If the language be written as simply as possible for the convenience of natives with the fount of type Phonology. 195 made tlie most of, the natives will read it right, but the European will be puzzled. The old king of Fiji was Cahobau, which the native will sound Dhahomhaii (au=:ow in cow), while the trader or planter will read it Kakobaw. The g-eneral solution is that the alphabet must be used to suit the native in the first place, and that the European must learn the value of the alphabet of a Melanesian language as he does in any other foreign tongue ; but that at the same time the letters should be used in the native alphabet scientifically and not arbitrarily. Bishop Patteson, who first reduced to writing the languages of many Melanesian islands, followed the advice of Professor Max Miiller in his OutKne Dictionary for the use of INIissionaries : he used no letters arbitrarily, but the Roman letter represented a sound in the native language the same in general character with that represented in English, and an italic letter was employed to show a variation in the sound. For example, in some Melanesian languages which have no hard g, there is a consonantal sound which is peculiar and cannot be represented by any letter with the power it has in English : this consonant is guttural and is represented by g \ Every native w^ho learns to read starts uith the use of the sound and associates the letter with it ; every European has to learn the sound and to apply it to the letter. It is true that in this there is danger. The European starts with the association of the English sound and the English letter, and will naturally give the native g the sound it has in Eng-lish. In the case of the sound of ng in the word ' finger,' it is in Melanesian languages a form of the guttural and is therefore represented by the italic g, not arbitrarily as in Fiji by q. In Fiji, where there is no hard g, except in a few words, they use g for ng in ' singer,' as they do also in the Southern New Hebrides. Bishop Patteson, using g for the peculiar guttural, which is not in Fiji, intro- duced the italic 11 for the ng in ' singer ' into the languages w'hich he wrote. In printing the words belonging to the Melanesian lan- ' The same sound is in the Loj'alty Islands arbitrarily represented by x. O % 196 Melanesiaii Languages . guag-es, not for native use but for European students, it is pos- sible either to use a scientific and accurate method of spelling- applied to all the languages alike, or to give the words as they are actually spelt in the method already adopted in the languages to which they belong*. The latter plan is followed here, with such occasional explanations as seem necessary, and a table giving the value of the letters in use has been pre- fixed. This, it is true, is neither scientific nor accurate, but it is almost unavoidable ; there are different systems already at work which seem to have a right to the words of the lan- guages to which they have been adapted. References to books in which some languag-es are already printed would be much more difficult if the words to be referred to were to be found there in a shape other than that given here. It is desirable here to give a brief view of the powers of the letters used in printing languages of Melanesia by the Wesleyan, Presbyterian, and London Missionaries respec- tively, in Fiji, the Southern New Hebrides, and the Loyalty Islands. I. The peculiar use of letters in printing Fiji is confined to b, d, g, c, q : b is always sounded as mb, and d as nd ; g is ng in ' sing' ; c is th in ' that' ; q is ng in 'finger.' 3. In the Southern New Hebrides, in Anaiteum, c is used for hard g-, d for th in 'broth,' g for ng^ in 'singer,' and j for te in ' righteous \' In Fate g* apparently stands for both the sounds of ng in English. 3. In the Loyalty Islands, in Nengone, or Mare as it is now called, g is hard g, ng is as in 'singer,' c is the English ch, 'm is a nasal m, x is the peculiar g-uttural common to most Melanesian languages, represented in the Melanesian Mission by g, but not existing apparently in Fiji, and not represented in printing the Southern New Hebrides languag'es. 4. To come now to the lang-uages with which the ]\Iolane- sian Mission has to do. The general principle being that the Roman letters represent the same sort of sound that they do * So in Mr. Inglis' Grammar, where h at the end of a syllable is said to be 'like X ill Oreek, or gh or di in Scotch.' Phonology. 1 9 7 in English, and italics variations of those sounds ' ; a good deal of uniformity has been ohtainal^le in printing the many languages that have to be reduced to writing. Thus, t stands for the hard dental, though it is nowhere the English t, and where a peculiar sound of t occurs t italic is used ; the nasal m is m, n is ng in ' sing.' But uniformity has imfortunately not been attained ; the alphabet belonging to each language has to be learned. For example, in Ysabel j represents either the English j or nj, but is used in printing Ureparapara or Santa Cruz for tch. The reason is that in the one case it was naturally used to represent the English or nearly English sound, and in the other, not being wanted as the English j, it was used for a sound which to the natives represents the English j. If the natives knew nothing of English spelling it would be an easier matter ; but when they call the English j che or tche it is better to meet them half way and let them spell chichi J iji. Another great obstacle to the carrjdng out of an uniform orthography has been that the knowledge of the variety of sounds requiring to be expressed has been obtained partially from time to time. When b has been settled as equivalent to mb a dialect appears in which m does not go with b ; it is un- avoidable, therefore, that b should have a different value in those two places : when j has been settled as representing tch in one or two languages another comes into view which has the sound of j and also of ch. It is practically impossible, therefore, to attempt a complete uniformity ; but a general uniformity has been attained, and the natives, for whom in fact the languages are printed, learn their own alphabet. The Alphabet as used in the Melanesian Mission. Vowels — a, long and short ; a short, and sharp, e, „ „ e, French e. ^ Italics in icrUing are marked with two dots above, not with a stroke below, because for dotting i and crossing f the pen goes over above the words. In printing words in italics the letter which is ordinary type would be italic, becomes Eoman, siw, sm, sing. 198 Melanesian Languages. i, long and short. 0, „ „ 0, German o. u, „ „ ?<, German ii. DilMhongs — ai, ae, ao, an. Consonanis — b, generally mb, in some places b. d, generally nd, in some places d ^ ; f/ see t. f, as in English. g, generally a guttural trill, in some places hard g ^ ; ^ is ngg, as ng in ' finger.' h^, as in English, it closes a syllable, j^, nj, or, as in English in Solomon Islands; tch in Santa Cruz, Torres Islands, and Urepara- para. k 1, more trilled than in English. m m^, nasal, n «, ng in ' singer.' gn for n ^. p, nearly the English. q, a compound of k p w, in which sometimes p is obscure, and sometimes k hardly heard'', r, trilled. * In Bugotu the difference of sound in d belongs to the village or the family ; in Araga it seems individual. At Saa it is sometimes dj. ^ It is difficult to determine at Saa whether the sound is hard g or k, k is written. In Wajio it is the hard g, and as there is no k distinct from hard g it might be well to use k for that sound. But at Fagani, close by, tlie peculiar INIelanesian g reappears, which is represented by a gap at "Wano ; there is no hard g, but k is sounded. To make the difference therefore between the hard g of AVawo and k of Fagani both letters are used, ^ At Lakona h approaches f, at Fagani, Ha'ani, it becomes f. * In Bugotu j follows d ; those who say nd sound nj : in some words some individuals at least sound j as tj. In Santa Cruz and other places the sound is much the same, but tch rather than tj. '■' Written in the Loyalty Islands 'm, in Southern New Hebrides mw. * In writing what in printing is the italic n two dots are put over n ; it is not easy therefore to use ii for the sound usually so represented, for fear of confusion, and the native g lends itself well to the combination gn, sounding as in French or Italian. '' The liii.s are closed upon the formation of the guttural and opened some- what suddenly to emit the breath. Tlie sound varies towards kw and pw, according as tlie guttural or labial is more fully formed. PJwnology. 199 s t, never the same as English, the tongue broader and not so far forward ; t in Lakona and Torres Islands ; the cheek to the breath is incom- plete ^ V, more labial than English v ^. w, closes a syllable ". z, as in English. II. Phonetic Changes, When in cognate languages, like the Melanesian, what is evidently the same word is found in two forms, the one form may often be seen to be owing to a phonetic change ; one may be pronounced the older form, the other more modern ; one may be shown, by comparison between many languages, to be a normal word, the other a modification of it. For example, the very common word for a canoe is in Maori waha, in Fiji loaq^a (wangka), and no doubt waha is the normal form ; k has changed to ngk by nasalization. But again the same word will appear in many languages in various forms, and no one can determine which form is the original, no order of change can be asserted. The same word appears as waka, vaka, Iiaka, and it is impossible to say that w has changed into V, or v into w, though h may be thought a change. These languages have no history that can be traced externally. If two forms of a word are found they are, if simple, parallel, not one original and the other derived ; kiko in Florida and tito in Maori are two pronunciations of the same word, and one has as much right to be called original as the other. The Greek TreVre {TT€fjLT:e) and Latin quinque are two forms of the same word, ^ The sound of t is not so blunt as in Maori, and varies slightly in different languages : that of t is something like tr, the breath passing over the tip of the tongue and vibrating ; the medial d is pronounced in the same way. ^ AVhen the Banks' Islands were first visited the names now written Vanua lava, Meralava, were spelt Vanua laba, JMeralaba. The native v is not the English. It sometimes approaches w. ^ The sounds of the three words in ]\Iota gau a fishhook, gao to spread like fire or news, and gaw to take in a handful, are quite distinct. 200 Melaiiesian LanztiCL^cs. brothers, not father and son, just as the Fiji lima and Maori r'lma. But all languages have a histor}"", however lately they may have come into view, and something of their history can be traced internally — some words show a phonetic change, some decay. In modern European languages there is historical evidence by which the old form and the new are certainly known ; in Melanesian languages there is nothing of the sort, but yet there is some certainty to be obtained that one form is older than another. That cinq is a modern form of qninqiie rests on outward evidence ; but it is just as certain from internal evidence that the Marquesas ima five, is later than the Fiji Ihna and Maori rhia ; and the Fiji liga and Maori ringa, hand, may just as surely be said to be later than lima, rima, five. It is important also to consider the question of the indis- tinctness and uncertainty of sounds, whether this means that distinct articulation of separate sounds has not been yet attained, or whether it is that people now pronounce sounds indistinctly which formerly were separate in their language. In the language of the Sandwich Islands there was so much indistinctness between t and k that one set of Missionaries used t and another k. The spelling is now settled to k, but the pronunciation is not settled to correspond ^. In San Cristoval in the Solomon Islands, at least at Wa?io, it is suf- ficiently ascertained that they use r not 1, yet a native who can read and write, and will tell you that they say r not 1, will pronounce some familiar word with 1, not r, and be per- fectly unconscious of it. When a native of Tikopia speaking a dilapidated Polynesian language, with a quid of betel leaf and areca nut in his mouth and his lips stiff with lime, was before him. Bishop Patteson himself could not ascertain the sounds he made. But with the organs of speech unimpaired, either ' This is Raid on the strength of a single example. A Sandwich Islander living in Norfolk Island pronounces the printed k plainly sometimes as t, mailcai he reads ma Hal; in other words he reads k with the guttural tenuis clearly pronounced : in many words it is difficult, in some it is impossible, to distinguish whether it is t or k, the sound is so obscure. The man himself believes the sounds to be all the same; one letter k is used, and he cannot perceive that his pronunciation varies. Phonology. 201 through carelessness or imperfectly exercised faculties, a gut- tural sound will sometimes be piade, not quite in the throat, and a dental a good way from the teeth, and what is produced is neither distinct k nor t. Or else from tlic same causes it is sometimes one and sometimes the other. The question is whether this double indistinctness and uncertainty are a primi- tive condition of articulation not yet settled into distinctions, or a degradation of articulation which has lost exactness. Melancsian examples go to sup})ort the latter view ; unless it be held that to pronounce a ^\■ord with a gap in it, where a consonant is sounded in a kindred tongue, is a more archaic practice than to pronounce the word with the consonant dis- tinct. In one region of the Solomon Islands, in Ulawa for example, it is the practice to say 'olu instead of the common numeral tola three, la for Iha fish, and words full of vowels are common ; and it is there that the learner is most puzzled with indistinct and uncertain consonants^. This indistinctness or uncertainty is plainly a ditferent thing from phonetic change. In the changes which do occur it is generally impossible to find a law of change. The two languages of Florida and Vatura^^a in Guadalcanar are so much alike as to be dialects of the same ; and between them there seems to be a certain law of change in the letters g^, h, s. The Florida g- (the Melanesian g), is always h in Vatura;^a, in words common to the two lang'uages, the Pronouns h'lla, liami are the glfa,gam'i of Florida and other tongues. No g therefore remains in Yatu- ra?ia. The Florida h, into which g has changed, becomes in Yaturawa s, sanavulii for hanavnin ten, e nisa for c ni/ia how mam\ Thus, by metathesis also, Florida ffe/ie is Vaturawa se/ie. But beyond this no rule can be made. Some words show Florida s turned to ch, written i,j'ji for sisl, some to t, ^ Some natives of Ulawa have been educated in Norfolk Island and read and write the Mota language ; but in writing Mota they use inditferently k and g, w and v, v and p. The variation of consonants in Santa Cruz, 1 and n, — p, V, b, — k and g, is not accompanied with any indistinctness in pronuncia- tion. 202 Melanesian Languages. tani for sani^ in some s remains. In no other of the Melane- sian lang-nag'es considered here can so much as this of a rule of phonetic chang-e be set down. The same words occur in dilierent languages in various forms, with equivalent sounds, but with no reg-ular law of change. In Bug-otu the Florida 1 changes to dh, hodho a pig" for holo ; but not every 1, hull, lima, vula, are the same. In Mota there is no h, which abounds in Motalava ; in many words, therefore, Mota has s where Motalava has h, sava for hav, us for ili : but there is no regular change, for Motalava very often has s where Mota has it. These are examples showing the general character of the Melanesian languages in this respect. Sounds which differ one from the other correspond one to the other in different lan- guages; and, interesting as the phonetic changes are, it is apparently impossible to show a law prevailing between one language and another. The reason for this probably is that the various languages and dialects have been brought irregu- larly into their present seats, not in successive and considerable migrations from one quarter or another, but by chance and petty movements of people whose language, though belonging to one family, was already much broken up and diversified. It is worth while to remark that some sounds do not seem to be constant in a language. In Samoa k has quite recently begun to take the place of t ; in Fiji the foreign p is coming into use and dispossessing the native v ^ : in Tahiti r and not 1 is now used ; but the old Pitcairn women in Norfolk Island, who spoke Tahitian with their mothers, cannot pronounce a word with r. Some years ago, in Wano of San Cristoval, the practice began to turn h into f, no doubt in imitation of their neighbours at Fagani, Ila'ani, but it was again discontinued. Such changes no doubt go on in languages which are un- written, and a language just brought into laew may show forms of words which are quite modern in it. But such * 'The tribes of Eastern Fiji have a p of their own co-existent with v, and do not confound them at all. Their p is the equivalent not of Bau v, but of its b.' — Rev. L. Fison. Phonology, 203 changes altjo are seen to take place in languages already printed and read ^. It will now be attempted to represent the sounds belonging to the Melanesian languages, with the changes which can be seen to be made, and the equivalents used in the various dialects. 1. Gutturals. — The tenuis k is absent in very few lan- guages. When it is absent it is represented either by ngg, {g, q) or by the hard g. In Lepers' Island in the dialect of Walurigi (j is used, while the neighbouring places have k, aga a canoe, and aka. The same is the case in Volow, Saddle Island, in the Banks' Islands ; g takes the place of k, og a canoe, where neighbour- ing Motlav has ok. g. — In Wa«o of San Cristoval the hard g takes the place of k. The lang'uages of Ulawa and Malanta, which are closely allied, have k, not hard g ; but there is not a very clear dis- tinction between the surd and sonant : there is a doubt whether to use both k and g, or k only. The hard g is very rare in Melanesia. In the Soloiuon Islands it is only heard in San Cristoval. In the Ncav Hebrides it is written c in Anaiteum. It is in the Loyalty Islands written g, but in Neng-one slightly nasalized. In Fiji the sound is rarely heard, and has no symbol. g. — In some languages where k is fully used it is very com- mon to use also^ (ng in finger), the Fiji q, as a change from k, which belongs to what no doubt is the original form. Thus the Fiji waga, the Araga ivaii;a, represent waka in languages in which there is no difficulty in using the latter form. It may be said that wherever g, or Fiji q, is found it is a change * In view of the question whether Melanesian languages, uninfluenced by the presence of Europeans, are fixed or changing in their words and gram- matical forms, there may be brought forward the case of that of Bugotu in Ysabel. In 1863 Bishop Patteson wrote down some descriptions of canoe- building, crocodiles, etc. from the mouths of natives, which in 1883 were declared by a later generation to be in accordance with their pi-esent speech, with the exception of an adverb used superlatively which has gone out of fashion. 204 Melanesian Languages. from an original k. Sometimes, but not often, the Fiji q is ng"k, not ngg", and so is g in Araga : sometimes in indi^adiial pronunciation the nasal sound is slight. That the sound is heard in Fate, though it is not represented as distinct from ng, is almost certain, since Bishop Patteson marked it in Sesake. It may be said that rj and hard g are never found in the same language, except that in Fiji some words written with k have the letter pronounced like g. But at the beginning of a word the nasal sound is often not so con- spicuous but that what is really g is taken for hard g. In Araga and in Santa Cruz k and g interchange ; they are one or the other at pleasure in the same word. n, — Another change from k is the nasal ng (in Fiji and elsewhere written g-), in the Melanesian Mission n. The change from k is shown in Lifu, Ambrym, Santa Cruz, and Duke of York in the suffixed first person Pronoun, the characteristic form of which is undoubtedly k. This in Lifu becomes ng, in Ambr3'm and Santa Cruz n, in Duke of York g = ng. Often however as the sound n occurs in Melanesian languages it is probably seldom that it represents k or an original guttural. g. — The guttural consonant thus written in the Melanesian Mission, and called hereafter the INIelanesian g, is very characteristic of the Melanesian languages, and yet is not heard, or is not recognised, in some parts of Melanesia. In Fiji it may be said that it is not heard ; in the Southern New Hebrides it is not recognised in print, though it certainly is heard ^ ; it has not been recognised in Duke of York. In the Loyalty Islands it has made itself so conspicuous as to receive a peculiar character, x. That it should exist and not be recognised is not improbable, because it may be taken for k or for r, or may be missed altogether. It has been written g (hard), r, ^, ;•, gh, rh, and k. That it resembles r is shown by the spelling of visitors ; Gaeta in Florida could never have * Bishop Patteson, vvliose aiitlidrity in questimis of sound is undoubted, and who took much pains with this sound, did not hesitate about it in Sesake. I have heard it plainly myself from a Fate native. Phonology. 205 been written Kita, or gam, to swim, in Mota ram \ if the sound had been hard g-. On tlie otiior hand, in tlie Mota printin*^, when the lang-uag-e was first committed to writing, the words takai for tagai, and ate for gate, show that the sound seemed sometimes very different from r. and sometimes was not caught. We may learn from this something- of the true sound that has to be represented, and we may understand how the sound may have failed to impress itself as one requiring- a distinct character -, The sound, no doubt, is difficult to Euroj)eans, and it is difficult to describe. It is written g-, because where it occurs there is no pure hard g, and because it is certainl}'^ guttural ; but it is never hard g in the mouth of a native, and no native who can write ever hesitates as to its use. Bishop Patteson A\as struck by its resemblance to the Arabic gliain, and Pro- fessor Max Mliller's description of the Hebrew ahi as ' a vibration of the fissura laryngea, approaching sometimes to a trill, nearly equivalent to German g in tage', closely suits it. There can be no doubt but that this sound in the Melane- sian languages represents k in kindred tongues, tis gagavu is Maori kakahu ; and that it is a step towards the break or gap which in Samoan represents k which has fallen out, symbolised by an inverted comma, and described as ' a sound something between h and k.' In fact in Melanesian lan- guages the break never represents k directly ; but indirectly through g, as in San Cristoval, it does. The loss oft makes the break, in words in which t and k have probably an equal original right, in Ulawa or Pek ; but in Wa;20, where k is re- placed by hard g, the Melanesian g Itills out and leaves a break. The common word for fish, ikn in INIaori, is i'a in ' Mota Vocabulary in Commodore Goodenougli's Journal. ^ Hebrew names written in the Greek of the LXX. and New Testament have the uin sometimes represented by T, sometimes onutted ; e.g. Gaza and Gomorrah, Amos and Eden, have equally in the original the initial ain. In the same way, as I am informed, Europeans have borrowed Arabic words beginning with this letter and made it sometimes k, carafe, sometimes g, gazelle, sometimes r, razzia. 2o6 Melanesian Lan^tiazes. i> Samoa, but g^enerally in Melanesia it is iga, which in Wa«o is i'a. Between the break, which is a sound in Samoa, and the Melanesian g, which is sonant, the difference probably is not g-reat. This sound, then, is not heard in Wa^^o, where the break represents it, or in the neig-hbouring- dialects of Ulawa and Saa ; but the natives there have no difficulty in pronouncing it. At Fagani near Wa^io and in great part of San Cristoval it is in use. At the north-west end of Guadalcanar it changes regularly to h^. In Florida, however, and Bugotu it re- appears, at the same time that w disappears. The connexion between these two semivowel sounds is shown both in the last named region, in Fiji where toa string is Mota gae, and in Mota. In ]\Iota one dialect substitutes w for g in many words, tawur for tagir, uio for ug ; in Florida g is used in pronouncing foreign words with w — woivut is j^ronounced go- gntii. The passage of k to w is perhaps through this g, as the Fiji hiviete, kuml, hune are the Mota toiimeto, wiiniif, tci(7ie. It must be added that this IMelanesian g sometimes repre- sents a more common u ; jjogi in Espiritu Santo is boni, night ; n and g constantly interchang'e in Ambrym ; ge and ne, he. In the dialect of Veverau at Mota g at the end of a word is pronounced like i, manfa'i, wurvai, for mantag, ivurvag. 2. Dentals. — It has been said that the Melanesian t is never the same exactly as the English ; it represents a blunter sound. Still there can rarely be any doubt but that the sound should be ^^Titten t. In Ulawa t comes near to d. But in Ulawa, as in Saa and Wawo, t sometimes drops and is represented by a gap or break, as in the word for head pa'u, ba'u, Mota qatii. In the word for ear ^alina, it is t that has been lost though the Wawo has garma. It is a remark- able peculiarity in the dialect of Pak, a small district of Vanua Lava, and its neighbourhood, that t is droi)ped in the same way where it is present in the common words of the Banks' Islands — q'i^igi is the Mota qatugl, Ulawa yyfl'z/. But t is ' The change of k to g and h is shown in the Vocabulary : ' Tree,' JIalay fcat/u, Fate kasii, Malagasy hazo, Florida r/ai, Vaturawa Iia!. Phonology. 207 not absent from the lan^uag^ ; it comes back partly represent- ing' n, and partly in words apparently borrowed from without. The word for a cocoanut in INIota, malir/, is me^ig \\dth t thrown out, but the Mota maniii a nose comes out as 7)ietigi, n havinof turned into t. Tn borrowed words some of the people at least support t with n ; nto for toa a fowl, t- — There is a second t which is printed t and occurs in the lang-uag-e of Lakona in Santa Maria and the Torres Islands. In this the contact of the tong-ue with the teeth is not quite complete, and a certain vibration is heard as the breath passes over the tong-ue, which has caused the sound to be represented by tr. It is a way of pronouncing in some words what is t in neighbouring districts, but more often it represents n ; taiun in Lakona, Mota tanun ; ten, nan ; tomiom of Torres Islands, Mota nomnom ; words, however, which are todiin, den, dom elsewhere. s. — The change from t to s is shown in vas, Lakona and Lepers' Island, for the common vat four, fei and taha, Wawo, for the Interrogative Pronouns sei and sava, and in many examples. r. — The change from t to r is found in Ambrym and the neigh- bouring island of Api. In Ambrym the common words mate to die, mata an eye, become mar; qeta qer. In Api the numeral vati becomes vari, tai one makes rai six. The change no doubt is due to the connexion between t, d, and r. j. — Along the west side of the New Hebrides and the Banks' Islands, and by the Torres Islands to Santa Cruz, there stretches a practice of turning t into tch, spelt j. This begins in Api, though chtia,ji(a, represents the numeral two as diia. In Ambrym Bishop Patteson wrote chene and tiene, showing that t before i changes to ch as it docs in Lakona, L^reparapara, Torres Islands, and Santa Cruz, and in some dialects of Fiji. In Espiritu Santo the sound is rather ts than tch ^ ; tajua, ^ Ti Sakalava, in Madagascar, is tsi Hova, fotsi white, Malay pufih, Lepers' Island macuti. ' Malay and Dairi have often c/a where Toba (Sumatra) has oth are pronounced. The two are equivalent and interchang-eable, but it will generally be found that, in the languages where both r and 1 are sounded, the word will have settled down in Melanesia into one form or the other. For example, the very common word for hair in the Ocean languages is in Melanesia always in a form with 1, nlii, like Malagasy rolo, not uru, like Maori huruhuru, r. — It cannot be said that r is quite uniform in sound, though trilled much more than in English. In Lakona at the end of a word it is cut off very sharply. The taking of d to strengthen r has been mentioned, and what is perhaps in some cases the consequent change of it to n. In some languages r cannot follow n without an intervening d, as in Mota nan ra cannot be pronounced except as nan dra. In Ambrym d comes between m and r, lom dro for lom ro. In some cases r disappears and leaves only d, and so rua two has become dua in Sesake, as daun in Malay is rau. In the same way t strengthens r in Fate, ratrua for ra rua they two ; and in Ambrym t and r are so far interchangeable that to and ro are forms of the same word. y. — Changes of a different kind are from r and 1 to y and dh. That to y is found in the Banks' Islands, in Saddle Island, and Ureparapara. In Saddle Island the children always sub- stitute y for r, and as they grow up use r ; but in one district, Bun, adults go on with y all their lives. In Ureparapara the use goes rather by district than by age. As in English, y with these people is both vowel and consonant ; vowel at the end of a syllable, consonant at the beginning. The Mota poroporo, Motlav borbor, is at Bun boiboi, or boyboy, if it w^re worth while to write it so : the Mota raioe becomes yaw : both consonant and vowel y appear in the name Vaiqalyaw for J arqalraio. In Fiji y is used as a consonant, but it some- times only represents i. In Ambrym y has been written. dh. — The change of r to dh occurs, or perhaps it should 2i6 Alelanesian Languages. rather be said of a reg-ion depopulated by the Labour trade, oc- curred, in a small district near Volow in Saddle Island, where, for the A^olow eresei, they said idJ/efiei, for Volow ^ge?', igedh. If this chang'e occurs nowhere else in Melanesia, it is parallel with that of dh for 1 in Bugotu and elsewhere ^. Very many words which in Florida have 1, in Bugotu have dh in its place ; ilatJie for flale child, botlio for bolo pig% vat/ie for vale house. The sound of n contained in the native d prevents the use of dh in printing the words. There is no rule, as has been said before, for the application of the change from 1 to dh ; in many common words 1 remains in both languages. It is not only with the neighbouring Florida that the Bugotu inter- changes 1 and dh ; tltepa, earth, is no doubt the distant INIota lepa. The same change is found in Fiji ; carji (dhangi) is the very common word for sk}^, rain, or wind, lau\^ rangi. 1, — The language of Cao, close to Bugotu, does not follow in the change from 1 to dh, but strengthens 1 with g^ Florida lano, fly, Bugotu tJiano, Gao glano ; lajn tongue, thajn, glapi. In Vaturawa 1 is left out, as k and t are in other languages ; the Florida tidalo, m.adola, become tida'o, madd'a. The inter- change of 1 with n at Santa Cruz has been already mentioned. 7. Sibilants. — s. These are entirely absent in Santa Cruz, in Duke of York, and in the greater part of the Torres Islands. At Santa Cruz they cannot, without practice, say s ; they substitute t for it in pronouncing foreign words : the English 'box' becomes haJcefi"^. In native words j=:tch represents sometimes the sibilant of other languages. In Duke of York w to some extent represents s, but the sibilant and aspirate are often left out in words which commonly appear with one or the other, as uri the Mota suriu^ Florida Iixli. In the greater number of languages which have both sibilants and aspirates h and s are equivalent. * In Gao _5rHoraj, yesterday, ecjuals Bugotu ignotha, r = dli : but the Florida is nola, Mota noi-a. ^ In this way the island properly called Sikopia has got the name of Tikopia, in charts Tucopia. The name has been taken from Santa Cruz. An island of the Fiji group has the same name Cikobia, c — dh, a form of s. Phonology. 1 1 7 z. — This sound is found, like the Eng-lisli, in Nengone, Save, and Cao. In Vaturawa it has a somewhat different value. In Savo it takes the place of s, and so is equivalent to h ; hlzi fing-er, the jNIota lusul ; hnzi rat, the Florida kahi ; aznazu smoke, the Florida aim, Mota asn. The sound takes the place of h in part of Florida, and equals there dh. In Cao it is eqiuvalont to the Bugotu dh, in words api)arently in which dh does not stand for 1 ; Bugotu thelie to die, Cao zehe. The sound also represents a more distant s ; the Savo kazu tree is Fate huso, Malag-asy hazo, Malay kcuju. In Vaturawa it is not easy to determine whether the sound is nj or nz ; in either case it is likely that d = nd is represented, which chang-es to the Bugotuj. But z also represents an aspirate, zare to sj)eak is hare, the Maewo ware, Maori kare. ch, j. — In Vaturawa s turns to ch ; Savo is called Chavo, the Florida slsi red is chichi, written jiji, as in Cao j in the same Y^ordijijia has the sound of j. c. — In Fiji c = dh often represents s in other lang-uag-es, which is indeed only to say that it also rej)resents h. The Fiji cake up, is the Mota sage, Florida huge ; cava is sava Mota, hava Florida, as cei is sel and hel in the same languages. 8. Aspirates. — In several Melanesian languages there is no aspirate ; in Fiji \ Fate, Mota, Santa Cruz, Duke of York. In Lakona and the Torres Islands h is rather explosive ; in Fagani it becomes f. Where it is absent, in Mota and Fiji, it is often represented by s ; but s and h are so fundamentally interchangeable in the whole family of languages that one cannot be said to take the place of the other ^. Except in the case of Duke of York, it can hardly be said that there is in Melanesia a dropping away of sibilant or aspirate without any equivalent, as when hage and sage, above, become Maori ake, and Samoan a'e. In Duke of York, as has been noted, w sometimes takes the place ; whiaga for sinaga or hinaga, a w'ord ' ' In the Nadroga, Navitilevu, dialect h is lieard and changes with s, Bau siga, Nadroga Mqa ; but not with every Bau s.' — Rev. L. Fison. ^ In South Cape, New (iuinea, sine is woman, which must be taken as a change from hine. At Teste Island the word is shine. Elsewhere sh is only heard in Nengone. 2i8 Melanesian Languages. which in Sesake is viuaga. The chang-e of k to h and Mela- nesian g shown in the words for ' Tree ' (Vocabulary No. 6^) extends throug-hout the lang-nag'es. dh. — In a considerable part of Florida the aspirate becomes dh, and in one part it becomes z, after the fashion of Save. Thus the neg-ative is fa/io in Boli, Halavo, and Ho_^o, (ad/io in Belag-a and Gaeta, fazo at Olevug-a. 9. Metathesis. — Consonants and syllables occasionally shift their places. This happens sometimes when there is no dialectical difference, as in Mota people in the same village may say either valakas or vakalas, or in Florida magora or maroga. Sometimes the people of some place will have their own form, as wesara for werasa in one Mota village ; in Fiji hahola or hokala, waqa and qawa. More commonly the same word appears regularly in two forms, in different, perhaps distant, dialects or languages ; as Florida dikl is Ysabel kidi, Florida Jiege is Bugotu gche, Mota gese, Fiji kece ; Mae wo iarisa is Mota sarita. Lepers' Island tatarise, Mota sasarita ^. Metathesis often serves to show the identity of widely distant words ; as the common Melanesian and Malay Archi- pelago word for fly, lano, rango, is shown to be the Maori ngaro by the dialectical metathesis rango. 10. Vowels. — No regular change of vowels between one language and another takes place. To take the example of the Banks' Islands, there is found on one side a preference for u, on the other for i. This belongs to a disposition either to use long and open vowels and diphthongs, or to cut the vowels short and sharp and do away altogether with diph- thongs. Thus the Mota tauwe a hill is at Motlav to. One set of people think the others speak ' thick ' or ' tliin,' ' large ' or ' small ' accordingly. Allowing for the shortening and lightening of vowel sounds, it may be said that the vowels in Melanesian languages change much less than the consonants. The shifting of a vowel by attraction to the one that succeeds it is not uncommon, especially in the Article and Particles of the New Hebrides and the Banks' Islands. * In Eotunia, hual for Itula moon, uas for nsa rain, /((//an (or fa Una ear. Phonology. 2 1 9 In Lepers' Island there is a singular inconstancy in the vowels ; a word will be pronounced first with one and then with another, without any apparent reason ; wai, water, or toei. 11. The phonetic character of lang-uag-es to the eye depends very much on the proportion of consonants to vowels which they present. Observers are not un^\'illing' to divide into distinct families lang-uag-es which show very harsh consonantal syllables or open syllables with abundant vowels. The Melanesian lang-uages differ very much among- themselves in this particular. The lang-uages of the Solomon Islands allow none but open sj'Uables, and are besides, in the South- eastern Islands of the group, very vocalic because of the fall- ing out of consonants. The lang-uag-es of the Southern New Hebrides present a g-reat contrast to these, exhibiting- very harsh combinations of consonants. Many lang-uages are of an intermediate character: Fiji closes no syllable \ nor, in spite of the appearance it may present, does Neng-one ; the Northern New Hebrides lang-uag-es dislike a close syllable ; Lepers' Island only closes with m, n, u, and w. In the Banks' Islands there is g-reat diversity mthin very little space : JNIota does not refuse or dislike to close a syllable or to bring- tog-ether consonants in harsh combination, but is very vocalic in g-eneral character ; Motlav, on the contrary, casts out every vowel it can, and is as consonantal as the worst of the New Hebrides languag-es. Mota and Motlav (Mota lava great Mota) are seven miles apart, inhabited by people identical in every respect, even in language ; but they speak their common language in very different ways, and have made their respective dialects so unlike that they are mutually unintelligible. Although, therefore, different regions present different characters of language in this respect, it is quite impossible to treat such difference as fundamental, or perhaps as worthy of more than particular observation when the several languages come under view. ^ ' This is true of the Ban dialect ; but some of the other Fijian dialects admit a close syllable ending in m, e. g. homhom, tarn. There are also unusual combinations of consonants, as mt, mn, tl, mbr.' — Rev. L. Fison. V. NUMERATION AND NUMERALS IN MELANESIAN LANGUAGES. I. Numeration. The three systems of numeration which are based on the practice of counting- on the fing-ers are found in Melanesia. ' To count the fing-ers of one hand up to five, and then g-o on with a second five, is a notation by fives, or, as it is called, a \ quinary notation. To count by the use of both hands to ten, and thence to reckon by tens, is a decimal notation. To g-o , by hands and feet to twenty, and thence to reckon by < twenties, is a vigesimal notation ^.' In some of the islands of the New Hebrides group and in the Banks' Islands the nota- tion is quinary ; in other islands of the New Hebrides, in Fiji and in the Solomon Islands, it is decimal ; in the Loyalty y Islands, New Caledonia, and in Anaiteum, the notation is, or was, vigesimal. It must be premised, however, that in none of these places, except in one part of the New Hebrides, is the system purely quinary or purely vigesimal. That is to say, the advance to hig-her numbers is not made by fives, but by tens, where the notation is quinary, and where it is vig-esimal the advance up to twenty is made by fives. In the decimal system each numeral is distinct, from one to ten, as in Eng-lish, and all further advance up to a hundred is made by the use of these numerals. But in the Melanesian lang-uages, whose system must be called quinar}^, the numerals up to five are distinct ; the digits of the second hand are named with reference to the ' Tyler, Primitive Culture. Numeration. 221 first ; ten has its own name without any reference to five ; and further advance is made by tens, not by fives. Thus, for ex- ample, in Mota tuivale, nirua, tiitol, nivat, tavelima are the first five numerals ; in lavearua seven, laveaiol eig"ht, laveavat nine, the numerals used for two, three, and four, are evidently repeated with a difference : but when ten, sarvavid, is reached the word is quite distinct, there is no repetition or recallin<»- of five ; and further advance is made by tens, not fives. This, then, is not purely quinary ; five is used to g-ct to ten, and then the notation becomes decimal. A purely quinary nota- tion would have no ten, that number would be expressed in fives. Such a notation appears in Tanna and Fate of the New Plebrides ; in Tanna karirnm is five, karirum karinim ten ; in Fate rua is two, lima five, and ten is relima, twenty relima rua ; ten in fact is two-five and twenty two two-five. In the same way, with regard to the vigesimal notation found in Melanesia, it is not purely vigesimal, but quinary up to twenty, and after that vigesimal. A purely vigesimal notation would provide distinct numerals from one up to twenty, as the pure decimal system does up to ten. If such a series of numerals be found elsewhere, there is none such in Melanesia ; the advance up to twenty is made by fives, by the fingers and toes. For example, in Nengone five is expressed by se dongo, which means that the counting of the fingers of one hand is finished by bringing them to a point together ; after- wards the counting goes on with five and one, five and two, till the second set of fingers is finished, and ten is called reioe tubenine, two rows of fingers. In this way by going on to reckon the toes they reach twenty, which they call re ngome a man. Having reached this number, they go on with twenties, with ' men,' forty is reive re ngome 'two men,' and so on up to a hundred, ' five men.' The vigesimal notation in Melanesia, curious and interesting as it is, is based on the quinary, and must be taken with it. In fact the Melanesian languages have a pure quinaiy system, and a pure decimal notation ; and between these a quinary system of notation, which becomes decimal when ten 2 22 Melanesian Languages. is reached in some languages, and vigesimal AA'hen twenty is reached in others. It may be taken for granted that the oldest method is the quinary^, and it is pretty certain that the decimal notation in jNIelanesia is comparatively recent there and introduced. It ^^^ll be well, therefore, to begin mth the jjurely quinary, and to take the decimal last ; the combinations of quinary and decimal, and quinary and vigesimal, being intermediate ^. 1. Pure Q^uhiary Kotaiion. — No word for ten is in use, except such a one as shows five to be the number really in view. In Tanna harirum karlrum, five five, stands for ten. In Eromanga, by an advance, ten is narolim two fives, in Fate relima is the same. In Sesake dua is two, lima five, dua lima ten, twenty dua lima dua two two-fives. In the neighbouring island of Api lua lima, and tua lima, is ten, lua and tua being two, lima five. The region in which this purely quinary notation is in use is very limited ; the southernmost island of the New Hebrides, Anaiteum, having a ^dgesimal form, and the more northern islands having either a decimal system or a word for ten. It should be observed that the power of rising to high numbers ' The way of reckoning on the fingers differs in various islands. In Nengone, as has been said, the fingers are turned up and brought together at five. In the Banks' Islands the fingers are turned down. This is often done with the spoken numerals, often without the use of words. The practice of turning down the fingers, contrary to our practice, deserves notice, as perhaps explaining why sometimes savages are reported to be unable to count above four. The European holds up one finger, which he counts, the native counts those that are down and says ' four.' Two fingers held up, the native, count- ing those that are down, calls three ; and so on till the white man, holding up five fingers, gives the native none turned down to count. The native is non- plussed, and the enquirer reports that savages cannot count above four. ^ The difference between the Melanesian numeration and the Australian is remarkable. In none of the Melanesian islands is found the incapacity for counting above two or three that is at least common in Australia. In Xew Guinea, among people close to Cape York in Australia, the natives of Erub, Fly River, and Tauan use the Australian way of counting : in Erub i netaf, 2 neisi, 3 neisi-netal, 4 neisi-neisi \ in Tauan i urapoii, 2 tikasar, 3 ukasar- vrapon, 4 ukasar-ukasar ; in Fly River i au, 2 ctoa, 3 netoa-nau, 4 netoa-netoa. There is nothing in the Melanesian languages here collected which cor- responds to two-one for three, two-two for ft)ur. Ntimeration. 223 is not impaired by this way of counting*. There are in Fate words for a hundred and a thousand ; rellma, althou<^h in fact it means two fives, and not one ten, has become to all intents and purposes a sing'le numeral. No doubt also Ina lima, thoug-h distinctly two numerals, two five, has come to occupy in the mind the place of a single word. Although they may say ' two-fives,' and not say ' ten,' yet in fact they count numbers above ten by two sets of fingers, and not by one set of five. When the second hand comes to be used in reckoning it is interesting to observe how the digits belonging- to it are named. In Tanna, where karirum is five, riti one, karirum riti is six : in Eromanga, sukrim nam, five two, is seven. This corresponds to the lima sa five one, lima zna five two, for six and seven, given by Humboldt in the Kawi Sprache. It is the simplest and no doubt the most ancient method ; but there is another, which Avill be considered further on, in which a sign is affixed to the nu.meral used on the first hand to show that a digit of the second hand is meant, and five is not repeated. It is as if seven were called the 'other two,' or the 'two above:' as in the Fate rua two, larua seven, tolio three, latolu eight. This appears to be an advance on five-two, five- three. 2. Imperfeci Becimal Notation. — In this system there is a word for ten ; after five is reached there is no further mention of this number. So far it is decimal ; but the digits of the second hand have not their own independent names as they have in a purely decimal notation : they are reckoned by words which correspond to the names of the digits of the first hand. In this the system is quinary ; the two hands are always present to the view, the succession from one to ten is not a simple continuous series but has a joint in it ; everything is measured Avith a two-foot rule. This, no doubt, is an advance upon a purely quinary nota- tion, and practically does as well as if it were purely decimal. In Mota they have invented for themselves a decimal series for a game, although they use this half quinary method in 224 Melanesian Languages. ordinary affairs^. An example from that languag-e will show the character of this imperfectly decimal method of reckoning- : — 1 tuwale 6 laveatea 2 nirua ' 7 lavearua 3 nitol 8 laveatol 4 nivat 9 Javeavat 5 tavelima 10 sanavul. It will be seen that the word for ten has no reference to five, and that five is not repeated on the second hand ; but with a different prefix the same numerals, rua, lol, vat, serve for the second, third, and fourth digits of both hands. In fact the word laveatea, six, is the same in construction, for tea is a form of the numeral most commonly used for ' one.' When this ]\Iota numeral series is examined it appears that the pre- fix ni is a verbal particle, the Numeral in that form is being used as a Verb. On the second hand lavea, which is prefixed to the numerals, is most reasonably taken as a word signifying the other side, or something above ; in the same way that, as will be shown hereafter, the units are expressed in quantities above ten ; in the same way, in fact, in which in ' eleven ' and ' twelve ' in English there is contained an element which sig- nifies that the number combined with it is in a certain relation to ten. It is plain, at least, that in the words for seven, eight, nine, the numerals two, three, four, are repeated, and with no express mention of five. In Fate, thoug-h there is no independent word for ten, the same system appears: la in latesa 6, lama 7, latola 8, lajiti 9, corresponds to the Mota lavea. This method of forming the numeral series up to ten prevails in the Northern New Hebrides, Banks' Islands, and Santa Cruz, and in the curiously isolated language of Savo in the Solomon Islands. In the Banks' Islands there is nothing very different from the Mota example given above. The language of Ambrym in the New Hebrides is very distinct, but the same way of forming- the ' There is a third way of counting in Mota, (for which see the Mota Grammar,) where Numerals properly speaking are not used, but descriptive expressions employed to avoid the common numerals. Numeration, 11^ numerals is found ; one hi, two ru. three snl., four fr on the first hand ; and on the second hand six Inse [se—tea in INIota), sev^en luru, eight lusul, nine I'lafer. In ^Nlalikolo, one sika'i^ six snkai, two e-tia, seven wJii-u, three eroi^ eight orol, four evatz, nine tvhi-vafz, show the same formation. In Santa Cruz there is the difference that a suffix marks the numerals of the second hand ; one eja, six ejame, two ali, seven elhne, three fli'?/', eight otmne, four «/;?/(?, nine ojm- eiiie. In Savo one eJa, three ^^/o, four agava, are repeated in /;c»yt»a six, 2^ogoro seven, kuava nine. The numerals of New Britain, given by Mr. Wilfred Powell, correspond, except in ten, to those of the Banks' Islands : one tikai, two urua, three oUil, four ivaf, five a lima ; six Up fikai, seven lov iirua, eight lov otule, nine lov ivat, ten tiir a lim. There is another word for ten in which five is not repeated, ave nun. The same thing is found in some of the languages of New Guinea. In Yule Island am one, aharaia six, rna two, aharua seven. In numerals g-iven bv Latham, a dialect of Seroei, near Port Dorey, has one hoiri, six boiri-kori ; two boroe, seven bor-kori; three boforo, eight boto-kori; four boa/i, nine boa-kon. Here it is evident that there is a prefix bo to the numerals on the first hand, and a suffix kori with those of the second. In another language, Ron, onemegnokor eight, onenfak nine, evidentl}" repeat ngokor three, and/a-^ four. These are all cases in which the digits belonging to the second hand are named with reference to those of the first, but without mention of five. There are some languages which, though they have advanced to a word for ten, still make up the numbers between five and ten by the addition of one, two, three, four, to five. Such is that of Duke of 1 ork Island, where seven is limadi ma ruacli. In the language of Yehen or Yengen in New Caledonia, given by Von der Gabelentz, the numerals after five are expressed in the same manner, nini five, nem wet six (i.e. nim we hets, Jiefs being one), nim toe Ink seven, nim we yen eight, nim jjo vits nine ; the Q 2 26 Melanesian Langtiages. numerals two, three, four being* Iie-luJc, he-yen, po-vits. The word for ten is pain-fhil-. In two lang-uag-es of the Malay Archipelago which have a name for ten, the same wa}^ of forming* the numerals of the isecond five is found. In Ende of Flores lima is five, lima a six, lima zua seven. The formation of the numeral eight is different; wutit, is four, rua hutu, two fours, is eight. In En- ganho, near Sumatra, alima is five, adoea two, and aliniei adoea is seven. A pure decimal series has not yet been formed ; the system is still quinary up to ten. 3. Tigesimal Notation. — The example of the Nengone lan- guag-e has already been g-iven, showing that up to twenty, which they call 'a man,' the notation is quinary, five being in terms the counting of one set of fingers, and ten the completion of two sets. Beyond twenty, though with multiples of twenty they use vigesimal notation, they have to recur to the quinary for intermediate numbers. Forty is two twenties, two men, rewe re ngome, thirty is ' one man and two sets of fingers,' i. e. one twenty and two fives, sa re ngowe ne rewe tuhenine. In Lifu also they count by ' men,' twenties, and advance in a purely quinary system to twenty. Five is tripi or fjipi, two is lue,ten is two-five, luejri, fifteen is three-five kreni-jn, twenty is cha-atre or ca-aij one man, a hundred is five men, ijiin o afj. The same method of counting by ' men ' as twenties is shown by Von der Gabelentz in two j)arts of New Caledonia. In the southernmost island of the New Hebrides something of the same system was found, and there, as in the Loyalty Islands, has been made away with by the Missionaries, who have substituted the less cumbrous English numerals. The Kev. J. Inglis does not admit the native numerals into his Grammar of the Anaiteum language at all, and only the first four into his Vocabulary. He gives us to understand, how- ever, that the Anaiteumese counted by fives up to twenty, using their fingers and toes ; but it does not appear that they used the word ' man ' for twenty, or indeed rose beyond twenty at all ^ ' 'The P.apiians proceed thus : They ccmiit the fingers on one hand till they Ntimeration. 227 These all belong- to one reg-ion, and there is no other part of Melanesia in which a vigesimal notation can be said to exist. At the same time there is a way of countino- bv twenties still in vogue in another part of INIelanesia, and there are traces of the same practice far away. In Bugotu, Ysabel, they have a pure decimal notation, hanavulu being- ten ; but for twenty they like to say tuiugu, and for multiples of twenty so many tutugu. At Savo they use neholo in the same way, mle being- ten. These words are rather collectives, perhaps, than numerals ; but the presence of a name for the unit above twenty, Usoa, different from that above ten, mpUi, is a mark in Savo that counting- by twenties, neholo^ is an ancient practice. This corresponds to the distinction in ]\Ialay between bias, the unit above ten, and Vikul, the unit above twenty, which seems to show that the counting by tens only was not the original practice in that language. In PoI;sTiesia also counting by scores, twenties, is part of the system of numeration. In the Marquesas ten is onohuu, twenty is fekaii ; which last word in the IMaori of New Zealand is ten. When twenty has been reached farther advance is made by tekan, thirty is tekait me ouokuu, forty e ua tekau two score, one hundred e iima tekait five score. In the Sandwich Islands the word used for ten when twenty is named is not the same word which is used for ten by itself or in any other multiple of ten. Ten is unu, for twenty iva kalua, two nines, is used, for thirty kana kolu, three kana. In these methods of numeration twenty is differently treated from other multiples of ten, which seems to show that it has a different history, that there was a time when twenty was the come to five, and then they say my hand, whatever that word may be in the language, for five ; then my hand and one for six, my hand and two for seven, and so on till they come to ten, for which they say my two hands ; they do the same thing with their ten toes, and then say my two hands and my two feet for twenty. All beyond this in Aneityumese is many, a great many, a great great many.' Mr. Inglis takes the Anaiteum people as an example of Papuans ; but his description is plainly inapplicable to the Melanesian people in the neighbouring Loyalty Islands, or in the other New Hebrides, or in any of the groups the languages of which are here examined. Q 2 2 2(3 Melanesian Languages. limit of counting". The cause of this may well have Leen that the natural limit of counting was the number of the fing-ers and toes ; but it does not appear that in any of the Ocean languag-es the feet were directly referred to, except perhaps in Anaiteum, and the term ' man ' to represent twenty is confined to New Caledonia and the Loyalty Islands. 4. Decimal Notation. — The purely decimal series of numerals, in which each number is expressed by an independent word, is found in the Melanesian languages, in the New Hebrides in two islands, Lepers' Island and Whitsuntide, in Fiji, and in the Solomon Islands. In the latter, the isolated language of Savo is an exception ; and in New Britain a quinary system has been noticed. The same decimal series substan- tially is in use in the Polynesian islands, and in the ]\Ialay Archipelago. It is important here, without considering the particular W' ords, to ascertain how far the ten numerals of the INIelanesian decimal series are the same with those of Polj^nesia and of the Malay Archipelago. Mr. Wallace^ in his Vocabularies of thirty-three languages of the latter division, gives the numerals, and it will be found that generally they are the same throusfhout. In 33 languages for one 22 ■ have some form of sa ,, two 30 rua » three 27 tol » four all pat » Jive six 31 30 lima an » seven eight nine 29 24 29 pifu walu sio „ ten 12 ,, piilu. It must be noticed that the Malay, from which this group of languages is named, by no means has a characteristic set of numerals ; the words for three, seven, eight, and nine, tiffa, tujo/i, delajjan, samlilan, are exceptional. The IMelanesian decimal series is identical with that just given of the great majority of the languages of the Malay Archipelago, not with the ISIalav. Nu mera tion. 229 The Pol^Tiesian numerals are substantially the same, with a general ag-reement amono- themselves, except in the case of ten ; and the Polynesian numerals are of course in use in the Pol3aiesian settlements in Melanesia. Have not then, it may be asked, the Melanesians who use this series of numerals borrowed them as a whole from the Polynesians ? It is certain that they have not. The first Polynesian numeral is taJii^ fasi, kasi, from which it is not possible that sa has been derived, thoug-h no doubt ia in tahi is the same as sa. So the Polynesian four is fa, tvha, ha, a, no doubt the same with the Malayan ampat, but not a form from which the Melanesian vat could have come. In Fiji, which is so close to Tonga and in communication with it, four is indeed va, not vati, but one dua and ten tini are altogether different from the Tongan taha and hongafnlu. We have, therefore, to conclude that the Melanesian decimal series of numerals is not borrowed from the Malay, from which it differs in four numerals out of ten, or from the Polynesian, than which it has stronger forms, but that it is identical with that generally in use in the Malay or Indian Archipelago and Madagascar. But there can be no doubt that, though not introduced as as a whole either from the Malay or the Polynesian languages, the purely decimal notation has been comparatively lately adopted by the Melanesian people, whose system was originally quinary. The numerals up to five, whether a quinary or a decimal notation be used, and ten, are generally the same ; it is the numerals of the second hand 6, 7, 8, 9, which have been introduced into Melanesia : and these most certainly not from jNIalay, in which three out of the four are different from those used in Melanesia. These numerals, and the general use of a decimal series, may possibly have been introduced from Polynesia, though the form of the numeral nine does not encourage the notion ^ How and when they were * Not one of the Polynesian words for nine has a form with a' ; Tongan alone has Ti, hiva. But the characteristic word in the Malay Archipelago, not Malay, is sio, and in Melanesia the sibilant is more common than the aspirate, and is represented in the Fiji civa. 230 Melanesian Languages, introduced, and how it came about that their introduction was partial, would be an interesting- and difficult inquiry. The consideration, hereafter, of the particular numerals may throw some lig-ht upon the subject. It is a remarkable thing- that in one island of the group in which the notation is quinary, but has a word for ten and counts by tens — in Mota of the Banks' Islands, there is a purely decimal series of numerals used in a g-ame ^. These numerals are all different from any that I can discover else- where ; they appear to be entirely indig-enous, and not in use even in the island nearest to Mota. We have, therefore, the phenomenon of a people inventing* a series of numerals for themselves which is decimal, and using- it in a g-ame, while they continue to use for ordinary purposes their old notation, the basis of which is quinary. It is probable that in the use of their words for seven or eig-ht, lavearua, laveatol, thoug-h the numerals two and three are expressed in them, the deriva- tion and orig-inal meaning- of the words are no more before the native mind in Mota, than it is in Eng-lish with ourselves when we use ' eleven ' and ' twelve,' or in French when they say ' douze ' and ' treize.' 5. Numeration beyond ten. — The methods of carrying- on the numeral series beyond ten in Melanesian lang-uag-es are two, one of which is of much interest. The addition of the unit to the ten with or without a conjunction is simple, and would deserve little consideration if it did not apjiear a sig-n of the simplification of a lang-uag-e. The introduction of the unit above ten, with an explanatory particle or desig-nation of it, has all the appearance of an orig-inal idiomatic method. If we look at Mr. Wallace's list of numerals in the collo- ' This decimal series is i qasa, 2 ivura, 3 lovi, 4 tama, 5 rina, 6 qaru, 7 lim, 8 naga, 9 viga, 10 loesu. These words are not all unintelligible ; sa, though never used as one in Mota, is in fact the same as tea, which appears in the Mota word for six, and qasa i, the first qa, corresponds to qaru 6, the second qa ; wesu in itself means completion, arrival, and is elsewhere used for a hundred. In Florida also there is a series of numerals used only in play, which, however, are mostly the ordinary numerals reversed ; ura for rua 2, lotu for tolu 3. Numeration. 231 quial Malay of Sing-apore we see satu one, dua two, and sapuloh ten ; eleven and twelve are sapnloh satu, sapuloh dua ; the unit is simply added to ten. But in his Java numerals, sa being- one, loro two, and pulah ten, eleven is s7oalas, and twelve rolas ; las added to the unit designates it as a sum above ten. Similarly in Salibabo in ressa eleven, there is no mention of mapuroh ten, but res is evidently the Java las, and sa is one. But in true ISIalay, not the ling-aa franca of commerce, sahlas is eleven, duahlas twelve ; hlas, evidently the same as the Javanese las and Salibabo res, is the desig-na- tion of the unit above ten. Another expression of apparently the same character is given in Masuratty : polo is ten, sia one, dua two, polo tern sia eleven, polo tern dua twelve ; in Wayapo polo is ten, umsium one, rua two, polo geren ensmm eleven, polo geren rua twelve. In these tem and geren appear to be words designating the unit above ten. It is plain, then, that in the Malayan reg-ion there is an idiomatic use of a desig- nation for the unit above ten, not a Conjunction or a Pre- position, but in fact a Noun ; and that where, as in the colloquial Malay, the lang-uag-e has been simplified and disturbed this idiom has been g-iven up. It is just the same in Melanesia. There is commonly in the languages which retain their quinary notation, a word designating the unit above ten, or above five, corresponding to the Malay hlas, and in some cases identical with the Masuratty tem ; while the languages which have adopted the no doubt com- paratively recent and foreign decimal notation have generally no such idiom. Thus in the Solomon Islands the isolated Save language, with its quinary system, has the designation nipili for the unit above ten, while all the neighbouring islands which use the decimal series add barely the unit to the ten. Among the Polynesian languages in the Sandwich Islands ten is umi, one hahi, two lua, eleven is umi kuma ma kahi, twelve umi kuma ma hia : kumi is described as a number or company, and ma as signifying company, and coming after the word to which it is applied. The explanation of the words signifying eleven and twelve, therefore, is ' ten, the 2^2 Melanesia^ Languages. number in company one, or two.' The same word is used in the Maori of New Zealand, tuma a number in excess ; and tuma may very well be taken to be the same word with tern of Masuratty and with the temei^ demei, uuniei of the Banks' Islands. In Melanesia a word of this kind is not used only to describe the unit over ten, but with the same notion the unit over five where the numeration is quinary, and the sum also, whatever it may be, over a hundred. The meaning- of the word is the ' sum over,' whether over five, ten, or a hundred. In the Lorjalty Islands, in Neng-one, the name of the unit above ten is cemene or xecene, eleven is rewe tuhenine ne sa re cemene, literally, two the sets of fing-ers and one the sum above. In Lifu ngemen is the name of the number above five, ca ngemen is six, hte ngemen seven, that is, the number-above (five) one, the number-above two. The designation of the number in the next set of five is ho, the digits belonging to the first set of toes are ca ko, hie ko eleven, twelve. The digits again between fifteen and twenty are called liua'i ano, ca huai ano, lue huai ano sixteen, seventeen. Thus each set of five has its appropriate name for the quantity above five, or ten, or fifteen. In the New Hehrides, in Fate, the designation of the unit above ten is teniatl, eig-hteen is relim iskei temati latolu, one ten, the unit-above eight. This temati recalls the tern, tuma, already mentioned. In Espiritu Santo the name of the number above ten shows itself plainly as a noun ; forty-four is sonovnl vat na vana movat, ' tens four, its unit above is four.' This word va is probably the same as ve of Araga, Whitsuntide, used for the number above a hundred, and the Santa Cruz wa ; both of which are, like it, constructed as nouns with the pronoun suffixed. In Araga and Lepers' Island ^ in Ijoth of which the decimal There is another way of couutiiig in Lepers' Island without the domagi, Numeration. 233 series of numbers is employed, the name of the unit above ten is dovaa, the word ah-cudy familiar ; in Lepers' Island, twelve is sanavula dovaagi fjalrne ; in Whitsuntide twelve is hanaviilii doma/i gairua, ten, its doma two. The same word is used in Aurora. In the Banks' Islands the same word in varying- shapes is universally employed ; in Merlav demei, in Gaua dome, Lakona jime ; in Vanua Lava denae, femei, femegi, nuvaegi ; in Mota nuraei; in Mota Lava dome ; Volow neme ; Ureparapara dem.e. It should be observed that in these lang-uag-es there is no need for the mention of ten ; for twelve it is enoug-h to say domagi gairne, numei nirua, as dua bias in Malay, or twelve in Eng-lish. In the Torres Islands the word is different, mahali, the meaning of which is a thing'-above. In Fiji the numeral above ten is simply introduced with the particle ka (the ga of Lepers' Island above), but maul is also used with or without ka ; tini mani foin, or lini ka mani tolu, thirteen. In Santa Cruz the unit above ten is ioa constructed as a Noun, naplu na wade tu thirteen. In the Solomon Islands the unit is g-enerally added simply to the ten, or with a Conjunction ; but in Fag-ani, San Cristoval, matara is the sum above either ten or a hundred. In Florida a Verb is often used, rua hawavulu me sara ma twenty-two, i. e. two the tens, (and) it has come up to two. In Savo the characterisiic Melanesian idiom reappears, the number above ten is nipifi, edo nipiii twelve, edo is two, and ten is not named. In this singular language, while mjAii is used to designate the number above ten, and any multiple of ten except twenty, another word, lisoa, is used for the unit above twenty. This corresponds remarkably with the Malay use of t'lkul above twenty, instead of bias above any other number of tens ; and it surely points in both cases to something of a vigesimal notation. In both twenty is treated as in which is instructive as showing the meaning of the word commonly used for ten to be a 'set ' not a numeral ; forty-eight can be expressed, navulu gevesi, navulu gailimegi gaiwelv, i.e. tens 4, the fifth ten (set), eight. 234 Melanesian Languages. many Melanesian languages a hundred is and as ten is, as the conclusion of a series beyond which counting goes on wath a new expression. From what has been shown, we may observe an agreement in this practice of using a word as the designation of the unit beyond ten, or of the digit beyond five, w^hich extends from the Malay to the languages of New Caledonia and the Loyalty Islands, and is explained as well as exemplified in Polynesia. The identity also is very remarkable of tern, ttme, tuma, dovm, mime, in BourUj New Zealand, and the New Hebrides and Banks' Islands. 6. Numeration heijond a Jmndred. — It is a natural extension, and in another point of view an illustration, of the way of adding units to tens, to use a word also to designate a number above a hundred. This use is not so common in Melanesia as the other. In the Loyalty Islands, where there is a designation for the digit above five, the number above a hundred is simply added ; in Lakona, where y?me^ is the unit above ten, there is nothing for the number above a hundred. In the Northern New Hebrides, however, and the Banks' Islands, it is common to use such a word, and it is commonly possible to learn the meaning of the word. A word in general use is in the Mota form avavm, a form which shows it to be a noun, and which is derived from the word av to pile one thing upon another. In Mota no is melnol vatuwale o avavin sanavul, hundred once, and the pile above ten : in Lepers' Island vudohie vagatuwale, avigi sanavulu. In Volow it is nivivm ' its number above ' so many. In Aurora the word used is Ian, a verb meaning to turn one thing over upon another ; 320 is medol tol, Ian ivonana sanwuhi ma, ' hundreds three, turn over upon it two tens.' In Araga, Pentecost, they use ve for the number above a hundred, no doubt the same word as va used in Espiritu Santo for the unit above ten. In part of Vanua Lava they use the same word above one hundred as above ten, feme. In Savo the same name is used for the number above a hundred that is used for that above the score, lisoa. ^ Numeration. 235 MELANESIAN NUMERALS. I. Quinari/ {no icord for toi). New Hebrides. I 2 3 4 5 10 Anaiteum e thi e ro e seij e manowan ikman Tana riti ka ru ka bar ke fa ka rirum karirum- kariruin Eromaiiga sai du ru di sil di vat sukrim narolim Fate iskei rua tolu bate lima relima 6 la tesa 7 la rua 8 la tolu 9 la fiti Sesake sikai dua dolu pati lima dua lima 6 la tesa 7 la dua 8 la dolu 9 lo veti Api tai lua tolu vari lima lua lima 6 rai 7 lua 8 tolo 9 vari Pama tai e lua e tola e hati e lime ha lua lim 6 a hitai 7 lu 8 tolu 9 hati. II. Imperfect Decimal {a loorcl for ten). New Hebrides, I 2 3 4 5 10 Malikolo sikai e ua e roi e vatz e rima singeap 6 su kai 7 whi u 8 roi 9 whi vatz Ambrym hu ru sul vir lim saMaul 6 li se 7 lu ro 8 hi sul 9 Ua ver Espiritu Santo tea rua tol vati Una sawovul 6 arave 7 ve rua 8 ve tou 9 ra tati Aurora tewa i rua i tol i vat tavalima sa/iwulu 6 lava tea 7 lava rua 8 lava tol 9 la vat. Banks' Islands. Merlav tuwale i rua i tol i vat tavali?» sa«avul 6 live tia 7 Uvea rua 8 Uvea tol 9 lev vat Santa Maria, Gog tuwal i ru i tol i vat tevelim saMovul 6 leve te 7 leve ru 8 leve tol 9 leve vat ,, Lakon tuwa ni ru ni tel ni vas tivilem gapra 6 le tuwa 7 lavu ru 8 lavi tel 9 la vas Vanua Lava, Pak vu wal vu ru vu 'ol vu ve 'evelem sawwul 6 leve 'a 7 leve ru 8 leve 'ol 9 leve ve >> Vureas tuwel ni ro ni tol ni vat tevelem sajMol 6 leve te 7 lovo ro 8 lo vo tol 9 leve vat » Mosina towal ni ru ni tol ni vet teveli;» sa?jwul 6 leve te 7 livu ro 8 livi tol 9 livi vet Mota tuwale ni rua ni tol ni vat taveliwa sawavul 6 lavea tea 7 lavea rua 8 lavea tol 9 lavea vat 2^6 Melanesian Languages. Banks' Islands {continued). Motlav I vi twag 2 VO ro 3 ve tel 4 ve vet 5 tevelem lO so»wul 6 leve te 7 livi ro 8 leve tel 9 leve vet Volow vo twa vo ro ve tel ve veat tevelem saHwil 6 leve te 7 leve ro 8 leve tel 9 leve veat TJreparapara Torres Islands vo towa 6 leve jea vu iia vo ro 7 leve ro vu rua vo tol 8 leve tol ve tal vo vet 9 leve vet ve vat teveliem tevelima saMowul hewawol Nifilole Savo 6 livi jia 7 lave rua 8 lave tal 9 liv vat. Santa Cbdz. e ja a li a tu a pue 6 e jame 7 lime 8 tume 9 pueme ni_9'i lilu eve uva 6 wele g\ 7 pole lu 8 po le 9 polo ve. Solomon Islands. ela edo igiva agava 6 pogo a 7 pogo ro 8 (kui) 9 kua va. III. Vigesimal. Loyalty Islands. na vlu nu na vlu vili nukolu ara atale Nengone Lifu '«\ Lepers' Island Whitsuntide Ilotunia I 2 3 4 5 20 sa rewe tini ece se dongo dongo ne sa fdongo ne '\ rewe grdongo ne fdongo [ tini "[ ece ne , f rewe tube IO< [ nine re ngom« ca lue kceni eke tji pi ca nge- flue g-ne- ' \ men fkcBni gne- ^Jeke nge- 10 lue pi men \ men ^y men ca ko 12 lue ko 13 kceni ko 14 eke ko 1 5 koeni pi ca huai ano ^ Jlua huai '\ ano fkojni huai , , feke huai \ ano 9\ ano ca atj. IV. Decimal. New Hkbrides. I gai tuwale 2 gai rue 3 gai tolu 4 gai vesi 5 gai lime 6 gai ono 7 gai bitu 8 gai welu 9 gai siwo 10 sawavulu I gai tuwa 2 gai rua 3 gai tolu 4 gai vasi 5 gai lima 6 gai ono 7 gai vitu 8 gai welu 9 gai siwo 10 ha«vulu. Fiji. I e dua 2 e rua .S e tolu 4 e va 5 e lima 6 e ono 7 e vitu 8 e walu 9 e ciwa 10 e tini I ta 2 rua 3 folu 4 hak 5 liam 6 on 7 hif 8 vol 9 siav 10 saghulu. I Numeration. So LOMON Islands. Ulawa I e ta 2 e rua 3 e 'olu 4 e liai 5 e lima 6 e one 7 e lii'u 8 e walu 9 e siwa lo tanahulu Malanta I e ta 2 e rua 3 e 'olu 4 e hai 5 e lime 6 6 ono 7 6 hi'u 8 6 walu 9 e si we lO tanahulu San Cristoval, "SVa«o I tai 2 e rua 3 e 'oru 4 e hai 5 rima 6 ono 7 bi'u 8 e waru 9 e siwa lO tawahuru ,, Fagani I i tagai 2 i rua 3 i 'oru 4 i fai 5 i rima 6 i ono 7 i pi'u 8 i waru 9 i siwa lO ta«avuru Florida I sakai 2 rua 3 tolu 4 vati 5 lima 6 ono 7 vitu 8 alu 9 hiua lO hanavulu Vatura«a 1 kesa 2 ruka 3 tolu 4 vati 5 jehe 6 ono 7 vitu 8 alu 9 siu lo sa«avulu Bugotu I sikei 2 rua 3 tolu 4 vati 5 lima 6 ono 7 vitu 8 alu 9 hia lO salage Gao I kahe 2 palu 3 tolu 4 fati 5 lima 6 famno 7 fa fitu 8 falu 9 fa hia 10 faboto New Georgia 1 meke 2 karua 3 hike 4 made 5 lima 6 I onoono sa 7 fopa 8 vesu 9 sia 4 ampat lO 5 na^uru ^ Malay 2 dua 3 tiga lima 6 anam 7 tujoh 8 delapan 9 sambilan 10 sapuloh Slalagasy I isa 2 roa 3 telo 4 efatra 5 dimy 6 enina 7 fito 8 valo 9 sivy lO folo Maori I tahi 2 e rua 3 e torn 4 e wha 5 e rima 6 e ono 7 e whitu 8 e waru 9 e iwa lO tekau. II. The Grammar of Numeration. 1. Cardinals. — Numerals in the Melanesian languages are used as Nouns, Adjectives, and Verbs ; that is, the same word expressing- number may be used grammatically in either way. For example, in Mota two is rua, sanavul ten ; when twenty is expressed as sanavul rua, it is plain that sanavul is used as a Noun and rua as an Adjective, two tens. If they speak of ten men, tamm sanavul, the Numeral is an Adjective which just now was a Substantive ; if in speaking of the number ten having been reached we say me sanavul vefa, sanavul is gram- matically a Verb, it is literally ' they have tenned.' This does not mean that in each INIelanesian language the Numerals are thus used indifferently as occasion may serve, ' The Numerals of Mafoor, N. Guinea, are i sai, 2 did, 3 kior, 4 fiak, 5 rim, 6 onem, J Jik, 8 war, 9 site, 10 samfur. 238 Melanesian Languages. but that Numerals may have either of these grammatical characters, though each languag-e perhaps prefers one use to another. In the case of a Numeral being* preceded by an Article it is of course a Noun ; if it takes a verbal Particle, it is in fact a Verb, though in English we translate it as an Adjective; and when a numeral plainly qualifies a preceding Substantive it is an Adjective, as in English. An Adjective in these lang'uages as a rule follows the Substantive it qualifies, and one Numeral following- another may qualify the one before it and be a numeral adjective, as in the Mota sanavul rua above, two tens. But two Numerals may stand together without any grammatical relation ; as in Florida hanavulu rua means twelve, not twenty, is ten-two, not two tens. The Numeral also as an Adjective may precede another Numeral which is a Noun. In Nengone tubenine a row or set of fingers is used for five, retve is two, ten is expressed by rewe tubenine two sets of fingers. In this instance tubenine can hardly be called a Numeral. When the expression sa re ngome, one man, is used for twenty, the Noun ngome, man, has even the Article re with it : but the same construction is seen in Florida rua hanavulu above, and in the Fiji rua sagavulu twenty, in the Malay dua pvloh, and the Samoan luafulu ; words and construction are the same. The idiomatic uses of the several languages differ in this respect. In Florida havi^avulu rna, like Malay (low Malay) sapuloh dua, means twelve, in Mota sariaviil rua means twenty: the Fiji rua sagavulu is not a possible expression in Mota. The Sesake dualima dua, twenty, combines two ways of speak- ing ; in dualima two fives, i. e. ten, dua comes first though dua qualifies lima, but dualima having become one word, a Noun, dua as an Adjective comes after it. It is not always plain, therefore, when a Numeral is cer- tainly an Adjective or a Noun Substantive ; but some- times the Article going with a word which is a Numeral, and not, like tubenine above, a Substantive used to express number, shows it to be in use a Substantive ; as in Malagasy the substantive character 0^ folo, the same as puloh, vulu. Numeration. 239 is shown in the expression roa amhy ny folo twelve, ' two above the ten ;' and in Fiji e nia na tini twenty, two the tens. The use of the Numerals as Verhs is perhaps difficult comparatively to ascertain, because the Numerals are com- monly obtained in that form. Any one learning- Mota would find the Numerals presenting* themselves to him as ninia two, nitol three, n'lvat four, and it would only be on con- sideration and by comparison with other languag-es that he would ascertain the true Numerals to be ruay tol, vat, and the prefix 7ii one which puts them into the place of Verbs. So in a Fiji Grammar, Numerals, under the name of Numeral Adjectives, are g-iven e di(a, e ma, e tolu, e vat, one, two, three, four, and ha diia, ka nia in eleven and twelve ; and thoug-h e and ka may be called ' numeral particles,' it is certain that they are the same with those that commonly go before Verbs. In Maori of New Zealand e and ka are the 'ordinary prefixes of the numerals below ten,' and those particles 'which pre- fixed to a word endue it wdth the qualities of a verb.'' It is often also the case that the common verbal particles of a lang-uag-e are not those used with Numerals. In Lepers' Island, for example, ga, which g-oes with the Numerals, is not used with ordinary verbs, yet it is no doubt the verbal particle in use in the neig-hbouring- island of Espiritu Santo, and the same with the Fiji and Maori ka. The true Numerals, and consequently the verbal particles attached to them, are ascertained by comparison of lang-uages among- themselves, and by observing the Numerals as constructed wath other words. In Mota ni is seen to be a verbal particle in ninia two, when sanavul rua twenty, is examined ; the very bare expression, as it seems at first sight, kanrumkarirum five-five, ten, in Tanna, appears in the light thrown upon it by other languages to be after all a verbal expression ; karirum five has the verbal ka, just as karu, two, is the same in form and grammatical character as Fiji or Maori ka rua. A prefix seen accompanying any series of Numerals may generally be taken as a verbal particle and a sig-n tliat the Numeral is used as a Verb, as in Anaiteum 240 Melanesian Languages. one e tlii, two e ro, three e seij, four e manoivan. (See Table of Numerals.) 2. OrclinaU. — Ordinals are naturally formed from Cardinals. It is remarkable that verj generally in Melanesia the ordinal 'first' is a distinct word from the cardinal 'one.' Some of these ordinals, which are unlike the cardinals of the lan- guage to which they belong, can be seen to be the same with the cardinals of a cognate language, as moa'i, first, in Mota is no doubt moi, one, in Gilolo. It is not uncommon also that a word meaning 'another' or a 'fellow' should be used for second, like the Mota tuara ^. The Ordinals are formed in the Melanesian languages by applying prefixes or suffixes to the Cardinals. In Fiji ha is prefixed, e rua two, karua second ; as in Malay duwa two, ka dmva second. This prefix ka in both languages gives some- thing like the sense of a passive participle, and is distinct from the verbal particle commonly used with cardinal nu- merals. In Fate ke or ki is prefixed in the same way, kenia second, kelima fifth. The causative prefix tvhaka in Maori makes an ordinal, wliaka tekau tenth, as faha in Malagasy does, roa two, faharoa second. The same causative prefix as vaga, or va, makes the ordinal in the Northern New Hebrides and Banks' Islands, with a substantive termination at the same time added to the cardinal : Maewo ma two, vagaruai second. Lepers' Island tolu three, va.gafoligi third, Mota vagaruei, vagatoliii. or vatol'm. This prefix, however, unless the word be thus made into a substantive, forms a multiplicative, as Fiji vakatoln, Mota vagatol, three times. The most common way in Melanesia of forming an ordinal is to suffix na, ne, oii, to the cardinal. In Nengone of the Loyalty Islands retve two, reirone third, tini three, iinone third ; in Eromanga of the New Hebrides durn two, dnrungi second ; in Espiritu Santo, with vaga also prefixed, vagaloluua third ; in Whitsuntide galrvana second. In the Banks' Islands vaga is * This is connected no doubt witli the practice of counting things in pairs. Compare the Mota xm expression for two, iami rahdia, the one that has its fellow. Numeration. 241 sometimes prefixed when this termination is used, and some- times not ; INTotlav cagrone second, vagtehie third, in Vureas rone, fobie. In the Solomon Islands this suffix, Savo being an exception, is g-eneral ; Illawa rnana, 'oliina, Florida riiani^ tolun'i ^. In Duke of York it is (Ji, Imadi ma ruadi seventh, i. e. fifth and second. In some of the Banks' Islands languages the word anai, noticed in the Vocabulary under the word ' Child,' is used to make an ordinal ; Mota n\ebiol hundred, mclnolanai hundredth ; Santa Maria, Gaua, rj(a van second, tol nan third. In the Banks' Islands generally the ordinal is a Noun and in the form of a Noun. III. Peadiar Methods and Terms used in Numeration. 1 . There is not, so far as I am aware, in Melanesia any way of counting by pairs like the use in Polynesia^. In Fiji and the Solomon Islands there are collective Nouns signifying tens of things very arbitrarily chosen, neither the number nor the name of the thing being expressed. Thus in Florida na kna is ten eggs, na hanara is ten baskets of food. In Florida these words are in no case the same as those in Fiji, and they are not so numerous, but the same objects are often counted in this manner. In Florida ten canoes or ten pud- dings are na gotji, which in Fiji are respectively a tidnudu, and a wai ; in Florida na ]}a^a is either ten pigs, or ten birds, or ten fish, or ten opossums ; in Fiji ten pigs are a rara, ten fowls a saga, ten fish a hola. There are many other words of the same kind naming tens of cocoanuts, breadfruit, crabs, shell- fish, bunches of bananas, baskets of nuts. In Fiji hola is a hundred canoes, koro a hundred cocoanuts, a selavo a thousand ' It is worth notice that with this termination the indefinite Numeral niha, how many, so many, q\iot, is made into the ordinal Adjective which has no English equivalent, the Latin quotus. '■^ ' In counting by couples in Duke of York they give the couples different names, according to the number of them there are. The Polynesian way was to use numerals with the understanding that so many pairs, not so many single things, was meant ; hokortin, twenty, meant forty, twenty pairs.' — i\Iaunsell. R 242 Melanesian Languages. cocoanuts. In Florida parego is a collective noun for ten of anything ; in Bug-otu selage is ten, tutugu twenty, things of any kind. 2. There are not in any Melanesian language, so far as I know, any ' numeral coefficients ' or ' numeral affixes ' such as are employed with numerals in the Indo-Chinese languages and in Malay. It is true that a word which is identical with the Malay buwah is used with things which strike the mind as globular^, but this is not used in numeration. There is nevertheless an idiom in giving a number in which a word precedes the numeral carrying with it the image which the things enumerated seem to present to the mind. Thus in Fiji four canoes in motion are a waqa saqai va, from qui to run. In Mota two canoes sailing together are called aka peperna butterfly-two canoes, from the look of the two sails. Using the indefinite visa so many, so many men together are tamm jmlvisa, from j'j?^^ to stick together, jpidsaviavul ten together, imltavelima five together, pnlvisa so many together ; if they are in a canoe they are sagevisa, on-board-so- many. Arrows shot, and canoes under sail are 'stand' so many tira visa, things in a bunch are sogovisa ' bunch ' so many, bats are taqa visa 'hang' so many, money is tal visa 'string' so many. In the Solomon Islands this use is not common, though in San Cristoval tae siha is the Mota sage visa. In Nengone with the number of sjiears they use naiu to strike, with the number of birds dede to fly, or te to sit, accordingly as they are flying or sitting. IV. The Melanesian Numerals. The consideration of the words used as Numerals is dis- tinct from that of the method of their use or grammatical arrangement ; their meaning in themselves and origin, if they can be discovered, must be full of interest and instruc- tion. We have seen that the Numerals which belong to the digits above five, of the second hand, are apparently of later introduction in Melanesia than those of the first hand ; as no * See Vocabulary Notes under the word 'Fruit.' WNIVER Numeration. 243 doubt the way of counting- by fives is the earliest to come into use among- mankind. It will be seen that in fact the series of the first five numbers is, generally speaking, the common property of the languag-es which arc here considered. There are many exceptional numerals to be found, but g-enerally speaking in an island language, whether in Formosa, in Madag-ascar, or in New Guinea, a list of Numerals will show the first five digfits substantially the same, and any one of these island numerals will be looked for in vain on the continent of Asia, Africa, or Australia. In New Guinea vocabularies, for example, there are often seen Numerals un- like those common in the Melanesian islands ; but it cannot be mistaken that the New Guinea numerals genemlly, in the vocabulary of any one language that may be taken, are to some extent the same as those of the Ocean languages, whereas Australian vocabularies show nothing whatever in the Numerals which is familiar to those acquainted with the island tongues. 1. — The first Cardinal is not one in which the greatest agreement prevails. There is a practice in some places in counting to begin with a word which is not used as the common Numeral : in Malagasy im is only used in counting, while iraij^ iraiJc?/ is used in composite numbers ; but this isa is no doubt the very common sa which, by itself or in a compound, is in many of these languages ' one.' The Florida sakai is compounded with it, jet in beginning to count a series keha is used for ' one,' not saJcai. Since the numerals of little known tongues are often got by counting a series, no doubt the true Numeral ' one ' is often wanting in vocabularies. The common first Numeral, however, is no doubt sa or ta. In Mr. Wallace's lists of the numerals of the Malay Archipelago, there are but five out of thirty-three in which sa is not present. It varies to se, so, si, Ida, but may be taken to be the same ; when as in INIala}^ ' one ' is satu, sa shows itself as the true numeral in sa puloh ten, sa Mas eleven. The Poly- nesian taJd, tasi, shows the same root. In jNIicronesia the Pellew Islands have ta7ig, the Kingsmill te. In New Guinea, E a 244 Melanesian Languages. at Guebe (the numerals having- the prefix pi), ' one ' is pi sa, at Arago ossa, in other lang-uages tata, sa, sai, in Redscar Bay ia, in the Gulf of Papua ta, tea. In New Britain and Duke of York takai is the same with sa/cai of the Solomon Islands, where also are iai and ela. In Santa Cruz where i often turns to ic/i, written/, ' one ' is Jia, which in the Torres Islands is vujia. In the Banks' Islands iea forms the Numeral laveatea six, and though not in use as a Numeral is well known to mean ' one^' In the New Hebrides tea is in Espiritu Santo, in Api ia, tai ; in Sesake sikai, Fate iskei, are the sikai, sakai, of the Solomon Islands, but latesa six, and the indefinite pronoun tea give forms of the common numeral. In Eromang-o one is mi, in Anaiteum etJii ; in the Loyalty Islands the word continues, in Nengone sa, in Lifu cha, in Uea Jiets [he a prefix) ; in Baladea of New Caledonia it is ta. The general resemblance, the g-cneral distribution of this numeral from the continent of Asia to the extremity of Melanesia is very remarkable. There is another quite distinct numeral, in Fiji dua, in the Banks' Islands tmva in Lakona, in Torres Islands ttiiva, tmvaga, tuioe ; with the verbal prefix votowo, votwa, vitwag, in the Banks' Islands ; where the same root no doubt makes tuwate, and, where t is dropped, vmoal, oival. In the three Northern islands of the New Hebrides teiva, tuwa, tmoale is the Numeral in use. The region occupied by this word is confined to Fiji, the Banks' Islands, and the Northern New Hebrides ; in the two latter it seems to have ousted tea, which still keeps its place in six, the first of the second hand. It is possible that the Marshall Islands dziion is the same. 2. — The second numeral is almost universally in some form Tua, the Malay dua ; thirty out of Mr. Wallace's thirty-three ag-ree in this, and quite as much agreement is found in Melanesia. There are some remarkable forms. To find dua ' one ' in Fiji and daa ' two ' in Malay, tuioa ' one ' in the Banks' Islands and tua ' two ' in Api of the New Hebrides, is not to find an apparent identity of numerals. But dua is only a ' 'In dialects of Fiji ta, taya, tia, are in use foi* one. There are besides in Navitilevu fla, hila, Ira, kila, kia.'' — Eev. L. Fison. Numeration. 245 form of rua^ which often becomes drua^ and the difference between t in some places and d in others is so slight that w^hat one European would wn-ite tua another would write diia. In Api the form lua is also found, and cliua with a not uncommon chang-e of t to cJi. The Malay dtia is in Celebes dm, which is paralleled by the form li for rn in Santa Cruz. In ]\ra]ikolo, as in Marquesas, r has fallen away, leaving- ua ; in Florida they sometimes make it rulca. In the New Guinea lists I have seen some form of the common numeral appears in the g-reater number, thoug-h it may be disguised as do. In Savo do also appears. 3. — This Numeral in the form of tol is also so nearly uni- versal in the region under view that it is only necessary to remark on some exceptions. One of the most remarkable of these is the Malay tiga, which stands alone in the languag-es w^hich are called Malayan. In Nengone three is t'mi, which is also the numeral in Tarawan of the King-smill grouj) of Micronesia. But tini in Fiji is ten, and is said to signify con- clusion. The resemblance can hardly be accidental, and raises a hesitating conjecture that there may be here a trace of counting by threes. The change of ^ to is given by Gabelentz as used for ten in New Caledonia. It is inconceivable that a \\ord should have found its way thither from New Zealand ; but if its meaning be some kind of tally there is no reason why it should not be used in both places. In fact tekau represents the tally and not the number. ' The native way of counting is by elevens, on the principle of i)utting aside one to every ten as a tally,' and thus tckatc in New Zealand means eleven as well as ten. A \\ord which in itself, though we may not ])c able to trace its original meaning, is used to signify the end of the countings Niinieration. 249 naturally rises as the practice of counting- advances to the sig-nification of a higher number than it expressed at first. Thus in Savo lale or sale is ten, which in the Torres Islands is a hundred ; the word no doubt the same. As tlni may pos- sibly have sig-nified the complete numeration as three in Ncng-one, and have ad\anced to ten in Fiji, and even to ten thousand in Maori, so tale may have signified the end of the counting- when no number beyond ten was counted, and have retained the meaning- of ten in Savo, while it has been ad- vanced as numeration improved to signify one hundred in Torres Islands. Many means more in a later generation than in an earlier : the Lakona gapra ten, means nothing- but ' many ;' tar, which in some languages is vaguely many, is in one a hundred, in several a thousand. Hundred. — There appears a quinary method of expressing a hundred in the Eromanga naroUm-?iarolhn, two-five-two-five, i. e. ten tens. The vig-esimal of the Loyalty Islands is se dongo re ngome of Neng-one, five men. The most common word in use in Melanesia, as in Polynesia, is ran a branch or leaf. The explanation of this use is to be had from the meaning of another expression used to signify a hundred in the Banks' Islands. In Mota this is mt?/ not, i. e. a whole vaele, the me/e being a kind of cycas. To count the days after a death a mele frond was taken, and beginning on one side of it a leaflet was counted for each day, one being pinched dowTi as a tally for every tenth. The frond when treated in this way on both sides furnished tallies for a hundred, and the final death-feast was commonly held on the hundredth day ; the whole xaele, vael not, was used and done with. The same practice is found in the Solomon Islands, where, in Ulawa and San Cristoval, not the simj)le rau but tanarau is the word in use. It is plain that this corresponds to their word tana/turtc ten, the same as the sanavul/c of other tongues ; that is to say the word for hundred means the double frond, counted on both sides, as the word for ten means the double handful of fingers. In Florida the word is //anatatu, 250 Melanesian Languages. in which Jiana has the same meaning of double, whatever latu may be ^. In the Torres Islands when hundreds are named they are called tale, the word used in Savo for ten, but in counting up to a hundred the hundred when reached is na won, the close, or completion. Thousand. — As high numbers are reached there is no doubt an increasing vagueness in their application, yet there can be no doubt but that Melanesians count with accuracy thousands of bananas, yams, and cocoanuts for feasts. The indefiniteness is shown in the word tar, which in the Banks' Islands is used for a thousand and also for very many, the same being a hun- dred in Espiritu Santo. In Nengone to count a thousand was to go as far as could be reached, e dongo, finish. The Fiji ndohi, thousand (the same word as nol in xael nol above), means all, complete. In Wa?^o of San Cristoval they have no word for a thousand. There is a word in use in Florida and Bugotu, mola, which is used indefinitely for a great number beyond count ; and this, but doubtfully, is given in Malanta and Ulawa for a thousand. To go accurately beyond a thousand is not commonly pos- sible, except as two or three or so many thousand ; if there be a word said to mean ten thousand a certain indefiniteness hangs about it. If the Malagasy alina means ten thousand, the meaning of the word is still ' night,' and there is a certain absurdity in saying alina roa ' two nights,' for twenty thou- sand, using a word for a certain number which denies the possibility of counting. In the Banks' Islands tar mataqela- qela is literally ' eye-blind thousand,' many beyond count. Figurative expressions show how the unpractised mind fails to rise to exactness in high numbers. In Torres Islands they use dor paka banyan roots, for very many beyond count, at Vaturawa rati na liai leaves of tree ; in Malanta they exclaim warehune huto I opossum's hairs ! idumie one ! count the sand 1 * How the meaning of a word disappears from men's minds in use is shown by the Bugotu form of this word, hafhanatu ; the syllables, as is very common, are transposed, and the meaning of the parts is lost. Nimieratioti. 251 In Fiji, however, the name of a tally like vatn loa, a black stone, no doubt is used with a definite number in view, thoug-h a number so larg'e as one hundred thousand is g-iven, and while yet oha is said to be used indellnitcly for a lower number as well as for ten thousand. In the same language vetelei, woka~ niu, are given for a million. VT. GRAMMABS. I. Banks Islands. The languages of the Banks' Islands are given the first place because that of one of them, Mota, much better known than any other to the compiler, has been the medium through which, generally speaking, information concerning the Melanesian lan- guages has been obtained. j\Iota has thus been a kind of standard to which the others have, more or less, been found or made to approach ; natives of other islands knowing Mota have explained the uses of their own languages with reference to it. The Gi'oup consists of eight islands, lying about the 14th parallel of South Latitude, and between 167° and 169° East Longitude \ The Islands are — (i) Vanua Lava, the largest. Great Banks' Islands ; (2) Santa Maria, (3) Saddle Island, (4) Sugarloaf Island, J/ota, (5) Ureparapara, Bligh Island, (6) Rowa, (7) J/erlav, Star Island, (8) J/erig, Sainte Claii'e Island^. Charts show in equal pro- minence a rock. Vat Ganai, misspelt Vatu Ehaudi. The languages began to be known to Europeans in the year 1858, and were first acquired and written by Bishop Patteson, the Rev. Lonsdale Pritt, and the Rev. John Palmer, of the Melanesian Mission. There is great difference between the languages, though there are many dialects where the difference is not so great as to prevent those who speak them from readily understanding one another. The Banks' Islands languages generally are closely allied to those of ' The Group was discovered and named by Bligh in his boat-voyage to Timor, after having been cast adi-ift by the mutineers of the Bounty. ^ The natives used another set of names when sailing between the Islands : Mota was Ure-kor, the place full of dried breadfruit ; Vanua Lava Ure-qauro ; Saddle Island Ure-wari; Santa Maria Ui'e-fiqalano ; places of different kinds of yams; Ureparapara t-Ve «*, the place of bows ; Merlav Urekere, the place of clubs ; Merig Ure-gave, the place of crabs ; Qakea, an islet of Vanua Lava, TJre-pug, the place of debt, because the shell of wliich money is made is abundant there ; Eavewa, another islet, Ure-marete, the place of a kind of holothuria. N? 1. \k' TORRES Ii>s Vat Garwui ' URE PARAPAH.A ^ SANTA MARIA \JjXX3JjCU 9 ME RIG- AND i^MERLAV I'^miRIES Iff /s K HEBRIDE S ^^«c A Iheh'uujtmix Mzlane.^ijixnj Lonuiuxues CZcureYuI .^t, Presx (Xrihixi. Mota. 253 the Noiihern New Hebrides, and are not far removed from the Fijian. Indeed, a ^lota man finds it easier to learn to speak Fiji than the language of ilotlav, close to his own island. Of the eight islands, Merig is the only one, being extremely small and lying between Merlav and Santa IMaria, that has no dialect of its own. Rowa is as small, but has its own dialect, with a peculiar mincing pronunciation. 1. Mota. Sugakloaf Island. The name of the island is Moto, with the nasal m : from the common practice of the Melanesians of prefixing a Preposition to the name of an island (see p. 162), it was at first known as Aumota. The island is very small, and its language has become important only by the accident of being used as the common language in the Melanesian Mission. Phonetic Character. — The languag e is geneea lly vocalic, though it does not I'eject close syllables and combinations of consonants of some little difficulty, such as sonnag, tarnag. There is a certain tendency towards throwing out vowels when suffixes are adcled^ as tarnag for tarauag, vapteg for vajmteg ; in reduplications, as gikig- lala for gilagilala ; and in names of persons and places. Accent. — It camiot be said that the accent usually falls on anyone SA'lloble, except in words of two syllables, in which the stress is on_ tlie fii'>t. It is certain that the incidence of the accent on one syllable does not, as in English, obscure the vowels in other syllables ; if the accent is on the first syllable in manigiu it does not prevent the full sounding of the succeeding i. In compound words, and words with affixes, the members of the compound, or the stems, retain their natural accent ; tdno viatitr a sleeping place, gasdl a knife, mawora broken, mavmravag break with. In fact, the accent falling on the latter syllable is a safe guide to the character of the words — ga-sal from sal to cut, ma-tur from the prefix ma and the root represented by the Malagasy turi. The accent in such words as nasasdna, ravevena falls distinctly on the penultimate, for the reason that sasa, veve are redujplications of sa and ve. Sometimes, it is true, in a long word, or in a clause sounding like one long word, the Accent in native speech will fall on some syllable on which there is no apparent reason for its resting, iloneia, pal2)aldtevat. Unless the word be one in which a dissyllabic suffix like iit is present, it may be said that the accent 2 54 Melanesian Grammars. can never be cast far back ; mdnig is the word, and with the ter- mination it remains mdnigiu, but idus ulusiu. Dialects. — There are two well-defined dialects on the island, one sympathizing with the neighbouring Motlav, the other with more distant Merlav. There are not many words of vocabulary dis- tinct ; the dijBference consists chiefly in the preference of u on the leeward side and i on the other, and the frequent substitution on the leeward side of w for g ; as tawur behind, in Veverau is tagir in the other dialect. It would have been well if one of these dialects had been in the first place chosen and followed, that of the leeward side by preference ; but the two have long been hopelessly confused in the speech and writing of foreigners. There is, besides what amounts to distinction of dialect, a variety in the way of pronunciation, which is paralleled in the neighbouring islands. The people of Veverau on the leeward side think that the Tasmate people, who are nearest to Merlav, speak thick, matoUol, and that the Maligo people, on the other side of them, nearest to Motlav, speak thin, mavinvin. The Maligo people say that the Luwai people, again, on the windward side, speak thick. The way of speaking, therefore, goes to some extent with the dialect, Veverau and Tasmate using one dialect, and Maligo and Luwai the other, and Tasmate tliinking that Luwai speaks thin as much as Maligo tliinks Tasmate thick. These are districts lying round the island. The Veverau peoj)le also are more inclined to introduce an euj)honic i, maros-i ava, for maros ava. With reference to dialect tlie Veverau people are called by the neighbours who speak ' thin ' ira we nao, and they call their neighbours ira tee talc, those respectively who say na and taJc. Some of the words in which they differ are these : — Ira we nao. Ira we tak. oia. ge do. mule, van. va, vano go. ttir. lira stand. rap. vega climb. le. la give. leo. lea\ un. ima drink. gang an. ganagana eat. gasavai. gasci how. vavine. tavine woman. sasae. sea different. * This word is hardly to be translated in one English word ; it may be law, custom, ways, speech. It ia the same with the Maori reo. Mota. 255 Song Dialect. — The Songs are always in a Dialect different from what is spoken, resembling the language of Gaua, Santa Maria, but not identical with it. Examples will be given below. Un words. — Quite distinct fi'om the words which differ in dialect are those which are used to take the place of such as form part or the whole of the names of relations by marriage. For example, one whose son-in-law, father-in-law, brother-in-law, is named Pan- tutun, hot hand, could not use the word 2)anei for hand or tutun for hot, but would have to substitute others. These substitutes are either common words used in an unusual way, as a knife may be called a cutter, or a bow a shooter ; teveteve for yasal, or venevene for us ; or as paito a shed may be used for ima a house ; or else words not commonly used in the language except under these circum- stances. These words, again, are either some common in neigh- bouring islands, as livna for hand, or else such as are only known in this use. To use a word in this way, in place of one which it is not correct to speak, is called to 'w?i.' A list of some of these words is subjoined. Un Wobds. limai for panel hand. liwv, for pel water. wanea >> som money. Jcakae nam yam. manarai )> j> >> sill ill qeta arum. repes „ pug debt. toqoii ta\s.a bag. Icarwae „ qoe pig. varea vanua place. sogae „ „ marapun vat stone. mawega „ villa moon. sasaqo loa sun. reivu „ ^vena rain. gire tuvag sell. samali ,, >j » tvana Vila lightning. tama „ galao left-handed. tatar vilog umbrella. molemole » Zan wind. saproro mate die. nanarag >> esu live. tignag ima drink. raio „ sur sing. niiv tagai no. raicraw „ as song. surata maran light. In Mota val is to put things one against another, answering to one another, the same word with valid to answei-, Malagasy vali. There is a way of counting when they wish to un ; ' one ' is val instead of tuwale ; ' two ' is tana valuna, i.e. 'and his fellow,' answering; 'three' is valuava, missing the match, the odd one; 'four' is valvalwia, match well, the reduplication re- ferring to the double pair. For ' ten ' the word is vawonot, va-ivono-t, ' make it complete.' I. Alphabet. The Vowels are a, e, i, o, u. These have the proper sound, not the English. There is a longer and shorter pronunciation of a, e, i, and on the broader or narrower a depends mostly the 256 Melanesian Grammars. tliickness or thinness of speech mentioned above. There is no shortened u. There are two sounds of o, the one approaching the English n in ' pull,' but o not u, and the other a simple o. The variation of sound depends, in fact, on the syllable being open or closed, no approach to u being heard in an open syllable ; for example, tollu an egg is to-li-u, but in composition tol manu, "bird's egg, makes a close syllable, which is in danger of being written tul man, because the sound of o is modified. For this reason there is no need for marking the change of sound by a change of sign. When the language waa first written many words were spelt with u which should have had o ; after- wards some were spelt with o that should have u. No native doubts which is right. At the end of a word u is often faint, often silent ; a word therefore may be written manu or man. Diphthongs are ai, ae, ao, an, clearly distinct one from the other. That these are diphthongs is shown by the difference of pronunciation between words like vagaus and ga-us, tinaena and naesuna ; where the word is compounded, ga-us bow string, na-esu the life, a break occurs between the two vowels which is not made when the vowels follow one another in the ordinary way. Examples of the distinction between au and ao, ai, and ae, are sau to lift uj), sao to take up a net (from both of which saiv, sawu, to blow, is distinct) ; gai, the conjunction ' until ' or an exclamation, gae a string. ^^ , The Consonants are k, g ; t ; jd, v, w ;fq ; m, m, n, n ; r, 1 ; s. The sound of g has been described (page 204^. In the Veveran dialect, on the leeside of the island, g at the end of a word after a is pronounced i, wurvai for ivurvag. The Mota t is blunter than the English. It is sometimes introduced to avoid the hiatus between two vowels, as before the suffix ia, me ge raJcatia sage, raised him up, for raka-ia. The Mota p is not so sharp as the English ; v approaches nearer to b : lava was at first written laba. A syllable is distinctly closed with w, e.g. mawmawui, to work, is a re- duplication of mawui, ma-u'ui, and the consonant from the second syllable ia taken into the reduplication with ita consonantal value, making a sound dis- tinct from man. The compound sound represented by q is kpw, though k is sometimes so obscure as to be missed, as in goqo to boil, tuqei a garden ; yet some, especially children, not able to pronounce both guttural and labial, will leave out p, and say tukwei. The nasal m is certain in the words in which it has ita place ; it does not vary with individual speakers. The old spelling of the name of the island Aumota for a lAola is instructive as to the sound. The following words are some of those that differ in sense according to the sound of m and m : ima to drink, ima a house ; tama father, tania as ; mera dawn, niera boy ; lama sea, lama to drum ; nom to think, «om thy ; mala a hawk, vaala a sow ; mata an eye, mala a snake ; manig cause, manig to dive. In printing for native use it is not worth while to use m as well as m ; natives will not go wrong in pro- Mot a. Articles. 257 nunciation. European students, for whose benefit m is used, will always be wise to mark the m in books and manuscript for themselves. By a native a syllable is closed with m, but mom is not very easy for a European to say. The trill of r and 1 is greater than in English : r cannot be pronounced after n and 1 without the intervention of d or u. Some will pronounce pulrua, two together, pnldrtta, some piditriia ; after n, u is not introduced. A true Mota man cannot pronounce h. II. Articles'. The ^lota Articles are o, na. Demonstrative^ and i, Pergonal. 1. There is no difference in nieaning between o and na; both answer to the English ' a ' and ' the,' but are in the native mind probably definite. The difference is that na is always used before the Noun to which a personal Pronoun is suffixed, and only with that construction. It is the practice to write it in one with the word so formed : o panei a hand, but na2)anena liis hand. Before terms of relationship na is not used ; tasik not nafasik my brother. The Article o is commonly used with Nouns where nothing very definite, perhaps, is in view, but something is indicated or present to the mind, for o no doubt is a demonstrative Particle. But when the notion is very general the Article is not present ; to catch a fish is rave o iga, to catch fish rave iga ; to shoot a bird vene o mnnu, to shoot birds venevene manu. 2. Thei'e is no distinction of Number ; o ira.a the house, o ima nan the houses ; nnimnJ:. luninal: iir/n, my house, my houses. 3. With names of places o ib used ; o Mo^a we, asau nan o Gaua, Mota is distant from Gaua. 4. The Personal Article is ^, which is used with personal names, native and foreign, male and female ; i Saravna, i George, i Rota- viro, i Sara. There is no need to use this Article, but as names of persons are generally words in common use, and i marks the word as a Proper Name, it is usually employed. 5. This Article applied to a word makes it at once a Proper name ; it has therefore the power not only of showing a word to be a personal name but of personifying the notion conveyed by the word. Thus gale to deceive, i gale the deceiver ; esu to live, i Esu ' Words are not divisible into Parts of Speech as distinctly as they are, for example, in Latin. The same word may be used as almost any l^art of Speech^, ^without change uf form. Nuvertholuss, tlicrc uv^' wurds wliicli cannut l>e so used, and there are certain changes of form whicli belong to words used in various grammatical characters, as well as certain ways of using them as different Parts of Speech. It is therefore convenient to divide words in the customary maimer, and to arrange the Grammar accordingly. S 258 Melanesian Grammars. the Living one ; ganganor wickedness, i Ganganor the Wicked one ; i Vavae the Word ; i Vaesu the Saviour. This, however, can only be properly done when a title or special appellation is in view, when a capital letter would be appropriate ; the Sower is rightly i Savsavur, but it will not do to translate a Greek Participle with the Article in this way, unless a sort of title is given. See John iv. 36. A special characteristic is thus designated: i Tanaro GUag'dala, ineia i gilaJa ape savasava n«n nanra iatasina, he was the one who knew about all sorts of things more than his brothers, i.e. he was the knowing one. It is a singular use of this Article by which i gopae is a sick man ; gopae sickness, gopa to be sick. 6. The Personal Article can take a Plural form with the Plural sign ra^ ira Tinqoro the Disciples ; and perhaps with a wider use than in the Singular, ira naro the widowers, though not widowers in general but the widowers of the place. 7. Wlien a native name is that of a female i is applied to it, but another Particle ro is prefixed, which shows the name to be feminine. This ro is not an Article, but combines with i to make the feminine Personal Article iro : i Taviro is a man's name, iro Taviro, or i Botaviro, is a woman's. This ro does not apply to foreign names, English or of other islands, which are only known as personal names ; a girl named Anne is / An, not iro An, a woman of a distant island i Ono. But if a foreign name is that of a known object, like a boat, it is tro Pout. A girl was called i Nas after a nurse, but the name was taken as a personal one. A mother is spoken of as i Veve, not iro Veve, and i sogoma is your relation by the mother's side without reference to sex. 8. The Personal Article in the feminine form personifies : iro Gale the female Deceiver, the Avoman whose title is Deceiver, iro Maranaga the Queen ; if the English word is used it is i Qin, not iro. The Plural form is iraro, with the same latitude as above, iraro naro the widows of the place. 9. In consequence of personal names having a meaning, meaning some thing, the Personal Article with the word meaning ' thing ' means a Person, or interrogatively with the word meaning ' what ' asks who is the Person : but always with reference to the name, not to the Person. Thus gene thing, gene the thing, but i gene the Person, ira gene the Persons ; i^'o gene the female Person : sava what ? i sava ? who ■? what man 1 iro sava 1 what female 1 The word gene does not call a man a thing, or iro sava ask in an uncompli- mentary manner what thing the woman is ; it is saying ' So-and-so ' instead of a name, asking 'what's her name?' not 'who is she?' See Pronoun sei. Mot a. Nouns. 259 10. Animals being personal enough to have names, their names take Personal Articles ; horses are i Bob, and iro Vitu, Star ; the cat of the house is i Pus, any cat pus. This is not old Mota, though correct and useful to illustrate the use. Pigs only had names in old Mota, and their names, like the shortened names of men, began with Wo, not i; but a dog in Mota now is i Pap, Bob. 11. The names of men and boys are often cut short, and the first syllable or two used with the prefix TFo; Wolig for Ligtarqoe, WogaU for Galepasoqoe. The Personal Article is not so commonly used with this. III. Nouns. 1 . There are, as in Melanesian languages generally, two di\dsions \ of Nouiis Substantive, viz. those that take the Personal Pronoun j suffixed (with the Article na), and those that do not. This division I is properly exhaustive. The principle of the division appears to lie in a nearer or more remote connection between a thing and the possessor of it. Parts or members of a body or organization, the ordinary equijiment and properties of a man, belong to the class the names of which take a Pronoun suffixed to show possession. Nouns of the other class are used with a Possessive sign to be hereafter explained. The application of this^ principle is not always clear, though it caij be applied without forcing the sense of it : a man's bag is natanana, na Article, tana bag, na third person Pronoun suflBxed, but his basket is non o gete, non his (in which, however, no is really a Noun and n the suffixed Pronoun), o the Article, gete bag ; a man's bow is na-vsu-na, his paddle non o icose. The bow and the bag are looked upon as closer appendages of the man than his paddle and his basket, being always in his hand or on his shoulder. With Compound Nouns the last member determines the class in this respect : na-ga-usu-na his bow-string, kere tvose anona the end of his paddlej non tano-togatoga his abiding place. It is possible that words belonging to the class which takes the suffix may be used as if belonging to the other class; but this can only be the case when they are used in a secondary sense, or when there is a purposely marked difference in the kind of possession indicated. Thus pane is not only an arm but an armlet ; in the primary sense my arm is na panek, in the secondary nok o pane my armlet. The latter difference but rarely occurs ; laok 2>UQ a debt owed to me, napuguk a debt that I owe. In Hazlewood's Fiji Dictionary the words that take the suffix are marked. It is of great importance that the words of this class should be accurately observed in all the Melanesian languages. Mr. Fison gives the distinction in Fiji of uluqu my head, and noqi( ulu the head I have for sale. See Duke of York. S 2 26o Melanesian Grammars. 2. Another division, but not cue of equal importance, can be made, of Nouns which have and have not a special termination as such. Very many words are Noun, Verb, or other part of speech, without any change of form ; but there are others which by their termination may be known to be Nouns Substantive, at any rate when they stand uncompouuded. Of these which have substantival terminations, many are Nouns belonging to the class which take the suffixed Pronoun, names of things which are relative to some other things, not names of things which have an absolute existence of their own. The Verb qeteg, to begin, is the same word with qetegiu a beginning, but the latter has a substantival termination which marks it as a Noun. 3. Of the Nouns with no special termination it is not necessary to say anything. Those that have such a termination may be divided into Verbal Nouns and Independent Nouns. 4. The Verbal Nouns are words which give in form as a Noun the abstract sense which is conveyed by the Verb, with a special termination added to the Verb. The terminations in ]\Iota are a, ia, ga, ra, va. Thus mate to die, matea death, nonom to think, nonomia thought, vano to go, vanoga a going, toga to abide, togara way of life, tajpe to love, ta2)eva love. There is no difference of signification according to the difference of termina- tion : mule as well as vano is to go, and nmleva is a going as well as vanoga. But a verb may assume two terminations, and make two nouns with a differ- ence of meaning ; as toga with ra is togara behaviour, and with va is togava station. A native thus defines the two worfls : o togara, we toga iaraa uvea, togava, loe toga avea, how one abides and where one abides. We may compare the English suffixes -ness, -hood, and the difference between hardiness and hardihood. Of these words matea alone takes the sufiixed Pronoun ; namateana his death, but nok o noyiomia my thought, nolc vanoga, togara, tapeva. 5. The Independent Substantives are so called because these are names of parts, members, things in relation to something which possesses or includes them, but are by tliis special termination shown to be in thought and in grammar free for the time from this dependence. Thus an eye is in the true form mata, as in so many languages, and when any one's eye is spoken of, since the word belongs to that class which is mostly made iip of this kind of Nouns, and has the Pronoun suffixed, this is the stem to which the suffix is applied ; namatana his eye. But if an eye is spoken of independently of any person, or any organization, the word assumes the termination i_ and becomes matai not mata. In Mota. Nouns. 261 thought the eye is independent, not viewed as a member ; in grammar the word is independent, not constructed; it assumes therefore the termination that marks it as such. The termination, when the radical ends in a vowel, is t, when it ends in a consonant iu or ?«", according to dialect : sasa-i a name, iuqe-i a garden, roro-i report, ulu-i hair; qeleg-iu beginning, qat-iu or qat-ui a head. There are words which have two forms, panei, a hand, &n(\ paniu ; the one from pane, the other from pan. The same account must be given of the word qarana a hole ; this would make qaranana, but o gasuioe we toga alo qaranina, a rat stays in its hole, there must be a fomi qaraniu ; so tavaliu is a side, but tavala pei, the other side the water, shows a form tavalai. There are Nouns with the termination e, vavae a word, gae a string, tinae bowels, gopae sickness, which have much in common with the Nouns ending in /, inasmuch as the radical is vava, ga, Una ; but this e is different from i. In the case of gae this is shown by the addition of i, making gaei, when a bunch of bananas is spoken of. The word tinae, bowels, takes the suffixed Pronoun without modification, tinaek, but Una as in tine vanua, the middle of the place, is the same word, though in an independent form it is tinai. 6. Words of this kind, those that take the terminations i, iu, ui, form compounds with other Nouns; but, since it is the true original form of the word which is the element compounded, the terminations never ajDpear iu composition. Where the tx'ue word ends in a consonant the composition is simple ; qatui, a head, in- dependently, qat the true word, qat qoe a pig's head, in construction : the termination is not dropped, for it has never been assumed. Where the true word ends in e or composition makes no change : tuqe a garden {tuqei independent!}'), tuqe sinaga a garden of food, roro, roroi a report, 7'0)'o vagalo a report of fighting. In the case of a word like ului hair, which when in composition with another is ill, it may be rather thought that ulu is the true form of the word, and that u is dropped in ul qoe pig's hair. Some, however, would say ulu qoe. It is common to introduce an euphonic / before a vowel, qati-alca i^ot qataka, the fore part of a canoe. Wliere the true word ends in a this termination in composition becomes e; sasai a name independently, sasa the true word and the stem to which a Pronoun is suffixed, na-sasa-na his name, sase tanun a man's name, a lightened in composition. 7. For further consideration of this subject the character of a word, whether it takes an independent termination or not, whether it takes a Pronoun suffixed or not, must be dismissed : what are to be kept in view are Nouns ending in a, and the fact that these 262 Melanesian Graimna7's. when compounded with another Noun, as the formei' of the two, change a to e. Sasai a name {na-sasa-na his name showing true fonn sasa), sase tanun a man's name ; ima a house, ime tanun a man's house (a word that takes the suffix, naimana his house) ; sinaga food (which cannot take the suffixed Pronoun), sinage tanun man's food. These words are said to be compounded together because of this modification of the vowel : the relation of possession between them is close, so that the idea and the word may be called compound. Hence the first member of the com- pound takes a lighter termination. There is an appearance of inflexion, but no true inflexion. It should be observed again that there is no case of composition where two Nouns are together and the second qualifies the first in the way of an Adjective : ima vat a stone house, ima vtd a spiritual house, difi'erent from ime vtii the house of a spirit. This is not always clear, partly because of the English idiom ; a house of prayer is one of that character that it is used for prayer, not one of which prayer is the owner or inhabitant; it should not therefore be ime tataro but ?'ma taiaro. The same modification of a to e takes place also, but not often, where there is no relation of a possessive kind, and where the second word does not qualify the first : o moegene, the first or principal thing, moai first, moa, moe, gene thing. One word seems an exception, in which an becomes o : naui a leaf, nan the true word by analogy, but no tangae the leaf of a tree. The account of this probably is that nau — rau being shortened into no in neighbouring dialects, has been taken up in Mota. 8. Prefixes to Nouns. — There are a few words in Mota, coi'- responding to a great many in Fiji, which are formed from Verbs by prefixing ^ : jpala to take up as with tongs, ipala tongs ; sar to pierce, isar a spear ; ras to bale, iras a baler ; got to cut, igot a cutter ; goso to husk cocoa-nuts, igoso a stick for the purj)ose ; lano to put rollei's under, ilano a roller. Another Prefix which makes Nouns of Verbs is ga : sal to cut with a drawing motion, gasal a knife ; nor to bear a grudge, ganor malice ; qisan to press down, gnqisan a weight ; pulut to stick together, ga2)ulut glue, paint. Such examples as these show that this prefix cannot well be ga from gae a string or bond, such as appears in garotrot a tie, from gae and rot to tie ; gatog\i the back- bone, gae and togoi vertebra, 9. Verbs are often used as Nouns without any change of form, or rather words are Noun or Verb indifferently. But a Verb used as a Noun will very often be reduplicated, rave to write, o rave- rave a writing. Aloia. Nouns. 263 10. Eedujdication^in Mota is either Jj) of the wliole word, vat a stone, vatvat stones, soasoai members; or (2) of the first syllable, (janor a malicious feeling, gagaganor malice as a characteristic quality ; or (3) of the first syllable closed by the consonant suc- ceeding it, rawoi a leg, raiwawoi many or great legs. The effect of reduplication is with Nouns to express number and size, and with Verbs, and consequently with Verbs used as Nouns, con- tinuance and repetition. Reduplication with the close syllable rather expresses number and size, and intensifies or exaggerates the notion of the word ; pisjnsui fingers, ranranoi legs, gate ran- rauona ! what big legs he has ! sulatalamoa jnspisui tve qoqo, the centijjede has many legs ; ganor malice, ganganor wickedness. Rcdiqilication in the case of the name of a plant signifies that it is wild or useless ; matig a cocoa-nut, metigtig a wild palm, qeta the esculent caladium, qetaqeta wild caladium. Compare Florida and Duke of York. 'When a reduplicated word becomes the first part of a compound, if the termination be a which, as above, changes to e, the reduplication is of the word as so changed, not of the true word. Thus soasoai, members, is the re- duplicated form oi soai, the reduplication signifying multiplicity, and the true word is soa : o soasoai members generally, na-soasoa-na his members, but o soesoe aka the component parts of a canoe ; siiia to shine, sinasinai a shining, o sinesine loa sun-shine. 11. Plvrnl, — The plural of Nouns is marked in three ways: (i) l)y iidupliLation as al)ove, (2) by the addition of a plural sign, and (3) by prefixing a particle. (2) The plui'al sign in common use is nan, which follows the Noun ; ima a house, iva.a nan houses. It may be presumed that this is in fact a Noun meaning a collection or multitude, but there is nothing to prove it to be so. It is sometimes sepa- rated from the Noun and placed after the Verb, qon we tvesu nan, days are coming, the plurality perhaps being extended to the Verb. In lama we reve nan ilo vanua it is plain that nan has not a merely plural meaning, the sea in many places runs up into the land ; the sea does not appear before the mind as one body of water but many. Another word is known and used, but seldom, at Mota, taure, the Vureas tore, taure ima houses, a collection of houses : the word is plainly a collective Noun. (3) The particles prefixed to mark j)lurality are re, ra, the latter of which commonly forms part of plural Pronouns. The use is only with words which describe persons with regard to age and re- 264 Melanesian Grammars. lationship : tasixL brother or sister, o retatasiu the set of brothers or sisters, ratatasik my brothers ; retutuai the set of sisters or brothers, ra tuiuak my sisters ; re tamtamai the fathers, the men of the generation above, ra tamak my fathers ; raveve mothers, and with a singular sense, mother ; rerelumagav the young men, o reremera the boys ; ira qaliga relations by marriage ; vaereata a male, rereata the men-folk, tavine, or vavine, a female, reta- vine, or revavine, the women-folk ; ira tamtamaragai the old men ; retawu the strangers. The use of these Nouns in Mota is pecu- liar, and from a point of view other than grammatical very interest- ing (see Vocabulary No. 43). It is to be observed that re is used when the whole class of persons is spoken of, ra when a certain number only are in view, retawu the body of foreigners, ira tawic the foreigners ; a man out of his own country being a tauni. The word ra is used also in what looks like the position of a Noun, ra ta Motalava the Motalava peoj^le ; where, however, ta being really a noun, as will be shown, the construction is the same. In these lan^ages the words ' brother ' and ' sister ' are used with refer- ence to the sex of the person relaiionship to tohom is in view: tasiu, tiduai is brother or sister as the case may be, if of the same sex tasiu, if of the other tutuai. A man's brother, a woman's sister, is tasiu ; a man's sister, a woman's brother, is tutuai. In some of these words there is reduplication to mark plurality, vaera is a boy, rereniera properly boys, but, like raveve mother, which is properly plural, used as singular, boy. The singular is mereata, the plural rereata ; mere is probably the same word with niera. VThen the Personal Article /, or, with the feminine sign, iro, becomes plural ira or iraro, it is this ra which is added. It is also an idiom to use ira, or ra, before a person's name to signify that person and his company, or the companions or people of the person : ira Bishop the Bishop's people, ira Wo- tcufris the gang at work with Wowutris at their head. In cases where a simple plural would be enough in English it is often idiomatic in Mota to use expressions which mean ' all kinds ' and ' every,' sale and val : o sale gopae sicknesses, all kinds, val gopae sicknesses, every sick- ness. These are combined and nan is added, o val sale gopae nan, sicknesses of all and every kind. 12. Since in a_lano;uage of this kind there is no Grammatical Gender, it is idle to say that lumagav, a young man, or qoe, a boar, are masculine, and malauiala, a girl, ma^rt, a sow, feminine. "\Mien it is desired to signify sex, mereata male, and tavine or vavine female, are added to qualify as Adjectives. Mota. Pronouns. 265 IV. Pronouns. 1. Personal Pronouns. — There are in Mota two distinct sets of Personal Pronouns, (i) those whicli are used as the subject or object of the Verb, and (2) those which are suffixed to a Noun Substantive. (i) Singular, i. inau, nau, na. ~ 2. iniko, ko, ka. 3. ineia, neia, ni, a. Plural. I. inclusive, inina, uina. exclusive, ikainam, kamam. 2. ikamiu, kamiu, kam. 3. ineira, neira, ira, ra. Dual. I. inclusive, inarua, narua, inara, nara. exclusive, ikarua, karua, ikara, kara. 2. ikamurua, kamurua, kamrua, kamra. 3. irarua^ rarua, irara, rara. Trial. i, inclicsive, inatol, natol. exclusive, ikatol, katol. 2. ikamtol, kamtol. 3. iratol, ratol. Observations. 1. For the probable composition of these words from the Personal Article t, a demonstrative 71 or k, and the true Pronoun w, ko, a in the singular, na, mam, miu, ra in the plural, see Comparative Grammar, page 116. 2. The use or omission of the Prefix * has something to do with the greater or less directness with which the person is indicated, but probably is often without any other reason than the caprice or convenience of the speaker. 3. Some of the forms are evidently shortened from the fuller, na from 7iau, ni from neia, ka from ko, kam for kamiu, and in the Dual na and ka, narua, nara, karua, kara, from nina and kamam ; but these shorter forms must not be taken generally as equivalent to the longer, ones. This may be so in the Dual nara, kara, kamra, but in the Singular and Plural it is not so. In the Singular and Plural the shorter forms, na, ka, ni, kam, can never be the object, but always are the subject of a Verb. There is again a distinction to be made among these ; na, ni, kam are used directly in an indicative sentence, ka is not, though one may ask ka ge sava ? where ka probably follows on an omitted si: na, ni are always used when the sentence is indirect, potential, subjunctive, optative, though they can also be used indicatively, as ka is not. For example, it is right to say na tee pute, ni tve pute, kam we pufe, I sit, he sits, ye sit; but it must be nau or na we pute si na rave, neia or ni we pute si ni or sin rave, I sit that I may write, he sits that he may write, na and ni, not nau and neia, in a subjoined clause ; and similarly ko ice pute si ka rave, thou sittest that thou mayest write: si ni 2 66 Mela7iesian Grammars. contracts to sin. So in the case of an optative or imperative sentence, na Ho let me see, ni muJe let him go, ka rave write thou. 4. It is evident that the Dual and Trial are not in fact more than the Plural _with the Numerals rua, tohi, two or three, suffixed; but inasmuch as both members of the compound thus made have been subject to change in the com- position, it is desirable to set them down as distinct persons. In karua the Pi'onoun kamam appears as ka, while the Numeral rua is entire ; in kara both parts, kamam and rua, are shortened to ka and ra. So nara, kamra, vara have ra for rua, and natol, katol, kamiol are for nina tol, kamam tol, kamiu tol. It should be understood withal that there is no true Dual or Trial as there is a true Dual in Nengone. It is necessary always to use the Dual and Trial when two or three persons are in view, never the Plural. The Dual is used in speaking of or to a single person when a near relation by marriage. 5. The third person plural ra presents some difficulty ; it is a Pronoun, but at the same time it is not always more than a plural personal sign. In an expression like ira Bishop, the Bishop and his companions or the Bishop's people, mentioned above, it is clear that ra is not a Pronoun but a plural sign added to the Personal Article. When o-a ta 3Iotalava, the Mota- lava people, is said, it may be questioned whether ra is not a Pronoun. When inanimate things are in view ra is perhaps never used. 6. The third singular a is never the subject, and only appears after a Verb or Preposition, and suffixed to it. 7. Suffixed forms of these Pronouns. — In Mota only the second and third singular and third plural are suffixed, in the forms ko, a, ra, to Verbs and some Prepositions. After a Consonant i, sometimes u, is introduced before the suffix ; nau we iloko, iloa, ilora, I see thee, him, them ; ni me rus-i-ko, vits-i-a, vus-i-ra, he struck thee, him, them ; mun-i-ko to thee, nan-i-a from him, *«;•- i-ra to them ; so palua, gaplotua, gapua : nanra, munra can be said without an intervening i by pronouncing d, nandra. When a. is suffixed to a word ending in a the euphonic i is introduced, laia for laa ; the same sometimes occurs after e, vus mateia kill him. To write these Pronouns as suffixes is not necessary, but comes naturally to the natives. The Mota language does not (like Florida, for example) repeat the object of the Verb as a suffixed Pronoun when the object has been already expressed. It dislikes the suffix of the third singular a to a Verb except when a person is spoken of; si ko qe ilo o tanun Hone amaira gaganag luea ma mun nau, if you should see that man with them point him out to me ; si namaiama qe ge iniko si ka futuag, toakele lue, savrag naniko, if thine eye should make thee to stumble, pull (it) out, cast (it) from thee. 8. Since there is no Gender, the third person singular is he, she, him, her, it, in English, as the case may be. But there is a certain dislike to using the Pronoun for inanimate objects. Y 2. Pronouns suffixed to Nouns. Singular, i. k. 2. wa, m. 3. na, n. Flural. I. nina, incl. 2. miu. 3. ra, r, mam, excl. It will be observed that the Plural forms are those of the Mota. Pronotins. 267 ordinary Personal Pronouns; the Singular consists of a perfectly distinct Pronoun. These Pronouns are suffixed only to one particular class of Nouns already described, giving a possessive sense ; and also form part, as suffixes, of the words which take the place of Pronominal Adjectives or Possessive Pronouns in English. Example — o panei, a hand. Singular, i. napanek, my hand. 2. napaneiwa, thy hand. 3. napanena, his hand. Plv/ral. I. napanenina, our hand, inclusive. napanemam, „ ,, exclusive. 2. napanemiu, your hand. 3. napanera, their hand. Dual. I. napanenara, hand of us two, inclusive. napanenkara, hand of us two, exclusive. 2. napanemurua, hand of you two. 3. napanerara, or -nenrara, hand of them two. Trial. i. napanenatol, hand of us three, inclusive. napanenkatol, hand of us three, exclusive. 2 . napanemtol, hand of you three. 3. napaneratol, hand of them three. Ohservations. 1. The first person singular Tc is sometimes ku ; na of the third person is often n before another word ; as m« is m ; and ra is r. 2. The sufl&x «, as distinct from na, points to some definite person, or thing spoken of as if a person, not an inanimate object. Thus ni ice pute ape hikin tanun he sits by a man's side, but ape kiki ivaa beside a house ; o tete toe tako ape sus tavine a baby hangs at a woman's breast, generally, but ape susun ravevena at its mother's breast, particularly. 3. There is in Mota only one use of this Pronoun suffixed to a Preposition, apena, about it, concerning it. In other languages this use is common. 4. It may be seen that in the first Dual and Trial, exclusive, panenkara, panenkatol, and third Dual panenrara , there is n after the Noun and before the pronominal sufiix. Tliere can be little doubt but that this is itself the third singular Pronoun suflSxed, napanen his hand, kara we two, the hand of him and me. It should be observed that kara is idiomatically used where we should say he and I ; kara Saraioia we two Sarawia, i. e. Sarawia and I. So imanrara, the house of them two, is iman his house, rara two of them, his house and the other's. 5. There is often added to the sufiix k the syllable sa, napaneksa my hand, nagaksa my food : it cannot be explained in origin or purpose. 268 Melanesian Grammars. 3. Demonstrative Pronouns. Hoke this, Hone that ; iJce this, ine that. The plural sign nan added to these makes the equivalent to ' these ' and ' those.' The particles he and ne in these Pronouns are demonstrative. The fuller demonstratives, nake and nane, are often added : iloke nake, Hone nane. There is a difference between iloke and ike, Hone and ine : iloke. Hone, can be used as Adjectives, o tanun iloke, Hone, this or that man ; ike and ine can only be used as Substantives, with the Article, ike this, o ine that, or with the Personal Article, i ike this person, iro ike this woman ; plural o ike nan, o ine nan, these, those. Note that iloJce and Hone, and in the plural HoTce nan, Hone nan, are used as Demonstrative Pronouns, this, that, these, those, without any Noun, but can never be used with an Article. There is another Demonstrative Pronoun in the Plural, which has arisen from a Vocative particle. If a man is called to, the ex- clamation is yai ! an Exclamation and not a Pronoun ; but the plural of this with the plural particle ra, and with the personal Article i, ragai ! or iragai, is both an Exclamation and a Demon- strative Pronoun ; ragai ! you people ! iragai those persons, ragai ta Luwai those Luwai people, ni me la at mien ragai he gave it to those people. In the Dual and Trial there is a shortened form, ragera, ragetol, for ragai rua, ragai tol, those two, those three, or you two ! you three ! The Demonstratives nake, nane, pointing here and there, can hardly be called Pronouns ; they go with and repeat the signification of iloke, Hone, Hoke nake this here, Hone nane that there. See Adverbs of Place and Time. 4. Interrogative Pronouns. The words used are no doubt really Nouns, sei who ? sava what 1 The Personal Article makes isei who ? irosei what woman ? irasei, irarosei feminine, who 1 plural. With sava the definite Article is used, o sava what 1 o sava nan what ? plural. The word sei stands as a Pronoun in the place of a person's name, not of the person himself. If in English the question is ' who V the meaning is ' who is he ?' the person ; if in Mota the question is isei 1 the meaning is ' what is the name?' To ask a name is not o sava nasasana? what is his name, but isei nasasana ? tvho is his name ? On the other hand, sava asks concerning a thing; it is asked, if a person is in pain, nasavama we vivtiff ? your what hurts you? (compare Duke of York). It is asked concerning a relation, 7iasa- Mot a. Pronouns. 269 vataa Hone ? your what is he ? your father, brother, or what ? But persons' names being themselves taken from the names of things, the personal Article with sava makes that also an Interrogative of a personal name : i sava 1 who ? iro sava ? what woman? sava referring not to the person, as has been said, but to the thing the name of which has become a personal name. Often sava becomes sa, o sal what ? This becomes a sort of interrogative exclamation, sa ! si Team gate gilala 1 what ! do ye not know ? When it is a question as to which or whether of two things, it is the idiom not to use a Pronoun but an Adverb, avea ko toe maros ? which, literally where, do you like? But if it be a question as to persons the Pronoun is used, isei nan vara ? which of the two ? Another meaning of sava is ' what or any sort or kind,' o sava manii Hone ? what bird? i. e. what kind of bird? o tol sava inanu ? the egg of what bird ? Icam qe kalo pata ilo sava ima, if ye enter into a house of any kind. The reduplication savasava means many things, all things, everything ; non savasava all his things, o savasava nan all sorts of things. A verbal form is used, ive savai : gate gilala si o vat ive savai Hone, it is not known what sort of stone that is. 5. Indefinite Pronouns. Both sei and sava, being in fact Nouns, are used as Indefinite Pronouns, isei some one, whosoever, o sava some thing, any thing, irasei some people, o sava nan some things, o savasava nan any sorts of things. Some one is tuara sei; o tuara tanun a certain man. Some persons or things is tuaniu, tuan, a tuan tanim some men, o tuan ima nan some houses : tuaniu is by its form a Noun, and is so used alone, o tuan iga nan aniaia, i^a ni me la ma mun nau o tuaniu, he had some fish with him, and he gave me some; o tuanimiu some of you. There is a Verb tuan to help ; and a Noun tua, helper, companion ; i tuanira he their companion, he and some others, i tuamiu sei one of you, some one your companion. There is another Indefinite Pronoun in frequent use, tea some- thing, a woi'd the same as the numeral tea one. It is used in the sense of something, anytliing, at all, le ma tea give me some, si ta lax tea if it be at all possible, o sava tea anything whatever, na gate lav moJc tea I have not received anything at all. When translated by the English ' at all,' tea has the appearance of an Adverb, but it is grammatically a Pronoun in Mota, anything-at-all. The distributive Particla val expresses ' each,' val neira, valval neira, they each of Uiem, val tanun each man, val sei each and every one, valvanua each island, or an island in each part. There are no Relative Pronouns, and care must be taken lest the Indefinite be taken for a Relative. When a relative would be used 270 Melanesian Grammars. in English, the sense is conveyed in Mota by the use of the demonstrative ; the man whom you sent told me, i gene me gaganag, ko me vatrania ma ti, ' the person told, you had sent him hither,' Or two Verbs may combine, without a conjunction or a relative clause ; i gene me ilo me gaganag the man saw, told ; i gene me gaganag me ilo the man told, saw ; the meaning being, the man who saw told, the man who told saw. Or by the use of the In- definite Pronoun ; ni me gaganag munrasei me vatatua, he told those who met liim, literally he tqld whatsoever persons met him, those persons whoever they were. The demonstrative particle nane is useful to do the office of a relative in pointing back, like the English ' that ' ; tanun nane Hone ko me vusia, or tanun Hone ko me vusia nane, the man whom you struck ; the man, that one, you struck. V. POSSESSIVES. 1. These are not Pronouns, though these words, which take the place of Possessive Pronouns or Pronominal Adjectives in English, have always a personal Pronoun suffixed. It has been said (page 259) that one class of Nouns takes the personal Pronoun suffixed to the Noun, giving a possessive sense, napanek my hand ; and that the other class takes, generally, before the Nouns a word meaning 'my,' 'thy,' 'his,' &c. which is not a Pronoun, but must be called a Possessive. These Possessives consist in fact of a Noun meaning a thing belonging or possessed, a Possessive Noun, and of the suffixed Pronoun, which shows to what person and number of persons the thing belongs, as ' my,' ' thy,' * our,' do in English. Thus my knife is nok gasal, and nok is no-k, no thing-belongiug-to, k me. The fact that these words never actually occur without a suflfixed Pronoun gives them so much of the appearance of a Pronoun that without consideration it is not easy to recognise their true character. It is worth notice and record, as showing how these words appear to a native, that Wogale was disposed to think no tangae, a leaf of a tree, not made up of naui a leaf and tangae a tree, but of no and tangae, no being this stem to which the Pronouns are suffixed, no the thing -belonging-to tangae a tree, i.e. a leaf. It was more natural to him to regard the possessive stem as a common Noun than to admit the shortening of naui to no 2. These Possessive Nouns in Mota are four : no, mo, ga, ma. The difference between no and mo is that no means a thing that comes into possession from without, mo a thing coming from Mota. Possessives. 271 within, possession of which ratlicr follows on the action of the possessor ^ Tliis distinction is not perhaps always clearly main- tained, but this is the distinction, and it is of importance. There is a closer relation signified by ga, generally of food ; ma is always of a thing to drink. These Possessive Nouns then take the suffixed Pronouns like the other Nouns : nok is thing belonging to me, mom« thing belonging to thee and of thy doing, gana thing belonging to him for his eating, mara thing for their drinking. The compound is the Possessive made up of nominal stem and suffixed Pronoun, and generally precedes those Nouns which cannot take a suffixed Pro- noun themselves, nok wose my paddle, mom o vavae thy word, gana o nam his yam, mhnina o 2}ei our drinking-water. 3. These words, though going so often with Nouns, and qualifying them as 'my,' 'thy,' &c. do, have a purely substantive use, as ' mine,' ' yours/ ' ours,' in English, and take the Article accord- ingly : nonsei Hoke ? whose is this ? (no-n-sei thing-belonging-to- him who) nok mine ; or na nonsei 1 na nok ; nagaksa Hoke a tliing- for-me-to-eat this. 4. These Possessives have often the Prefix a : anok, amoma, agana, amanina. Although the best native authority makes this the Preposition a it may be doubted. In Florida, where there is no Preposition a, it is equally anigua, anina. In Mota also the same appears in other words, a avin sei ? a avik, whose fire ? my fire. It may be a Noun which survives also in the Preposition. WTien anok, am.ok, &c. are used, they generally follow the Noun ; nok, raok, &c. generally precede it. 5. Observations. 1. no. — The meaning of no being simply a thing-belonging in a general way, there is not much that needs explanation. It should be observed, however, that the translation of the word, when in the third person, is often made by the English preposition 'of ; o parapara non iamana his father's axe, or the axe of his father ; / pnhalana me ronotag non Qat o rararao his friend heard the crying of Qat. Care must be taken to keep the distinction clear; there is no 'of in Mota. It is necessary also to bear in mind that no cannot be used as an equivalent for 'my,' 'thy,' &c., unless the thing can be rightly spoken of as a kind of property; 'my father' cannot be nok mama. Juxta- position of two words conveys the possessive or genitive relation : it is not so idiomatic to say ima inait, my house, as i'mrt^, but it is right. Sometimes, as often in Motlav, ' with him' will be used for 'his,' o gasal amaia his knife. ^ ' A noma tavaa apeniko, we van ma, amoma ^ama ko me ge? Your no as if a thing with you that comes to you, your \\\o as if you had done it : a native explanation. 272 Melanesian Grammars. 2. mo. — It is common to iise mo^, mom«, &c., after a Verb with trie sense of 'for my part,' 'myself : si na ilo moTc let me see, nau qara iJo goro vi\ok rereuiera I am now for the first time looking after boys, tama ni me vet vciona as he said himself; anxonsei me ge save o siopa 1 who tore the garment ? whose doing was it ? In another way it is said, ko te ge momam you shall do it for us, as our agent ; I'sei te mnle vaomam ? who will go for us ? Again, in a way apparently inconsistent, one will write in a letter, nau we gaganag moma I tell to you, rather, I tell a piece of news for you. Since this word signifies a thing done by or proceeding from a person, it is conveniently used to translate an English passive participle ; at the end of a book ' Printed by A. Lobu, H. Silter and others,' it is Namora A. Lohti, S. Silter, &c. me qisan, i.e. the doing of A. L. &c. (they) printed (it) ; amonsei me ge scire ? torn by whom ? ga. — This word only accidentally resembles the word gana to eat ; the radical notion in it is of something which is in a very close relation to the one who has it, and things to eat are so regarded. When it is said gcui o tano his gi-ound, gar nolmeat their edge of reef, it may be because food is got there, which makes the place a peculiar possession ; but there are uses of the word which have no reference to food. A charm prepared for any one's destruction is nagana, gan talamatai ; an arrow meant to kill some one is gan qatia ; ni me vanan tamatetiqa, nagaku, he loaded a gun, for me, to shoot me with. So also rain, sunshine, wind, calm, procured by a weather-doctor, is nagana his, gan loena, loa, Ian, taro. ma. — This is only used of things to drink, including sugarcane ; mam pei tva! here's your water, Hoke ton, namama, here's the sugarcane, for you. 6. There are two other words, not different indeed grammatically from these, and equally translated by English pronominal Adjec- tives, but not likely to be taken for Possessive Pronouns : ^;M/aj, anai. A pig, a fruit tree, anything which is a choice possession, \s pulai; pulak som my money, pidan o qoe his pig, napulanina nol Hoke all this is ours, our property. It may be a pulak. With persons, not property but dependents, anai is used: tanun anak my man, a man who follows me, rowroicovag anana his servant. A man of the place or of the veve is o tanun anai. See Vocabulary No. 13. The first syllable a must not be taken to be the same with a before no, mo, &c. VI. Adjectives. 1. Adjectives properly so called are few in Mota ; that is to say, words which are not Nouns Substantive used to qualify other Nouns, nor words which would be translated by an English Ad- jective, but which in grammatical form are in Mota Verbs. If ima vat, a stone house, be considered, it is seen that vat stone is just as much a Substantive as ima house ; it qualifies, but is not an Ad- jective. So tanun we tatas, a bad man, is translated by an Mo! a . A djcc lives. 273 Adjective ' bad' in Eiifilish, but we tatas is in grammatical Curni a Verb K It may be said, jjrobably, that no word used to qualify as an Adjective in IMota refuses to be put into form as a Verb : but some are used simply as Adjectives without verbal form. Such are mantagai small, Ihvoa great, nun true, words with the adjectival termination ga, or with the prefix of quality ma. These are used as Verbs, some frequently ; but they are used, and j^rojierly, as true Adjectives ; ima mantagai small house, tanun liivua great man, vavae nun true word, qon malakalaka joyful day, matesala taniniga straight path. These words are all primarily Ad- jectives. 2. Some of those words have a form which belongs to them as Adjectives, owing either to termination or prefix. Adjectival Terminations. These in Mota are ga, ra, ta, ga. — The examples of other languages, Lepers' Island, Florida, show that this termination is added to Substantives or other words to make Adjectives ; but in Mota many words evidently of this character have no such apparent stem ; such are taniniga straight, aqaga white, turturvga blue. Of some the stem is found, silsil of silsiliga black, rono, as in ronronotar multitude of possessions, of ronoga famous, wmimai dust, wuwuaga dusty. Sometimes i is inserted ; mamasa dry, mamasaiga parched ; tala to be careless, wanton, tata- laiga wanton, wicked. ra. — This is seen in llgligira fluid from ligiu fluid, tvottvotora rough from ivot to stick up. ta. — This is no doubt the same with sa in Maewo ; sasaritd level, equal, from sar to be straight with ; taperaia dish-shaped, tapera a dish ; mamani- gata full of ulcers, maniga an ulcer. It is probable that sa is to be added to these Adjectival terminations ; magarosa, pitiful, has probably the stem garo, with aflixes ma and sa ; and garo prob.ibly is the same with uro, the stem of the Florida arovi. In gaela tough, stringy, there can be little doubt but that la is a termination, like ra, added to gae string. Adjectival Prefix. There is a Prefix of condition ma, which may be seen commonly in the Adjectives given in the Vocabularies. It is prefixed usually to Verbs, and then makes a word which seems i:)articipial, sare to tear, masare torn, late to break, malafe broken. But though this Prefix no doubt demands consideration with verbs, it is no less certain that words formed with it arc ver}' often not such as can * Inau malinsala I am hungry ; mat insula is here an Adjective used as a Substantive, a hungry person. 2 74 Melanesiati Grammm's. be called Participles, but are most convenieutly at least called Adjectives, and those pai'ticularly which are formed from nouns. Examples, matoltol thick, mageregere weak, mamarir cold, malum- htm soft ; mavinvin thin, from vinhi skin, rnanaranara bloody, from nara lood. There is no doubt but that ta in taniniga straight, taiitapa'pa flat- sided, taplagolago cylindrical, is a prefix of the same kind. 3. Comparison of Adjectives. Degrees of comparison are expressed either by the use of a Preposition, or of an Adverb, or by a simple positive statement which impliesli comjDarison made in the mind. The Preposition used is nan fromj o qoe we poa nan o gastnce a pig is bigger than a rat; ilolce we maiio jioa nan this is rather larger than (it); or without the Preposition nan, but with the Adverb mano ; ilolce we mano poa this is rather large, i. e. larger. Another expression is vara, vara poa larger. A Superlative is expressed by an Adverb, loe poa aneane very large, i.e. largest. In the use of these Adverbs the statement is positive, but there is an enhancement of the force of the Adjective expressed by the Adverb. When the statement is merely ilolte ive poa, this is large, there is nothing in the words to express comparison, but it is understood that the estimate is relative ; there is something smaller which makes this large. This is more plainly the case when it is said iloJce tve wia, Hone ive tatas, this is good, that is bad, meaning 'this is better than that,' not that the one is positively bad, but the other is good and makes it seem so. So it may be said, iniko tuwale we wia gai you only are good, meaning not much more than that ' you are very good.' If the comparison is of numbers the Adverb, or perhajDs Pre- position, sal, over and above, is used ; kamam ive qoqo sal neira, or sal avunara, we are more than they, over and above them. 4. There are some expressions which may conveniently find a place here. There are two words which come before Nouns and qualify them as good and bad, matai and mala : o nialai laiiuii a good man, o mala tanun a bad man. The latter is said in a depreciatory way also, without a positive state- ment of badness, as was said of King Cakobau of Fiji, o mala maranaga Hoke o tausis tagai a poor kind of a King this (with) no trowsers. Matai is possibly the Polynesian maitai. The word sokore before a Noun makes it decidedly bad : o sokore iangae a bad tree. This appears to be itself a Noun Substantive sokorai ; another such is parasiu ; a paras qoe a poor sort of pig. The word Wirt MO, sometimes an Adverb, is also used with Nouns, diminishing the significance, or expressing a certain contempt or pity ; iagai wa ! o mano poroporo toia, not at all, (no harm meant) merely a little joke ; ineia o mano tanun ta Taluga he was a Valuga man, poor fellow ! or, only a Valnga man. Mot a . A djec lives . 275 An exjtression meaning the real thing, tlie true genuine thing, is o fur sava, tiir being the same as turiai body or trunk: o t ur vava ta Mota real Mota language, o tur ineia gal ! his very self ! of a picture. Of any tiling big about a person, mat'nj a cocoa-nut is playfully used, na- matig-manuna his cocoa-nut nose, namaiig-toqana his cocoa-nut belly. This is rather used when, for example, it is a small man with a big nose or big belly ; o matig mona is a big package with few things in it, o matigi aka a large canoe with few men on board. Vat, a stone, is used in something of the same way : o vat tangae a large thick piece of wood, o vat tantin a big heavy man. Of food it is said to be matig wia, nae ivia, ivoiaga wia, cocoa-nut good, almond good. Of one who has an abundance of sonietliing it is said that he is mere, child, of it ; meresom rich, som money. One who is fond of something is said to be a bird with regard to it ; mansom one fond of money, 7nanurtna fond of drinking. The image is taken from a bird haunting a bush or tree of the fruit of which it is fond. VII. Veebs. 1. Almost any word can be a Verb in ^lota, being made so by the use of the Verbal Particles to be mentioned below ; qou night, me qon veta it is night already, / Qat qara ukeg o qon sin qon, Qat then let night go, that it might be night ; the substantive qon is in a verbal form : mantagai little, is an Adjective,. ima me mantagai mun nina, the house has become too small for us, shows it a Verb : siwo down, an Adverb, ni me siwo ma, he has (come^ down hither : mun to, a Preposition, na te munia mun tamana I wilt (be) to him a father ^ : he ! an Exclamation, ni me ke ! he (cried) ke ! Veve mother, isei me Veve inau ? who called me Mother 1 ' ^lothered ' me. A clause of a sentence may become a Verb, o matava iva ravrav me qon vagaruei the morning and the evening (were) the second day. Any word then used in a verbal form may be called a Verb, but there are some words which are in their own proper nature Verbs; nonom to think, vava to speak, sua to paddle, and such like, are Verbs ; words wliich are names of actions, not of things, and are not Nouns Substantive. To think is nonom, a thought is nonomia ; to speak vava, a speech vavae ; to paddle sua, a paddling suava ; the * The Preposition being in fact a Noun makes it possible for it to be a Verb here, otherwise it might be said that the clause was the Verb. These Particles, judging by the English translation of them, might be thought to be Substantive Verbs, but they are not. T 2 276 Melanesian Grammars. Mota Nouus and Verbs are as plainly distinct as the English. It is possible, no doubt, to make these words Nouns by putting an Article before them, but they are then Verbs made into Nouns, and not words indifferently one or the other ; te rusagia a'pe non rnawmawui he will be paid for his work. Besides these words which are naturally Verbs, there are Verbs which liaA'e a particular form as such, either by means of a prefix or a termination. The causative prefix va makes esu, which is either Verb or Noun, live or life, into vaesu to save ; th'e transitive termination makes rono, to be in a state of feeling, into ronotag, to hear or feel something ; and vaesu and roxxotag are words the form of which shows them to be Verbs. 2. Verbal Particles. When it is said that these Particles are the means by which a word shows its character as a Verb it must not bje supposed that one of them invariably accompanies a Verb. There are exceptions, to be hereafter explained. Verbal Particles, besides marking the word as a Verb, express Tense and Mood, to some extent at least. The}^ may be divided in Mota into Temporal and Modal. The Verbal Particles are wi'itten apart from the Verbs to which they belong, we vara, me nonom, te sua, not wevava, meno7wm, tesua, the manner of writing Maori having been followed. It is useful to keep the word which is the Verb distinct from the accompanjdng Particle. (i) Temjwral Particles — ive, me, te, ti. u-e. — The temporal force is hardly anything; but yet, as me is decidedly past and te future, we does express the present by diiierence from the others. It is better, however, to dwell as little as possible upon its temporal character ; •whatever may be the time present to the mind of the speaker, when it has been already marked as past or future by me or fe, we continues to be used ; if no time is marked, the tense can be only said to be present. In the case of a narrative the past particle me sets the time ; subordinate actions require no more mark of tense, and go on with ice : but -successive stages of action, if of sufficient consequence, ai-e introduced again with me. JVan ira tatasina me valago nina alo vanua, ice iJo i Qat tana rasoaiia we pute, wa neira me mamakci lava apena : Then his brothers ran and reached the village, and see Qat and his wife are sitting, and they were greatly astonished at it. In this me gives the time past ; as they reached the village they saw, it was not a subsequent event, therefore it is ice ilo ; Qat was sitting, but the scene is present to the mind, therefore it is ice pute is sitting ; astonish- ment followed on the sight, the narrative resumes with me. We should say they came and saw and wondered ; in Mota they say that they came and see, and wondered. In this way an Adjective, or what answers to an Adjective, being in the form of a Verb, has ice for the particle, whatever may be the Mola. Verds. 277 tense of the sentence: ni me puna Ho ape matig si ice tala.i be smelt the cocoa-nut and found it was bad. It is the same if the time is marked as future by te. If the action is a kind of compound, with no successive stages, there is only the mark of time with the first Verb : Team te ganayana toa we imaima ye shall eat and drink. me. — Though the past is certainly indicated by me, it is strengthened by an Adverb veta already : ni me mute ret a he is dead, has died already. The past particle also can be used for the future in anticipation : na me mute nake 1 have died now, o aJca, qa, me tul ! the canoe has sunk: the apprehension is lively, and the strictness of the tense is lost. te. — Here again, though te certainly is future, an Adverb is used to make it more definitely so, anaisa hereafter. When a thing can be considered sure to happen, te is used without a future sense as ' will ' in English : te tamaike val tail it is so every season, will be so. In narrative, when events now past are in view, the future te is still used: ni me vet si te van ma, paso nan me gisraka, he said he would (will) come, and then he started. ti. — This particle conveys the notion of immediate succession of one action on another, and of continuity, regularity, invariable occurrence. Thus it is commonly used in narrative, as one thing succeeds another without any con- siderable interval. There is very little of a temporal character about it. As an example of narrative of successive actions closely connected making up one event, A'ctn i Qat me ronotag si o qon a Vava, ti ligo raka o ratce, ti map alo aka, ti ganio i Vava, ti tun o qon nia; Qat heard that there was night at Vava, ties up a pig, puts it in a canoe, sails to Vava, buys night with it. Invariable condition or recurrence is not very different ; na imana ti taqa pan matesala his house stands (leans forward or overhangs) by the road ; ti tirutira kelkel apena he keeps standing about near it; o gaviga ti taivaga alo rara, the Malay apple, Eugenia, flowers in the winter; o no paka ti nun saru, ti awisiga gupilot kel, the banian sheds its leaves (and) soon buds again, i.e. every season. The following native stoiy gives an example of the use of these Particles : — Concerning a woman and her child (who) slept, and a ghost took Ape tavine tana natina me matur, pa o tamate me la a basket put them two in it then hung them two on the 2>ora me sogon rara alolona, qara siplug rara avaiuo top branch of a tree decayed, then her child woke says INIother toot tangae we kor, nan natina me mamata ti vet lua, Veve 1 see stars many, and her mother woke says na lue ilo o vit we qoqo, nan ravevena me mamata ti vet wa, my child don't move about, we two shall die. And they two called natuk ni2)ea risris nara te mate. Nan rara me sutvare the birds, but (they) not fly could, the fan-tail then at last flew with o manu nan 2)Ci gate gava lai, o tage qara gavag them two, and they two rewarded him with an umbrella palm leaf, and he rara tea rara me rusagia mun o vilog, wa ni rejoiced at it, saying it was because it matched his tail. me nudakaluka apena was ape me taram na golona. 278 Melanesian Grammars. (2) Modal Particles — qe, ta. ' ' These mark something like a Subjunctive or Potential Mood ; si kamiu qe ronotag sava if you feel anything ; lue pnte nare ira tasina a vawo aha ti qe olo sage sits waiting for his brothers still on boai'd when they should bring the canoe to shore ; ta taro if it should be calm. The difference between the two is small, qe is less potential perhaps than ta, and the latter is more used with a view to the future ; ta nawo te nowo, ta tete nawo tete noivo, if there should be surf will weed, if there should not be surf will not weed. The conjunction si, if, is u^sed with both, but less commonly with ta. (3) The Particle— ti. This is distinct from the narrative or continuous ti before men- tioned. Its use is double, one to throw back the time so as to make a pluperfect, the other to mitigate or moderate the direct- ness of a statement. In the one case it rather belongs to the tem- poral particles, but it cannot be classed with them. It follows the Verb. 1. The Pluperfect in English is not always used where ti is in Mota, but in most cases it can be used to translate it ; neira me matur tama ni me varegira ti they went to sleep as he had commanded them ; »an neia loa, ilone o gene nau me vet ti mun kamiu ajiena, then said he, that is the thing I told you of, i.e. what I told you of before, what I had told you of. Since this use oi ti throws the time back a stage it is evident that it cannot be applied except in a narration in which successive stages of time may be brought into view. It cannot be used with the Present. 2. There is no exact translation of ti in its other use, but 'just' is like it, when one says ' Just come here.' A Mota man would not say to his father- in-law, mule ma gai, come here ; na apena te maragai, he would be shy of doing so; he would say mule ma ti just come here; tlie directness of the request would be moderated. The use is a matter of feeling. The same particle, which is not a Verbal Particle as ice, me, te are, but here is rather adverbial, has another use : it signifies that something still remains ; mantagai ti a little still remains, o kereai ti there is still some-at-tbe-bottom. The sense is intensified by e: mantagai ti e aka me lul, the canoe was a very little short of sinking. This is probably the same ti that following a Verb gives the sense of in- completeness to the action described. It goes with ti of continuity or te : ti tiratira ti he keeps standing about ; manu te roroico ti birds will (or, in narrative, would) keep flying off. 3. AA^erb is used without a Verbal Particle before it (i) in the Imperative, (2) in a Subjoined clause, (3) in a Negative sentence, (4) after certain Adverbs. (i) ■I^nperative. — The simple Verb is enough, 2'>^(tt^ siico sit down, Mota. Verbs. 279 •mule go : but in the Second Person singular it is common to use the Pronoun ka, lea fide siwo ma sit down here, ka mule at go. In the First Person na mule, nara, nina, kara, mule, let me, us, us two, go ; but kamam a mule with a Preposition ; in the Third, Tieira mule let them go, ni mule let him go. In the Dual for the Second Person ura or wura is used ; ura m,ule ilo tuqei, ura nowo valis, go you two into the garden, clear away the grass ; pa alo vie nowo qet wura mule alo takelei, and when the weeding is quite finished go you two on the other side. In ura the numeral rua is present as ra. When three jiersons are spoken to tol is used, tol mule, tol nowo, you three go, weed. In a respectful way a qaliga, son-in-law or father-in-law, will say ura, you two, to his qaliga ; and a woman with an infant is spoken to, and of, as two persons ; and in addressing more persons than three tol is sometimes used. But whenever ura or tol are used, two or three persons are in the mind of the speaker. In the Plural tur is used ; tur mule indpul rua gese, go two and two. Sometimes the future te is used in an Imperative sense ; kamiu te ge tamaine you shall do so. For Negative Imperatives see below under Negatives. (2) In a Subjoined clause there is no Verbal Particle ; na me vet si na mule at, I said that I would go, si neira mule, that they should go. (3) When it is said that in a Negative sentence there is no Verbal Particle it is meant that none is apparent, except in the case of qe and ta. JVau gate maros I don't wish, I won't, or (because in the absence of a Verbal Particle there is no note of Tense) I did not wish, I would not : in the Future na tete maros ran I shall not like it at all, I certainly shall not be willing. But si na qe tete maros, or nau ta tete maros, if I should not be willing. This is enough to put qe and ta upon a different footing from ive, me, te ; shows them more like Conjunctions. It will be seen, when Negatives are con- sidered, that there are, in fact, Verbal Particles in gate and tete. (4) After certain words which may be called Adverbs there are no Verbal Particles, perhaps because the time is given by them. These are qara, qale, kere, and teve. The meaning of qara is, now for the first time, upon that, immediately, just now, recently ; nau qara raxe u\ok letas, I now for the first time write a letter ; 28o Melanesian Grammars. ni me ffopa o qon nitol, qara mate, he was ill three clays, then died, ov it may be ni qara mate, he then died, ni qara mate ti he is just dead. The meaning of qale is still : kamam no! qale esuesii gese we all of us are still well. It is a matter of dialect whether Jcere or teve is used ; the meaning appears to be something like ' only ' in the way of diminishing the importance of the action ; see, however, ker in Motlav and Gaua : ni tv& tantan apesa? ho me vusia ajpesa ? why is he crying ? what did you beat him for ? Tagai tva, nau here, or teve, tut gap neia. No, I only just hit him with my fist. These correspond to what have been called expletive particles in Fiji. Thus mani appears to answer to qara : sa taura e dua na vatu ko hoy a ka mani viriki au he took (Mota taur) a stone and threw it (Mota vivir) at me. So also ' bagi gives a tone of surprise to a statement.' Rev. L. Fison. (5) It often haijpens that a Verb without a Particle appears in a kind of Infinitive, and it is then really a Noun ; kamiu me mule ma si a maivDTaivui you came here to work. Perhaps ^yrti until, till, can hardly be called an Adverb, but a Conjunction ; the Verb after it has no Particle, na te goard amaiko nau gai mate, I shall abide with you till I die, neira me vagvagalo gai mate qet, they went on fighting till they all died. A Verb with a Particle may equally be used as a Xoun ; ko we piite mamasa gai, ape sava ? ape we nala qa. You are sitting idle, what for ? because of being tired ; neira me ilo me silsiJiga they saw it had become dark ; ape me taram na golona because it matched his tail. In these examples it must not be supposed that the Nominatives, 'it' in English, are omitted ; tre nala, me silsiliga, me taram, are Nouns : the being tired, the having become dark, the having matched. 4. Suffixes to Verbs. These suffixed terminations make a Verb which without them is Neuter or Active definitely transitive ; there must always where one is employed be an object before the mind, though it may not be expressed in words, upon which the action of the Verb passes over. The Verb vava, to speak, expresses an action of a general kind ; when g is suffixed, giving it a definite transitive force, the speaking is shown to be directed upon or against some object ; vavag to speak against ; gava to fly, gavag, as in the story page 277 to fly with, convey by flying : rono is to be in a state of feeling generally, rono vivtig to be in pain, rouo puna to have a sense of smelling, ronotag to feel or hear something, ronotag vivtig to feel a pain, ronotag punai to smell an odour; vano is to go, vanov is to put, vanogag to go with something, to convey. The suffix may ^PPl.y to a word which is not commonly a Verb, mata an eye, matag to eye, making a transitive Verb. These Suffixes do not in Mota, as in Fiji, take altogether the place of Pre- Mota. Verbs. 281 posltlone ; sometimes a Preposition, especially goro, is used after a Verb in this form, kokor goro to protect something from or against what may do harm. These Suffixes are of two forms, Consonantal, the addition of u single Consonant, and Syllabic. (i) Consonantal Suffixes. — It is evident tliat these c.an only be added to Verbs that end in a Vowel. The Consonants suffixed are, g, t, V, r, s, n, n. For example, mana, spiritual influence, or to have it, manag to enable by passing over that influence ; mava to be heavy, mavat to be heavy upon, to weigh down ; sora to lay a plot, entertain a design, sorav lay a plot against some one ; koko to shut in, kokor to protect ; kokos to enclose ; rau to thrust the hand into a bag, ra'iin to thrust in the hand and take out some- thing ; tiqa to shoot, let fly an ari'ow, tiqan to shoot some- thing. The most common of these suffixes is g. (2) The Syllabic Suffixes are ag, gag, tag,>vag, rag, sag, mag, lag, nag, nag. Examples, taleag turn, fi'om tale about ; vanogag take, from vano to go; altag to look after, from al to go about; sirvag to cut close, from sir to shave ; mataray to gaze at. from mata eye ; maraesag to laugh at, from marae to laugh ; saromag to sheathe, from saro to go in ; tigonag to pole a canoe, from tigo ; lilnag to spread, from HI to unfold. The efiective part of these Suffixes is ag, the consonants preceding serve only to introduce this. In the case of either the Consonantal or Syllabic terminations, it is ini' possible to connect any particular force with the form of the Suffix. The Verb has assumed the Suffix which use has appropriated to it. Sometimes two terminations are in use without any variation of meaning, as sarovag or saro- mag to sheathe : or in some cases a variation of meaning accompanies a varia- tion of Suffix, as from koko is made kokor to enclose in the way of protection, kokos to enclose in the way of preventing escape, kokot to enclose in the way of straitening ; cases in which it is evident that use only, and not any force in the consonant suffixed, gives the particular signification. The Separable Siffix — vag. There is a Suffix vag which must be distinguished from the definite transitive termination vag above mentioned. This is always equivalent to the English * with,' and can be separated from the verb to which it is usually affixed. Thus mule to go, mulevag to go with, o reremera we mulevag o tajwra the boy is going with a dish, ni ive mule raveaglue o tinesara vag o tapera he is going through the courtyard with a dish. 282 Melanesian Grammars. The Preposition ' with,' to which this separable vag is equivalent, is that of accompaniment, not ' with ' instrumental : iwasvag o tapera to fall down with a dish, matevag o gopae iuiunsag die with a fever. The distinction between this and the other V(W, which is one of the Syllabic Suffixes above enumerated, and is not separable from the Verb, throws light upon the character of them all. The Syllabic Suffix vag, which makes a Verb definitely transitive, may indeed in some instances be represented in translation into English by the Preposition ' with,' but so may also the other inseparable Suffixes ; it may also, like the rest, be represented by some other Preposition. But vag which is separable is always equivalent to 'with.' In proportion, therefore, as it appears to be separable, it appears to differ in signification and power from the inseparable Suffixes, and to assume the appearance of a Preposition. Two questions then arise : is the separable vog the same with the inseparable ? If the same, is the separable vag an example of a Preposition on the way to become a Suffix, and left behind by tag, rag, sag and others, which are inseparable Suffixes 1 or is it an example of a termination beforehand with the others in becoming detached and turning into a Preposition ? The fact that all the inseparable Suffixes, including vng, are identical in signification is against the probability of their having been Prepositions. In Fiji some at least, if not all, of the corresponding Suffixes are separable, the Suffix coming at the end of a number of words and combining them into one Vei'b, as miile-raveaghieo-iinesara-vag may be regarded as a compound Verb. In Volow hea = sag is equivalent to Mota vag. There are some Adverbs usually or often written in one with Verbs as if Suffixes : such as reacj, vitag ; mapreag to put away, nomvitag to forgive. This proceeds partly from the dropping of i in vitag ; toavtag for toa vitag to go away and leave ; raJcavtag, raka vitag to take up and away, leaving something behind ; tanov- tag take hands off, tano to touch. It should be observed that a transitive suffix and causative prefix frequently combine in the same Verb : vapitfeg to seat, from^i(^e to sit, vakokot to close, vasinar to make to shine, valeasag to dispute. It is with an uncommon force of the Suffix tag, more like Fiji taka, that vegatag means to climb for a person ; isei te vegatag inau ? who is there to climb (cocoa-nuts) for me ? 5. Prefixes. — These are Causative, Eeciprocal, of Condition, and of Spontaneity. (i) The Causative is va, sometimes raga; esu to live, vaesit to make to live, to save ; qoqo many, vagaqoqo to multiply. This Prefix is in very frequent use, but yet the Verb ge or wa," to make, is often used in place of it, me ge esua instead of me vaesna, saved him, vat ge lot a pestle to make lot. But ge often means to act, not to do, ni we gege loloqow, he acts like a fool, not makes others foolish. The Verb va to go, combined with another word, may sometimes appear like the Causative : vailo to visit, vatatii to encounter. (2) The Reciprocal is var\ vara we varvus they two are beating Mota. Verbs. 283 one another. This may be reduplicated ; varvarvus, keep on beat- ing one another, (3) The Prefixes of Condition ma and ta have been ah'eady mentioned under the head of Adjectives ; as they are prefixed to words which cannot be called Verbs, it is not possible to con- fine them to this connection. Prefixed to Verbs ma often gives much the sense of the Passive, or of a Participle : sure to tear, masare torn. Inasmuch as words which would be translated by English Adjectives have generally in ^lota the form of Verbs, i.e. are used with Verbal Particles, it is hardly possible to keep a distinction in the case of this Prefix between the words beginning with ma of condition which should be called Adjectives and those which may be called Participles. Examples illustrate the value of the Prefix : wora asunder, mawora parted, to come apart, tapera me mawora the plate is broken ; late to break, malate to be broken ; laka to kick up the heels, malukalaJca to rejoice, to he in an exultant condition ; luqe to fold, maluqe folded ^. It may be re- duplicated mamagese lonely, mamagarosa pitiable. The Prefix to. is not so common as mo, and may be thought to have more of the meaning of spontaneity. It is found in the words tatiotio to stagger, taavaava to miss footing; in taj^kigolago, from ta and qolago a cylinder, a word now used for a wheel as qolago is for a barrel ; in taptapapa slab-sided, from ta and ^^opa, a word which jierhaps is not used as a Noun in Mota, but is found in Malay and ^laori as a plank or slab. It is reduplicated in tata- wihvil to come rolling over and over. To these may perhaps be added sa ; sasaroro to come or sink down. (4) The diff'erence between ta and tava is that the latter shows the condition indicated to have come about of itself: %d to untie a rope, tali me tavaul the rope has come untied of itself; tavamasu to fall down, tavaroro to sink down, tavaraka to rise up, spon- • Nouns are formed from these Verbs or Adjectives : masare, torn, o masarei a rag; maluqei, malqe siopa, a roll of cloth. A singular word of this kind is magesei, for which there is no English translation : it is compounded of the prefix of condition ma, the Adverb gese only, and the nominal termination t. The Noun thus compounded takes the pronomii«il Suffix, and magesek, ma- gesema, magesena, is used where in English the Adjective 'alone' would be used; I alone did it, by myself, inaii mageselc, itiiko magesema, inela ma- gesena, thou alone, he alone, literally my lone, thy lone, his lone. The usage of suffixing the Pronoun belongs to Florida and Bugotu, hegegu, gekegti, and 284 Melanesian Grammars. taneouely; to^v is sometimes used, tavsare torn. In reduplication tatavaul, tavtavaraka. Another form of rare use in Mota is tama : gae me tamarurvs a line run out to full length. There is no difference in meaning between this and tava- ruruSf and tama is the Prefix in Lepers' Island. It may be doubted whether in the word tawaga to come open (taioagasiu a flower) tawa is not another form of tava, for gaga is to split, gape. Other- wise it may possibly be ta the Prefix, and waga = tcaka to open. The Prefixes var and va can be applied to the same word, and make up one Verb, as varvaesu save one another. It is usual to write the Causative and Reciprocal Prefixes together with the Verb, and it is convenient, by doing so, to mark the distinction of them from the Particles by which the Verbal character of the word, or compound, is marked : thus neira we varvaleleasag gese, they do nothing but dispute among themselves, might be written we var va leleasag, but with less clearness. 6. Convpound Verbs. — There is a use of two Verbs combined in one word by which the actions described are represented as succeeding one another in time, in such a way as would seem most naturally to be given by two Verbs connected by a Conjunction. Such a word is roivopute, roivo to fly, pute to sit ; vianu me roivopiUe a bird has flown and sat, rowotira jump and stand, light on the feet, valago nina, run and arrive. Tt may not be possible to separate this use from that of a Verb added to another as an Adverb. It is common to use paso, finish, after a Verb, to mark the completion of the action, qalo, to hit, to mark success, and ilo, to see, to convey the notion of attempt or approach to success : o aka me kel paso ma nau qara kel siivo avunana, after the vessel has come back I shall go home on board her, i.e. the vessel has come back, that complete, I shall return ; ni me vivir toa, gate vivir qalo, he threw at a fowl, did not hit it, i.e. did not throw (and) hit ; nau me ge ilo apena, I tried to do it, i. e. did (to) see (if I could) ; tagai, nau gate ge ilo mok, not so, I did not do it at all. 7. Impersonal Verbs. — There are some Verbs which are regu- larly used without a nominative ; we vivlig nau it pains me, me vule veta inau I am already tired ; me raktitia he is in great pain ; me tama mun nau I was confused, it came awkwardly to me. These Verbs can be used otherwise : o sava we vivtigiko 1 what hurts you ? o vivtig we raktitia pain causes him agony. All Verbs can be used in what may be called an Impersonal way, and a Passive sense is so frequently conveyed. 8. Passive. — There is no Passive form. The Mota Verb, in fact, has no Voice ; the Verb names an action or condition as a Noun names a thing ; if the signification is such as would in English be Mot a. Verbs. 285 conveyed bj' a Passive form, the Mota Verb convej'S it as well us an Active sense. In some instances tins is j^lainly the case, in others the ex- pression can be explained as an impersonal form of sentence. To build a house is we taur ivna, they build a house neira we taur o ivaa ; but ixaa me taur veta cannot be otherwise translated than as 'the house is already built,' although, according to the form of the sentence, o ima is now the Nominative to the Verb taur, as neira was, and the Verb, without change in form, has become Passive in sense. It is the same when no Nominative is directly expressed, but clearly understood ; map siwo ma iaj^era put the basket down here; me map veta wa (it) has been put down already; avea nom tausis ? where are your trowsers 1 me asa qa, me tvor, (they) have been washed, (and) laid in the sun to diy. The Verbs ma]> to put, asa to wash, wor to spread out in the sun, can be certainly called Active Verbs, yet they are clearly used in a Passive sense. It is just as correct, however, to say me taur jx^iso ima the house is built, me map ta2)era the basket has been put down, we asa suqya the garment has been washed, in sentences in which, to speak according to European grammar, ^ma, tapera, siojya, are no longer Nominatives, but the objects of the Verbs taur, map, asa. Such sentences, therefore, can be taken as im- personal, vie taur im.a, as if neira, they, have built the house. But it cannot be too positively stated that the Mota Verb must not be considered as capable of Voice. It has been already said (5. (3) page 283) that the Prefix ma gives something of a Passive meaning to the Verb to which it is prefixed ; ni me sare non s'lopa he has torn his clothes, non siopa me masare his clothes have been torn. But it would not be right to call that a Passive form which may equally carry an active sense, as malahalaka rejoicing; the Prefix ma shows a con- dition, which may -be Passive or not. A Verb added to a Noun to qualify it may be Active or Passive, that is, may be translated in English by a Passive or Active Participle. Thus qat vusvus is a stick to strike with, a striking stick, vus to strike, and gene vus is a thing struck ; taur to catch, gene taur the thing caught, vasigtag to abhor, tanun vasigtag an abhorred or despicable j^erson. In these cases the Verb is added in a bare form which does not show, or perhaps when thus added and qualifying does not even allow, it to be 286 Melanesian Grammars. really a Verb. Sometimes, however, a full Verbal form with the Verbal Particle qualifies a Noun, and may have the signification of a Passive Participle ; nau me ilo o aha me lamas aqaga, I saw a ship (that) was painted white. 9. Reflective Verbs. — A reflex action is described by the Adverb hel, back : I strike myself, nau we vus hel nau I strike me back ; he strangled himself, m me ligo mate kelua. An example of another sort is ni m,e gana vasusumagia he ate to excess, to stufi" himself. Ko we vara Jcel o sava apenikol What do you say of yourself? 10. Negative Verbs.— It has been said (3. (3) page 278) that Verbs in a Negative use have no Verbal Particle except qe and ta ; nau gate ilo I do not, or did not, see, nau tete ilo I shall not see. A comparison of Mota with other languages makes it apjDear, how- ever, that the negative lies in te, and that probably gate and tete, written as one word, are ga te, te te ; ga being the Verbal particle, present, with te negative, te the future particle with te negative. Sometimes mate is used in Mota as the negative with a Verb, nau mate taka ineia I won't obey him ; it is perhaps borrowed from Vureas, Vanua Lava, and is thought stronger than gate and tete. It is correct to say ineia tagai iake he is not here, as well as ineia gate lake. Tlie Negative Imperative or Dehortatory form is with the use of the word pea, which probably means ' naught.' A person tired or wearied will say na me pea ran I am good for nothing, have become utterly naught : so of anything given up or put down as objectionable it is said me pea veta it has been done away, has been brought to naught. To forbid anything the expression is ni pea let it be naught, i. e. let it not be. With this the Verbal Particle may or may not be used, nipea maitir don't sleep, or nipea to.e matur, matur or we matur being both, in fact, Nouns. The use varies to some extent with the different persons : inau nipea matur (or we matur') let me not sleep ; in the Second Person nipea matur, or iniko nipea matur, in the Third ineia nipea. With the Dual ura iS used, as ordinarily in the Imperative, vra pea matur don't you two sleep ; of three persons tolpea ; and with the Second Person Plural tur, turpea matur don't sleep. It may be, however, kamurua nipea or nipea ineira ice matur, the difference being rather that between a direct and indirect prohibition. Another word, wa lest, may bf used : ko wa 7nasu don't fall, rather take care lest you fall. 11. Auxiliary Verbs. — There are certain Verbs which may be called Auxiliary, having rarely a separate existence of their own, and being Verbs, not Prefixes. For example, ni me ti nanagona a2)e vanuana, he set his face towards his own place, shows_^' in all respects a Verb, yet ti usually occurs in composition, tikula to turn the back. Another is .so ; ni we so naranoua a matesala he sets, advances, his foot in the path ; in composition sonago to scorn, set Mota. Adverbs. 287 the face against, so-wus to hate, set the lip against, sororo to set about a report. These words may thus be translated to show their use and meaning, yet ti and so only mean to set when in such combination and connection as this. 12. lieduplication of Verbs. — Verbs, like Nouns, are reduplicated in three ways, either by (i) reiietition of the first syllable, or (2) by that of the first syllable closed by the succeeding consonant, or (3) of the whole word ; for example, j^ute to sit, ]^/U2)ute, putpute, pute2mte. The force of these reduplications is — (i) Continuance, prolongation of the action, pupule keeji on sitting ; the redupli- cated syllable may be repeated as often as the idea of continuance or prolongation to be conveyed requires. (2) Intensification, the force of the word is magnified, piutpute to sit closely down, siksike, from sike, to seek earnestly. (3) Repetition, 2>utepute to sit, get up, sit again, sit from time to time, ima 2>ute2)ute a sitting room. Chai'acteristic examples are : aka me sale-sale-sale sasasale the canoe floated and floated and went floating on ; neira me toga totoga they stayed and stayed ; oiaw ira Qat me tootoga a Tetgan Qat and his company prolonged their stay at Tetgan. Reduplication of this latter kind depends more on the tone, feeling, and gesture of the speaker than on anything that can be reduced to rule. The form of Reduplication (2) is evidently only applicable to words having an open first syllable, to close ■vvhicli in reduplication the consonant is borrowed from the second : pu-fe, put-pu-te. Soruetimes two syllables are repeated, with perhaps the character of this second form : liwoa great, llwoUwoa ; purei unskilled, pnrepiirei. VIII. Adverbs. Some words used as Adverbs are Nouns, assisted often by Pre- positions ; ^ome are Verbs ; some ai'e words not used in any other grammatical form. Adverbs can be divided into those of Place, Time, and Manner, but place and time are generally conceived of as the same. It is the habit of the native mind to have place constantly in view, to use continually Adverbs and Particles of direction pointing hither and thither, up and down, landwards and seawards. The Particles ke and ne point here and there, and form part of many Adverbs of Place and Time ; ma, hitherward, and at, outward, are in continual and conspicuous use to indicate motion, or direc- tion of thought, or course of time. 2 88 Melaiiesian Grammars. Examples of Adverbs. 1. Adverbs of Place: — Here iake ; there iane, aia ; where avea. Hither ialce ma ; thither at aia ; whither 1 ivea ? Hence iake at ; thence ma aia ; whence 1 ma avea ? Up kalo, raka, sage ; down siwo, sur ; afar aras ; everywhere valval ; off away, reag ; near ^xtu, j^^ten, arivtag ; a varea outside. The primary meaning of the last is not in the house, but in the varea, the village area, but the Noun vareai has come to mean the outside generally. There is a difference between iane and aia ; the one is demonstrative, the other indefinite. 2. Adverbs of Time : — Now ilokenake ; then alo Hone ; when ? anaisa ? ananaisa of past time. To-day qarig ; anaqarig of past time ; yesterday ananora ; day before yesterday anarisa ; to-morrow amaran or maran ; day after to-morrow arisa. Hereafter anaisa; heretofore, ah-eady,'ye^, and ma, men, can be seen to be in fact Nouns. (i) a is simply locative, at; uvea? a Moto, where? at what place 1 at Mota. This has been seen in Adverbs, and enters into the Compound Prepositions. It is very important to observe that the native idiom uses this Preposition where in English we should use ' from ; ' the place in the native mind is that from which the motion starts, not where the object in view is. Thus ni we mule ma avea ? a Mota nan qa where does he come from ? from Mota to be sure, literally, 'at what place does he come hither ? at Mota;' ni me masii avune iangae he fell from a tree, literally, he fell on the top of a tree ; Tcamam me ilo o aka a matenua we have seen a vessel from the cliff, at the cliff. See p. 160. This Preposition is used before the Infinitive Verb, as before a Noun, much like the English 'to;' a na sava nial to do what with it? (what are you going to do?), a na gagarat nia, ni matro apena to do my skin disease with it, that it may heal ; ni me risa si a matur he lay down to sleep. (2) i is of Motion to, ni me van ivea ? i lau, where has he gone to, to what place ? to the beach. It is probably this which is used in expressing direction, i siivo, i rowo, inexactly. West, East. (3) sur, sir, is of INIotion to a person, not to a place ; mule suria go to him, neira me gayanag ma sur kamam, they came and Mota. Prepositions. 293 told us, literally, they told it hither to us. Without personal approach, however, sur may he used, ni me, rave ma o letas sur nau he has written me a letter. This preposition is never merely Dative in Mota. J^ singular use of the word is in the phrase ni me ilo sur anona he has chosen for his own. The use of the word in songs has a much wider scope, as it has in other islands. In the phrase mule sur (or sir) vanua go along through the country, gamo sir j}an vanna sail along the coast, the word sur is a different one ; the same which as reduplicated is susur to sew, and which means to sing, sur o as. (4) cforo is the most difficult of ^fota Prepositions. There is always the idea of uinlioii in it, and of motion against, although in fact no motion may take place. If a man stands by a house ni ive lira a/;e ivna, if he stands leaning against it, 7ii we pesinag goro ima ; if a rail leans against a fence the preposition is a/;e, but if it props up the fence it is goro ; if a man sits before another without any reference to motion ni ive 2mte ape nagona, he sits at his face, but ni we imte goro nanagona if he be spoken of as coming and sit- ting over against his face. With this is connected the sense of round about as in the way of guard or protection ; with a view to motion inwards or outwards, to fence round a garden is geara goro o tuqei, as if to keep the garden in ; to fence against pigs is geara goro qoe, to keep them out. In many cases goro will be translated ' over,' but not with the simple sense of superposition, with the sense of ' over against,' of motion or of action. To put on clothes is to saru goro nataraperaa mun siopa, to clothe over your body with a garment, the notion is that of the garment being made to pass over, and o siopa ive toga goro natara2)ema is over your body, in the way of protection or concealment. In the same way to paint over a surface is lamas goro. If clothes are spread out to dry in the sun, we wor goro loa, to meet the sun's action ; if one sits by the fire simply with the notion of position, ni we 2nite 2>cin av, but if he sits to look after it, ni we 2>ute goro ; so ni tve masil goro av, he warms himself at the fire, sits against it, to meet the warmth. Then follows upon this the sense of ' after,' ilo goro look after, in common use as an exclamation of warning, ilo goro ! look out ! that is against something occurring ; hamam we ilo goro mantag 7iai- mama, we look well after your house, take care of it lest any harm should come to it. Beyond this there is the meaning of 'after' as in going after to fetch, rnule goroa go after him, not in the sense of following only but of coming to him. 294 Melanesian Grammars. The word is often in use as an Adverb, as in ilo goro ! look out ! ni me mule goro veta he has already gone to fetch, o tuqei me geara goro nan o qoe the garden is fenced round to protect it from the pigs. It should be remarked that goro after a Verb refuses to be separated from it by an Adverb. Unlike other Prepositions it adheres to the Verb, not to the Noun it governs ; as above, ilo goro mantag naimama, not ilo maniag goro naivciaraa, look after well your house, not look well after. This is probably because of the adverbial use of the word. The Verb ilo is transitive, and therefore can govern, to speak in the way of European grammar, the object ima, and goro may lie taken as an Adverb ; but the same is the case with intransitive Verbs also, as tira to stand, ni me tira goro maremare itiau a matesala he stood firmly against me in the path ; goro is plainly a Preposition, but it follows closely on the Verb. The Verb goro, goro late o iangae snap off a plant, goro mot o gae break short off a line, is not connected with this Preposition. (5) nan. — The Preposition of Motion from is nan, with the simple meaning of ' from ; ' not used, however, as has been observed above, in all cases where 'from' would be used in English. Like an English Preposition, this is used at the end of a sentence without a Substantive, referring back to one that has gone before ; Hone naimana ni qara rowolue nan, that is his house he has just gone out from. Such a use may be thought adverbial, as nan may be called rather an Adverb than a Preposition in the following use, nan ma area Hone? nan a "Mota ma from whence is that? from Mota; literally, hither from at what place ? from at Mota hither. The Preposition in this sentence which answers grammatically to the English 'from' is a 'at;' nan may be translated by 'out,' out hither at Mota, Mota the place in view, the motion outwards and hither- wards. So ni me sere lue nan nlo vatifnara he went out from among them, literally, from in the midst of them. It may be conjectured from the Duke of York na (see that language. No. 34) that nan is in origin a Noun. The use of nan in comparison has been noticed, p. 274. It is used also as 'more than,' ni tee gilala nan ratatasina he is more knowing than his brothers. Another use is with loa lest, nan ni tva Jcel ma lest he should return. (6) mun. — As the Dative nmtn is simply 'to' and 'for;' la ma m,un nau give it to me, munsei Hone ! who is that for ? mxm nau Jcel for myself. Thex'e is a use of mun not precisely equivalent to the common use of ' for,' though often properly so translated ; lav nau mun mereima, na te roivrowovagiko take me for one of your household, I will work for you ; neira me lav rasoana mun rasoara, wa naa- kana mun aJcara they took his wife for their wife and his canoe Motel. Prepositions. 295 for their canoe. In this use the Noun after mun takes no Article, as it would in the ordinary sense of 'for,' innn naakansei o e.pa iloJce we susnr ? for whose canoe is this sail being sewn ] mun naakak nake qa, for my canoe here to be sure. (7) mun. — The other use of 7nu7i may almost always be trans- lated by ' with,' and is mostly instrumental ; ni me vusia viun o here he struck him with a club, neira me galea mtm galeva nan toe qoqo they tricked him with many deceits, o qolarjo %ve ura mun 2)ei the barrel is filled with water, equivalent to uravag. It is sometimes to be translated by ' by,' ni me vusia mun o galeva he struck him (killed him) by guile. The notion of a means or in- strument depends on the previous conception of something accom- panying, so mun still translated by ' with ' has a sense not instrumental ; ni we mule ma mun o sava ? what has he come here with 1 mun tana kumara with bags of sweet potatoes ; o tanun mun o rurus a man with a rheumatic complaint. There is an interesting question concerning this word, as concerning nan, whether it is not possible to view it as a Noun. In neighbouring languages, see e.g. Motlav, Volow, Pak, the corresponding word has no final n, from which, following the analogy of other Prepositions, it may be conjectured that mun is an archaic Noun, «»« = ?«?', with the Third Person Pronoun suffixed. This is confirmed by the use of mura to them, murara to them two, otherwise munra, munrara, in which mii certainly appears as a complete Preposition. Since this is so it is allowable, though not perhaps desirable, to write tntinau, munina, muneira, not mun nan, mun nina, mun neira. There is a further difficulty in the fact that in Vanua Lava, e.g. me corresponds to instrumental tnun and is also the equivalent to the Mota me (12), which has the sense of accompaniment. It may possibly be that the Mota mun ' with ' of accom- paniment (the same with mun 'to' dative), though in but one form, repre- sents two Prepositions of the neighbouring languages. (8) nia. — The use of this is peculiar, inasmuch as it always comes after the noun to which it belongs ; iloke here ni me vusia it nia this is the club he struck him with. The English equivalent is perhaps ' withal,' sometimes ' thereby.' There is a peculiar use of hm after the Verb ris to change, torou Hone te ris rupe nia that caterpillar will turn into a butterfly. It is said also la mun mama nia give it to father, in which, as in the other, it is possible to use ' withal ' as a translation. The question thus arises whether it is a Preposition at all. The Banks' Islands languages generally do not have the word, but in Merlav and Gaua ni is the instrumental Preposition, and in the New Hebrides in Espiritu Santo, gini, ginia, show the same. Florida, though far off, uses nia as a Preposition, and uses it also with the word liliu to change. (9) ta. — This is a Preposition of relation with regard to place only; tanun ta Moto a Mota man, tangae ta Maewo a Maewo 296 Melatiesian Grammars. plant, vava ta Meralava the language of Merlav, ta ^iotalava a Motlav person. The meaning is 'belonging to' a place ; it cannot therefore be used of a part of an island, as of Veverau a place in Mota, or Takelvarea a part of Veverau, which is takele Mota, takele Veverau. In o ta Moata alo mate tironin ilo iiwa he climbed by the window into the house; alo translated 'by' signifies that he went inside the window. The translation of talo is ' of,' the notion being that of having a place in and so belonging ; tangae talo vc\ot a tree of the forest, o linai talo tve tuai a custom of old times. The use of alo as an Adverb, whilst, when, follows on this signification of the Preposition. Mota. Prepositions. 299 Ohserrafion. — To some Prepositions it is usual and natural to write the Pro- noun suffixed, mnniko, nuria, amaira, nauia, goroa; to the others it is not possible so to suffix a Pronoun. With some Persons it may be doubtful whether the Pronoun should be so suffixed, whether it is correct to write mtinau, nanina, not mun nait, nan nina. It should be remarked that the Pronoun following these Prepositions is never (except in apena) that suffixed to Nouns, but the short form of the Personal Pronoun that is suffixed to Verbs. But there is not in Mota a shorter form of «a« and nina used after Verbs, and it would therefore seem more correct to write mun nau, than munau, while munia, mtiniko, munra are undoubtedly right. See above, (7). 3. There are many words, constructed of Simple Prepositions and Nouns, which naturally are written as single words and are represented by the English ' upon,' ' above,' ' below,' ' between,' ' within,' ' beside,' ' beyond.' These, as the Noun takes befoi-e it the Prepositions a, i, or ta, change their signification accordingly. For example, vunai is a Noun, the uj)per side, avune vat is on a stone, iviine vat on to a stone, tavune vat from on a stone. These are evidently not true Prepositions, but since they are translated by Prepositions they have their place here. In most of them the Noun is, like vunai, of the kind that takes an independent termination ; a few are made with Nouns that have not that form, such as vawo, lele ; and these latter cannot change their termina- tion when constructed with another Noun ; it is avune vat, because vu7ia, the independent form of which is vunai, becoming the first of two substantives, pai"t of a compound word, lightens its termina- tion a to e ; it is avawo ixaa, alele ima, and there is no use of vawo or leU in Mota except in this way. Examples of these words are avawo on, alele in, of one kind. Of the otlier, vunai, avune as above; Jalanai the under side, alalane ima under the house, that is, at the under side of the house; ilalane, talalane -with the change of meaning belonging to the Preposition ?'and ta ; gasutoe me valago ilalane ima, the rat ran under the house, loetoii talalane ima a bottle from under the house ; vatitnai the middle part {va causative, titinai, tina inside), alo vatitnai, alovatitne, ilovatitne, talovatitne, between ; panel a hand, apan by. These are naturally written in one word, but there is nothing but convenience to cause it. Other words, though translated by an English Preposition, would never be written otherwise than as Preposition and Noun, tavaliu a side, a tavala pei beyond the water. All these Nouns are used with the sense of Prepositions when the Pre- positions are in fact left out, vaivo vat, vune vat, as well as avawo, atnine ; tavala pei, lele ima : but in Mota the use of the Preposition is correct. When a person is spoken of, the Pronoun suffixed to the Noun destroys the appearance of a Compound Preposition: vunai is 'top,' a Noun, vune vat, a compound of two Nouns, top-of-stone, avune vat on top of stone, on a stone ; it is natural to write tangae me m.asu avune vat a tree fell on a stone ; but 'top of me' is vunaJc, tangae me vaasu a vunak a tree fell on me, though 300 Melanesian Grammars. avunak, not a vunalc, may be written, on top-of-me will not be taken for a Preposition ; a is always the Preposition really, but avune appears to be one. 4. There ai'e words also which may be called Complex Prepositions, because,\ though no part is a Preposition, the compound has the force and use of one ; such is raveaglue, raveag in a straight direction, lue out, o sinosino loa we sina raveaglue o tironin the sunbeam shines through the glass. Other words, which are properly Adverbs or Verbs, are used as Prepositions ; wal'iog around, leas instfead of, ^^eien near, tataga according to, lagau over. X. Conjunctions. 1. The common Copulative Conjunction is wa, and. 2. The Adversative, jpa, has but very little adversative force, and is often to be translated not ' but,' but ' and.' A decided Adversa- tive is nava, but. These two together pa nava, commonly written in one word, make a strongly Adversative Conjunction 'but not- withstanding.' Nan i Qat me ronotag, wa ni me gilala ran si ratatasina me lav rasoana wa naaJcana, pa sin siwo ga2)lot, nava gate lai, ape wot naraga me rip gese ; pa ni me galoi matila sin siwo, pa ti tantan gese, Then Qat heard it and knew that his brothers had caiTied off his wife and his canoe, and (j^>a) he would have got quickly down, but (nava) he could not because the boughs of the nutmeg tree had all swelled to a large size, and (jm) he tried in vain to get down, and {pa) does nothing but cry. 3. The Conjunction nan is a connective in narration, without any logical force or sense of time. It may begin a narration like * now ' in English ; and it may be convenient to translate it ' now ' or, as above, ' then.' 4. The Disjunctive is si or, Hoke si Hone this or that ; and may be repeated as an alternative, si Hoke si Hone either this or that. 5. The same word si is the Conditional si if, and ' whether.' Si is not always expressed when the Verbal Pai-ticle is qe ; qe toena qarig if it rains to-day. 6. Another sense of si is illative, 'that,' 'in order that;' ni me siwo i lau si ni, or sin, sugsug aia, he went down to the beach that he might bathe there; or with a and the Infinitive Verb si a sugsug to bathe. This also is declarative like ' that,' neira me ga~ ganag si ni me siwo i lau, they said that he was gone down to the beach ; nau we nonom mok si te tamaine, I think that it will be so. Mota. Conjimctions. 301 In narrative si is used as if a clause was understood before it ; / Q^al iloke ni gale toga ran ma, nava si ravevena apena, This Qat was not from ever- lasting, but (the story goes) that he had a mother ; nan i Maratoa me tvurvag gaplot kel mulan o aka, si me tvurvag mun napisuna, so Marawn quickly repaired the canoe again, (they say) that he repaired it with his fingers. In another use it seems expletive, but is idiomatic, ira tatasina we maros we la naakan Qat mun akara, toa irasoana si mun rasoara, his brothers wanted to take Qat's canoe for their canoe) and his wife (as if) for their wife. 7. In quotation tlie declarative si is used, but with tlie sign of quotation wa ; ni me vet iva si ni ie van ma he said that he was coming ; wa si is sometimes was : wa interposed shows that there is something more than the decharation of a fact, that the words of the speaker are quoted. When a mere fact is declared wa cannot be used ; it is incorrect to say neii'a me ilo wa si ni me mate veta they saw that he was already dead. The mark of quotation wa cannot be called a Conjunction. When the quotation is direct wa alone is used ; nan ni me vet mun ravevena tva, Veve, van ma savrag inau ilo lama then he said to his mother (wa), Mother, come and throw me into the sea. If the quotation were indirect it would be ni me vet mun ravevena wa si ni van ma he said to his mother that she was to come. There is no need for a Verb; nan neia mun rave- vena wa then (said he) to his mother that — . 8. The same sound {loa) is used again in a third and distinct sense, that of ' lest ; ' ilogoro ko wa masu take care lest you fall. The same is used as a word of warning like ' ware,' wa iniko nan o vat get out of the way of the stone, ware stone ! wa iniko take care of yourself, out of the way. 9. To express cause there is no Conjunction, but a Substantive manigiu is used, which may be translated ' cause,' - ' reason,' or 'purpose,' and stands therefore for the English ' because,' 'therefore,' ' wherefore,' ' in order that,' ' by way of.' For example, nau gate ronotag, manigiu ni gate vava jpoa I did not hear» because^ (the cause) he did not speaJ? loud ; manigiu sava ko me ge ? manigiu nau we maros why did you do it ] because I like ; mani- giu (or manigiu) o ine therefore, wherefore, on that account ; ko me kos gajma ? Tagai, manigiu apena did you ill treat him with- out a cause? No, there was a reason for it; ni me vega kalo o manigiu sin ilo mantag he climbed up in order that he might see well ; manig gale by way of, with the purpose of, deceit. 1,0. Till, until, is gai ; nau tete loloqon laiko gai mate I shall not forget you till I die ; nau me mamata vires gai matava I was awake all the time till morning. 302 Melanesian Grammars. But this Conjunction is often left out : nau me mamata maran I was awake (till) morning, kamam me ur qon we have done nothing (till) night, i.e. all day. In many cases in which ' and ' is used in English an expression is used in Mota which must be mentioned in this place. It is when, for example, persons or things which can be looked upon as companions or fellows are spoken of: I and my brother, tak tasik, Peter and John, Feter tana John, you and who besides ? tama isei ; a man is standing between the almond tree and the breadfruit tree o tanun toe tira alo masaoi o nae tan o patau ; we and our fathers taniam tamamam. It is plain that it is in fact a Noun ta, which may be translated companion or mate, with the Suffixed Pronoun : tak tasik my companion, my brother, Peter his mate John, who is your companion? and so on. In the Plural there is a little difficulty, tamam tamamam we and our fathers, i.e. our fathers our companions, we and our fathers with us, tara tasira they and their brothers, their brothers with them, tamiu rasei ? who with you ? In the Singular there can be no confusion, but in the Plural it is not easy to perceive whose companions are spoken of, whom the ta represents ; and in consequence tanina, tamam, tamiti, though undoubtedly correct, are by some natives viewed with disapproval, and there is some divergence of opinion concerning the Third Person Plural. Is it correct to say tarasti or tara rasei for 'and some persons,' taragai or tara ragai for 'and those persons?' to say ira manua tara gopae the orphans and sick persons, or ira maniia tara ra gopae ? There is no number in the Noun ta ; it may be companion or com- panions : number is shown by the Suffix, tak my ta, tama thy ta, tana his ta ; I, thou, he are one person, but tak does not tell how many are my com- panions; if I say inau tara I call myself their companion, not them my companions. When tara is used the ta is the companion or companions of many, not one, of them. If then the phrase is o manua tara gopae it is evidently correct if the orphan is one, it is the orphan the sick men's com- panion ; if it be ira manua tara ra gopae it is equally correct, meaning the orphans (and) their companions the sick men. Tak is a common expletive, esi tak ! but the meaning is ' my mate.' XI. NUMEEALS. The numeral system in Mota is imperfectly decimal ; tliat is to say, there is a word for 'ten,' and when ten has been reached all higher numbers are expressed in tens, but there is not a series of independent numerals up to ten. The basis of calculation being the five fingers of one hand, the first five numerals which belong to one hand are repeated with variation for the numerals belonging to the second hand up to nine. 1. Cardinals. — These as in ordinary use are — One tuwale, two nirua, three nitol, four nivat, five tavelima, six laveatea, seven lavearua, eight laveatol, nine laveavat, ten sanavul. It is seen at once that the true numerals two, three, four, are Mot a. Nttmerals. 303 r«o, tol, vat ; wliich appear again in seven, eiglit, nine. Three is often tolu, and four vati ; as in counting, rua, tolu, vali. The Prefix ni is in fact a Verbal Particle, and can be applied also to tuwale, tavelima, and sauavul. The Particlea in common use with Verbs, except we, fi, are applied to all these Numerals, taking of course the place of ni with rua, tolu, vat : me tuwale it was one, te tuwale will be one, ta tuwale let it be one, qe tuwale if it be one. In this way Numerals are constantly used in Verbal form, as Adjectives are. It may be said that rua, tolu, vati, are hardly ever used except in a Verbal form. The other Numerals counnonly appear as Adjectives : o tanu7i imvale one man, o q'oe sanavul, tavelima, lavearua, ten, five, seven, pigs. Sanavul may be said to be naturally a Substantive. The meaning of the Prefix lavea in the Numerals of the second hand has not been made out. In laveatea six, tea is the Numeral tea, one, obsolete in Mota, but in use elsewhere. The Prefix tave in tavelima, five, has not been explained, but may be tavaliti, taval lima, hand on one side. Twenty, thirty, forty, sanavul rua, tolu, vat, and so on up to ninety sanavul laveavat. To mark exactness of reckoning wonowono, full, is used; sanavul tuwale, rua, wonowono, full ten, twenty. For the units above tens a substantive is used, o numei ; eleven is sanavul tuivale o numei tuwale one ten, the unit above it one ; twelve numei nirua, and so on. It is not necessary always to mention the ten, o numei nirua by itself will signify twelve. The Verbal particle may equally be me, ti, ta with the numei ; sanavul tol numei me vat thirty-four, when past time is declared, ' three tens the number-above was four.' A hundred is vnelnol, that is, a whole mele leaf. This may be made a Vei-b, qon me melnol the days were a hundred ; or may be used as an Adjective o qon melnol a hundred days. The sum above a hundred has the name avaviu, from the Verb av to put one thing upon another ; a hundred and three vnelnol tuioale (or vatuwale) o avaviu nitol, two hundred and fifty melnol vagarua o avaviu sanavicl tavelima, twice hundred, the sum-above- hundred five tens. A Thousand is tar. Substantive, Adjective, or Verb ; tar tuwale one thousand, tanun tar a thousand men, o vetal me tar vagavat the bananas were four times thousand. The exact use of tar for a thousand is settled in native usage as ten hundreds, but the word is also loosely used for any gi'eat number. There is no name for a sum greater than a thousand; tar mataqelaqela is a very large indefinite number, mataqela blind. The Mota numeration thus exhibited is clear if lengthy : 1884 is tar vatu- 304 Melanesian Gramma7's. wale, melnol laveatol (or vagalaveatol), avaviu sanavul laveatol, mimei nivat, thousand once, hundreds eight (or eight times), the sum-above-hundreds eight tens, the unit-above four. 2. Ordinals. — The Cardinals assume a Substantival termination in becoming Ordinals, and the second, third, ajid fourth take also a Causative or Multiplicative Prefix; varuei or vagaruei second, vatoliu or vagatoliu third, vavatiu fourth, taveltmai fifth, laveteai sixth, laveruai seventh, lavetoliu eighth, lavevatiu ninth, sanavuliu, or -ui, tenth. Though the terminations are those of Nouns it cannot be said that the words are generally used otherwise than as Adjectives, o qon vatoliu the third daj^ It is, however, equally correct to say qon vatoliu ; and sanavului, or -?«, is the tenth, and the tenth part. It must be observed that the Causative vaga or va replaces the Verbal ni in second, third, and fourth, and that a is dropped in sixth, seventh, eighth, and ninth. There is no Ordinal Numeral in the first place ; the word in use for 'first' is mocd. For 'hundredth' the expression is melnol anai, anai being the same word explained under Possessives (p. 272) as meaning ' belonging to.' To express such Ordinals as twelfth, twenty-fifth, where the nnmei occurs, the Numeral belonging to the numet alone becomes an Ordinal : twenty-second sanavul tuwale o numet varuei, twenty-fifth sanavul rua numei tavelimai. The same applies to Ordinals above a hundred, -melnol vatuwale o avaviu tave- limai hundred and fifth. There is a word, tuara, which may be generally translated * other.' If there are two things, either of the two with reference to the other is tuara : tamafe ti Jcur tuara gaviga, ti la mun o tanun o t^lara the ghost eats one of the (two) gavigas, and gives the other to the man. So tuara is the next, alo tau tuara next year, not this, the other. The word also, like ' other ' in English, is used indefinitely, alo tuara qon the other day, some day, tuara tanun a certain man, tuarasei one or the other, some one. ' 3. 3IuUi2)licatives are formed with the Causative Prefix vaga or va ; vatuwale, or vagatuivale, once, vagarua twice, vagatol thrice, vagasanavul ten times, vagamelnol a hundred times. 4. A word which cannot be translated in one English word is visa, which means interrogatively 'how many?' and indefinitely 'so many.' This has the Verbal Particle, 7ii visa? ho\V many? ta visa let them be so many, me visa ? how many were they ? te visa ? how many shall there be ? As a Multiplicative vagavisa ? how often ? how many times 1 5. When persons or things under certain cii'cumstances are reckoned the Numeral is not simply used, but is introduced by a word wliich more or less describes the circumstances. If ten men Mota. Exclamatiojis, Expletives. 305 are spoken of regarded as in a company together it would not be o tanun sanavul, bi^t tanun 2)ul sanavul, 2>ul meaning to be close together ; ten men in a canoe are tanun sage sanavul ; sage visa ^ avune aha 1 how many in the canoe % sage meaning ' up' or, as we should say, ' on board.' Using visa instead of a numeral, so many things in a bunch together are sogo visa ; gaviga sogorua two Malay apples in a bunch ; arrows are lira or tur visa standing so many, canoes sailing also are tira visa ; bats taqavisa hanging so many ; money is reckoned tal visa string so many ; so many things or persons at once are sorako visa ; ni me tiqa mes sorakorua he shot two parrots at a shot, ratol we vavava sorakotol the three men are talking all three at once. With something of the same manner of speaking two canoes sailing together are said to sail butterfly-two, we gamo pe2)erua. The only use of wo with a Numeral is with tuwale ; natuna wotuwale his only child. 6. It is very remarkable that in counting in the game of tika, and in that onl)', a regular decimal series of Numerals is in use ; one qasa, two ivura, three lovi, four tama, five rina, six qani, seven lini, eight 7iaga, nine viga, ten tvesu. These Numerals appear to be peculiar to Mota, and of native invention, none of them corresponding to the Numerals of other islands. Of the words, wiira, two, may be the same as that used in Dual Imperative ; wesu, ten, means arriving at the conclusion ; qasa one, qaru six, seem to refer to the first and second hand respectively, sa being elsewhere ' one : ' but the use of a pure decimal notation, not of foreign origin, in a game, by people who in common \ life use one of a quinary character, is very remarkable. XII. Exclamations, Expletives. 1. Exclamations are some of them such cries as naturally or conventionally express pain, pleasure, surprise, and other emotions of the mind ; and though they are not uttered indiscriminately, they are incapable perhaps of exact definition as to meaning. Such as these are e !, ke ! which express a general emotion rather of admiring surprise or satisfaction. With e ! are combined other sounds,*e ke ! of admiration, e we ! of disapproving surprise, e qe I more emphatic than e ke !. The simple sound a is not used, but a wo ! is a cry of astonishment and general emotion, a qo ! perhaps stronger, a wa ! of grief or fatigue, a re ! of pain. With e are combined words which give a meaning to the Ex- clamation that can be translated ; e si ! I don't know, si meaning * if,' ' whether ;' e wun ! I suppose so, wun probably. 2. Other Exclamations are more articulate, and become words, X 3o6 Melmiesimi Grammars. To express afl&rmation and assent, besides the upwai'd backward movement of the head, nodding upwai'ds, or a cluck of the tongue, the Exclamation we ! is used ; and various degrees of assent and satisfaction are shown in the character of tone and prolongation of the sound ; wje.may be strong and decided, hesitating and timid, or very faintly acquiescing ; uwe is strong, we-e apprehensive. A Vocative addressed to persons is gai ! very commonly used, and sometimes with a force of asseveration. The prefix ra by which it becomes plural has been noticed ; but gai ! is applicable to one or more persons. Another less common is ara ! but these are not necessarily addressed to individuals. The meaning of alova ! is ' indeed ! ' The Negative gate is used in admiration ; gate tanun gai, Oh ! what a man ! 3. Thei-e are other words which must be called Expletives, which indeed are by no means devoid of signification, and add clearness and vivacity to native speech, yet cannot be classed with any Parts of Speech. Such are wa and qa ; tagai wa no, certainly not, ineia nan qa he, that one, to be sure. It is hardly possible to give in words the force of gina and gita, though they. have a certain demonstrative character, as gitie is ' that ' in Vureas ; avea ko we maros ? ilohe gina, which do you like ? this one ; avea ineia ? alo mot gita, where is he 1 in the bush. Of these two words some natives say gita, some gina, most of them gina. XIII. Example of Mota Nakeative. By way of a specimen of native Mota writing is given the following part of the story of Qat. It was written by the late native Deacon Edward Wogale. O Kakakae apen Qat. I Qat iloke ni gate toga ran ma', nava si ravevena apena, wa nasasana iro Qatgoro ; wa iravevena ilone si o vat me mawora iiania, pa gate gilala^ si o vat we savai ilone. Wa ira tatasina mulaH apena, tuwale nasasana i Tawaro Gila- gilala, ni we lolomaran' ape savasava nan, pa ni we gaganag lue mun neira apena, wa o varuei nasasana i Tawaro Loloqow, ni we loloqow* o savasava, pa ni we gege loloqoM. Ineira sawavul tuwale o nuwei nirua, we log tataga o no-tangae ta Mota gese ; o varue nu>nei ^ i Qat. Wa ineira me togatoga a Vanua Lava, alo vatiu o sasai alo Sepere"; wa neira me togatoga aia i Qat qara iintbi o savasava. Pa ni me ti« paso o savasava nan, nava ni gate gilala we tin o qoii, pa o maran me marmaran gese''. Naw ira tatasina me vet munia wa, Gai, Qat, gate wia nake gai o mai-an vires, ka ge gasei mulan ti* nia ; paso nan i Qat qara sike sin te ge o sava ape o maran ilone. Na« ni me ronotag si o qoH a Vava, ti ligo raka o rawe, ti map alo aka, ti gawjo i Vava, Mota. Narrative. 307 ti tun" o qon nia mim i gene ta aia'". Wa ni me la mulaH nuinia o toa sin te gaganag o iiiaran si me maran kel mulaw. Paso nan ni me kel ma siir ratata- sina, ti vet muu neira wa, Ilokenake tur ge taurmate o tanoi kamiu te risa sur ilolona ; na»« neira me ge o no-matig, me van, me wosalag niantag natanora MaM. Na» i Qat me varus neira wa, Kamiu me taurmate paso, ragai ? Naw neira wa, We, me paso. Na« i Qat qara ukeg o qo» sin qo»''. Na» ti vet muneira wa, Kamiu qe ilo o nago vanua qe savsavai, pa ineia veta nan, wa tur risa sur siwo ilo tano epamiu. Naw neira wa, We-e'^. Paso nan neira me ilo me silsiliga ; na?i neira wa, sava nake, Qat ? Na» Qat wa. Pa ineia veta nake"; wa kamiu qe roHotag o sava ape matamiu tur risa rorowo nan. Ni me vet taj«aine ape matamaragai, ape neira gate gilala o matamaragai. Naw me silsiliga mantag, neira me rojiotag namatara we maragai ; wa neira iiie vet mun Qat wa, Qat, o sava me ge namataiaam ? Na*» neia wa, Hone o gene nau me vet ti " mun kamiu apena, tur risa roroMo mantag, wa vataqav namatamiu, tur matur. Wa neira me matur ta>Ka i Qat me varegira ti. Naw me qou maul paso, i Qat me la o mavin'^ memea, me teve o qo?t nia, wa o maran qara rowolue mulaw ma, si ape o qo« me gara goroa ti. Wa neira me toga maul alo vauua ilone Lo Sepere; ni me tintiw o savasava aia. Literal Translation. A Story about Qat. This Qat (he) was not from everlasting, but (thej say) that he had a mother, and her name was Qatgoro ; and that mother of his (they say) was a stone tliat burst asunder from him, but it is not known what sort of stone that was. And he had also brothers ; one his name was Tawaro Gilagilala, he understood about all sorts of things, and made known to them about it ; and the second his name was Taiiaro Loloqon, he was ignorant of everything, and he behaved like a fool. There were twelve of them, all called after the leaves of Mota trees ; the twelfth was Qat. And they lived at Vanua Lava at a place the name (of which) is At the Sepere, and (while) they were living there Qat began to make all sorts of things. And he finished making all kinds of things, but he did not know how to make night, and the day was always nothing but day. So his brothers said to him, Hallo Qat this is not good, this nothing but daylight, just manage somehow again about it. Then Qat sought what he should do about that daylight. Then he heard that there was night at Vava, (and) to begin with, he ties up a pig, puts it in the canoe, sails to Vava, buys night with it from a person of the place. And he gave him also a fowl to make known the daylight that it was light again once more. After that he came back to his brothers and says to them. Now get ready a place you will lie down into ; and they got cocoa-nut leaves, plaited them, spread well their places (with them). Then Qat asked them. You fellows, have you finished getting ready ? Then they (said). Yes, it is finished. Then Qat let go the night that it might be night. Then he says to them, If you should see the face of the land should be something or other, but that is it already, and lie down on to your mat-places ; then (said) they, Ye-es. After that they saw that it had become dark ; and they (said). What is this, Qat ? Then Qat (Isaid), But this is it abeady ; and, if you should feel anything about your eyes, lie quietly. He spoke to them thus about eye-quivering, because they did not know eye- quivering (sleepiness). Then it became perfectly dark, they felt their eyes quivering, and they said to Qat, Qat, what has done (something to) our eyes ? X 2 3o8 Melanesian Grammars. And he (said), That is the thing I spoke to you about, lie perfectly still, and shut your eyes, sleep ; and they slept as Qat had ordered them. Then, after it was long dark, Qat took a red piece of obsidian and cut the night with it, and the daylight thereupon came thi-ough again, (they say) because the night had come down upon it. And they lived a long time in that place, Lo Sepere ; he made all sorts of things there. Notes. — ^ toga to abide, ran right throughout, ma hither. ^ Impersonal use of the Verb. ^ we lolomaran has become a Verb, but it would be better Mota to write nalolona toe maran the inward part of him is light. * Similarly nalolona we qou ape savasava his inward part is night about everything, is better, but the metaphor has become a Verb, which takes the object without a Preposition. ^ The second number-above-ten, so the twelfth. ^ The names of places are generally 'At' a tree, or some other natural object, see p. 162. '' maran is both light and day. Observe that maran is reduplicated to show prolongation, and becomes a Verb, ' day only dayed.' * This is the ti of civility. * tun is to buy with a great price. ■" This Person, a Vui like Qat, not a man but a spirit, was i Qon, his name was Night ; ^' qon becomes a Verb, 'that it might night.' '^ This shows a little apprehension, hesitating assent. '' nake spoken of a thing already present, ' this,' corresponds to nan, nane, said above, and again below, of the thing still future, not here, therefore ' that : ' the demonstrative gives emphasis. '* ti throws the time of his telling them back beyond that of the sleepiness coming on. " Obsidian was used to cut native string and threads; the Mota idiom says that the dawn cuts, not breaks, o maran ti teve, XIV. The Song Dialect. Mota songs are never made in the language commonly used ; nor is their language that of any neighbouring place. Tliis holds good of the Banks' Island songs generally ; they are always in the song dialect of the Island or district in which they are sung, and that is never the sj)oken language of any other part. The language of the Mota songs resembles that of Gaua, in Santa Maria, on the one side, or of J/otlav on the other ; but the Mota song is not in the Gaua or J/otlav spoken language ; the example of the Bishop's song here given shows that a J/otlav song, in the song dialect of that place, when sung in Mota differs from the same when sung in jl/otlav. There is, that is to say, belonging to each Banks' Island language its song dialect. It would seem natural to assume that this song dialect is an archaic form of the spoken language, but thex'e is probably nothing to prove this to be the case. The characteristic differences of the Song Dialect from the common speech of Mota may be seen in these examples to consist (i) in the casting out of vowels, and consequent contraction of the words, (2) in the occasional addition of a final vowel, (3) in the use of words not used at all or used differently in common language, (4^ in the imitation of foreign forms. Examples may be Mota. Song Dialect. 309 Been of (l) in nalnik, nasrik, for na lawik, nastirik, in Song i ; of (2) in mae for ma in both songs, ree, lumagave, in Song 2 ; of (3) in nirmanior the Third Person Plural Pronoun, and the Verbal Particle se in Song I, sa for sage, and the Verbal Particle e in Song 2, sur for ape in both ; of (4) in ice and tvu in place of the Article in both, Vano lave, and na as the Article in Song i, vonue for tanua in Song 2. The contractions and elongations are no doubt adapted to what may be called the tune. Besides this there is the change of a to e in lave, lame, wore, &c. To compose a song is to measure a song, ive foivo o as, an expression which Bhows that there is a fitting of words to a sort of metre. This is not a measure of lines or number of syllables, but an adaptation to a sort of tune. To sing a song is loe sur o as, sur being the word which when reduplicated means to sew, and the notion that of drawing out a thread stitch after stitch. The string of vowels at the beginning sets in some way the tune or the character of it ; and a new strain of different character is introduced in Song 2, in the same way. The song is called the song of the person who is the subject of it, na-asina; it is the 'measure' of the poet who composed it, na towona. A third person will give money to the poet to compose a song in honour of the subject. To make a song about a person is to fara him. The Bishop's Song was composed in honour of Bishop Selwyn the elder, by a ilfotlav woman. The Mota version of it here given difiFers from the original, which may be seen at the end of the ilfotlav Grammar, but it is in the ilfotlav style. The other song was composed by a boy at Norfolk Island, and written down by him : it is after the Gaua fashion of singing, but not in the Gaua song dialect. I. Naasin Besop. Oeoewa ! wu roro sa 1 naroron i Besope ni gajw tal na Vano lave ; naljiik na laM lave, nasrik na ar ilferlav, ni se turtur ale lame ; gis nok melov ok ; melov rer me rere levran Pohenqon, nam loslos wore sur na te mul XJlsilane, ro Ti«gormew se tur gor norue. Oeoewa ! Wu roro la»^ ni se lul ma ale lame, wu roro ak Besop ga»j mae, naroron i wowut ga?» ma ; na me rowtag mas naroro?«, na te ilo nangowia ve ■ Kirman sororo ma napasi nago/wa, na pasi gar manuma, ni se rer le varan Kelepe. Nam tantaw, nam loslos wore sur na te mul Ulsilan. Translation. The Bishop's Song. Oeoewa ! News of what ? news of the Bishop, he sails round Vanua Lava. My wind is a great wind, my bcnes are the Casuarina tree of Merlav ; he stands in the sea. Oh ! my sailing of the ship ; the sailing on the flow of the sea has flowed into the bosom of Eohenqon ; I am nothing but rejoiced because I shall go to New Zealand ; Eo Tiwgormew withstands us two. Oeoewa ! The noise of the wind ! it has sounded hither on the sea, the news (noise) of the Bishop's ship sailing hither, the news of the hero sailing hither ; I have heard merely the report of j'ou, where shall I see your face ? They have brought the report of the beauty of your face, the beauty of the root of your nose, it has flowed into the bosom of Eelepe, I have wept, I have only rejoiced because I shall go to New Zealand. Notes. — lou for Article ; na for 0, never used in this way in prose ; nalnik — na lanik, in prose 7wk larx; nasurik aru ta Meralava; se a Verbal 3IO Melanesian Grammars. Particle never used in prose, no doubt the Fiji sa\ melovoTc — molovi oka; levran = lo varan ; nam — na me ; sur for ape, use of the Preposition only made in songs; norue = narua, nara ; mas = mamasa bare; na7igoina = na>iagoma; nirmaii a Demonstrative Pronoun only used in songs. The song is from jlfotlav, but is not identical in the form of the words with that used in ilfotlav ; which see at the end of the Grammar of that language. 2. Naasin Me?. Eale ! inam sa, na me ile we ak me sale erow, pe «ere ve ? pe were vate. Me tug lue e rer le varan Mel, te mul soror le vonue, ero vano mae. Jialawale taw lolowonwon sur paka lave, nin te mul vee ? lumagave enin teve laklake, ro» naronse me gam mae, e von naroron kavten te mol kel Nusilane. aieoe, o eieoa, eoeieoe, weae, nin laklak sur we save ? we ak tavea ? we ak ta ion, ni te mul kel. Translation. MeVs Song. Eale ! I was sitting, I saw a ship had floated in the East, at the point of land where ? at the rocky point of land. The flow of the sea drew out into the breast of Mel, he will go and give the news in the village. They two are coming hither. Damsels weep with sorrow for the ship, whither shall we go ? Youths, let us rejoice to hear the news of someone who has sailed hither, to hear the news of the Captain who will go back to New Zealand. O aieoe ! oeieoa ! eoeieoe weae ! what do we rejoice about ? of what place is it a ship ? it is a foreign ship, it will go back. Notes. — inam sa = ioiau me sage, the ilfotlav hag to sit; loe aJc = o aha, Gaua use for the Article ; e row = i roivo ; nere face in Ureparapara, lip, beak, in Vanua Lava ; ve = vea where ; vate, e added, as in wae, vee, below according to the use of songs ; tng to untie a string, used for the flow of the sea ; e rer = orere; le=lo; mul soror = mule sororo ; vonue = vanua; ero = irara; surin prose cannot be used except of motion to a person; enin = inina; laJclalce = Idkalaka, the final a changed to e, as in vonue, &c. ; ron = ronotag ; se = sei; gam. mae=gaino ma; e ron = we ronotag, e for "Verbal Particle; sur tve save = ape sava, we for o. The song is after the Gaua fashion of singing. 2. 1/oTLAv, Saddle Island. The South-western end of Saddle Island, which lies seven miles North of Mota, is if otlav, i. e. Mota lava, Great Mota. Attached to it by a reef is the islet a Ra, in Mota a Rao, in which the same dialect is spoken. To the east of J/otlav is Volow, separated by a district in which the variation of speech is insignificant. On the other coast is Bun, the dialect of which is not very different from that of J/otlav, but is characterized by the change of r to y. Between Volow and Bun there was a dialect resembling that of Volow, and changing r to dh. The little reef Island of Rowa to yiotlav. Alphabet^ Articles. 311 the North-east of Saddle Island has a distinct dialect, not much unlike J/otlav, but changing k to g, and known by a peculiar mincing pronunciation. The diflference of ifota and J/otlav language to the ear is very great, though the peoi)le are the same in race and customs, and the Islands are so near. The J/otlav speech casts out as many Vowels as possible, shortens those that remain, and changes p to mb, n very often to nd. It is what in Mota is called mamnvin, tliin. An example, in which the same words are used in both languages, will briefly show the difference ; J/otlav, nahte metioeh woqtin, mo lolo (/or namtan, Mota, patau me tawosa vaivo qatuna, me lolo goro namatana, a breadfruit smashed on his head, poured over his eyes. A characteristic of J/otlav is the shifting of vowels to accord with an anticipated sound. Such words as Articles and Verbal Particles can have no fixed form, na han, ni oiitvaer, ma van, mo tog : in the above example the Article has a in nabte because that is the Vowel cast out of hte-=^2)atau, it is metiveh because e belongs to fweh=tawasa. In consequence of the shortening of words and the shifting of Vowels it is difficult to write the language clearly, Jfotlav people will write to one another in Mota, because they say they do not know how to write their own language. The examples here are almost all taken from a Phrase-book by Eev. J Palmer. I. Alphabet. The Vowels are a, e, i, 0, u, and a short o. There are no Diphthongs ; the Mota laii, maur, tamve, sao, are le, mir, to, he. The Consonants are k, g ; t, d ; b, v, w ; q ; m, m, n, n ; 1, r ; s, h. There is no p, b = mb replaces it ; Motl&v people in Mota words will put v for p, mav for map; b sometimes represents Mota m, boros = maros. Mota n is often represented by d = nd. Mota s is sometimes h, ih = us a bow ; h closes a syllable. The sound of q is kmbw, because b = mb = p; before a vowel the whole compound can be heard, >u qil. II. Articles. 1. The Demonstrative Article is n — combining with a Noun which begins with a Vowel, and before a Consonant having a shifting vowel in accordance with the first vowel of the Noun ; 7iet, n-et, a man, niJi, n-ih, a bow, nok, n-ok, a canoe : na tar a calm, ne tenge a tree, ni til a certain fish, no totgal a picture, no to a hill, nu bug a debt. 312 Melanesian Grammars. When, as commonly, the first vowel of the Noun with an Article is cast out, the vowel of the Article represents it : ndbte for na hate, namtan for na matan. But as the language loves to shorten vowels, the Article has often a shorter vowel than that which has been cast out of the Noun : gohow rat, naghoio a rat. Sometimes also the Article has a vowel which properly belongs to the Noun, but is shortened when pronounced in it : nagmel for na gemel, Mota gamal. It is evident that because of the elisions the Article must be generally written in one word with the Noun. 2. The Personal Article is i, Plural ir. The Feminine sign is r- with shifting vowel : Ra Las, Re Sem, Ro Q{o)vur, Ri Tit, Ru Bur. /personifies as in Mota : i meren = i gene the person, i hav who ? III. Nouns. 1. There is a like division as in Mota between Nouns which take a Suffixed Pronoun and those that do not. 2. Verbal Substantives. The terminations are e,r,g; mat to die, mate death, tog to abide, natgar way of life, vano go, navnog going. Reduplication often gives the same sense, dem think, nedemdem thought, tab to love, natahtab love. 3. Independent Substantives. The terminations are ge, n ; na Tiege name, stem ha ; nabnege hand, ban ; nenten child, nat ; netlen egg, tel. This termination has no place in composition ; nabnege, nahne men bird's wing, nelwege, nelioo eg fish's tooth, nahege, nahe et a man's name, nahan his name ; the termination of the first member of the compound is lightened. The true form of the word meaning tooth is leivo, of that meaning name ha ; the sufiBxing of _$re shortens o and a to e, nelwege, nahege. 4. Plural. The sign of Plural is geh, the Mota gese ; na ge geh things, net geh men. The Prefix for persons is ra, ratelki some. That^eA has the same meaning with_9'e*e is shown by Icemem to ^lotlav geh we are all J/otlav people and no others ; del, Mota nol, is ' all' in the sense of totality, but is also used with the sense of plurality. IV. Pronouns. 1 . Personal Pronouns. Singular. 1. ino, no. 2. inek, nek. 3. ike, ke. Plural. I. 2. 3- inclusive, iged, ged. exclusive, kemem. kimi. iker, ker. Dual. 1. incl. doro. excl. kamamro. 2. komro. 3. koro. Trial. I. 2. incl. detel. excl. kamam tel. kemtel. kertel. Vioilav. Pronouns. 313 Observations. — i. These forms are all used indifferently as subject and object, and in indicative or subjoined sentences. 2. The Prefix / gives more personal emphasis. 3. In the Third Person ke is a demonstrative ; r (Mota ra) makes the Plural. 4. The Dual and Trial are the Plural with the Numerals ro and tel. The vowels belonging to the Pronoun shift in accordance with those of the Numerals do, ko, with ro ; de, ke with tel. Why it should be kemem in the Plural and kamam in the Dual and Trial cannot be explained. The inclusive First Person Plural ged gives only d to form the Dual and Trial, showing the true Pronoun (see p. 119). 2. Pronouns suffixed to Nouns. Singular, — i. k; 2. (m) ; 3. n. Plural. — I. excl. mem; 2. mi; 3. r. Examples. — nalnege a hand. Singular, l. nabnek, my hand. Plural, i. I'wcZ. nabnenged, our hand. excl. nabnemem. 2. nabne, thy hand. 2. nabnemi, your hand. 3. nabnen, his, her, its, hand. 3. nabner, their hand. na Tiege a name. — Singular: 1. na hek ; 2. na he; 3. na ban. Plural: I. na hanged, na hamem ; 2. na hami ; 3. na har. nangege a face. — Singular: i. nangek ; 2. nange ; 3. nangon. Plural: I. nangonged, nangomem ; 2. nangomi; 3. nangor. Dual: i. excl. nango- mamro; 3. nangorro. Observations. — i. The absence in general use of a suSixed form for the Second Person Singular is very remarkable, and cannot be explained ; as in Volow and Vanua Lava ; it appears only with the Possessives no and go. The Noun in the form to which the independent substantival termination ge is suffixed is, as it were, in a Genitive Case with the Second Singular : niqtige a head, niqti thy head, namtege an eye, namte thine eye. It must be remembered that the termination ge is not always suffixed to the true form of the word (III. (3) ) ; Aa is the stem from which nahege, nahe come, mata that of namtege, namte ; a modification of the termination of the true word makes a Genitive. 2. For the inclusive First Plural there is no other form than ged, though d would seem natural, see Pak ; the others are modifications of the Personal Pronouns. The introduction of n before ged is parallel to the same in the First exclusive in Mota, napanen kamam. 3. If the Noun to which the Pronoun is suffixed ends in a consonant, a vowel must be supplied before k, n, r; em a house, evaar their houses, bug debt, nebgor their debts, nebgun his fault. 4. For the Dual and Trial there are no special forms ; ro and tel are added to the Plural : but mem becomes mam before ro. 5. In the different Persons the vowels of the stem come out differently before the various Suffixes : e. g. ha has been said to be the true word, which, with the independent termination ge, becomes liege; the shortened vowel remains in na hek, na he, the true vowel appears in each of the other Persons. 314 Melanesian Grammars, So the stem of nangege is, in its true form, 7iago, and o comes out in all the Persons except the First and Second Singular. 3. Demonstrative Pronouns. Igol, gol, igoh, goh, hagoh, this. Inen, nen, hanen, that. Examples. — Ne tenge e toe gol hih a good tree this for a bow, non he ne helmismis igoll whose is this knife? navno litvo igoh a large land this ; «om ■malsab inen your garment that ; na hav nen 1 no qo, what is that ? a pig. The Personal Pronoun ke is also used as a Demonstrative, hardly as more than a Particle; wos hir ho no totgal he nail (it) under that picture, the picture there. A demonstrative expression is iJce genha goh ! he, that one to be sure, Mota ineia nan qa ! The Demonstrative which in Mota is iragai, those people, is irge : irge ta Qauro te ter geteg nevaar how do the Bauro people build their houses ? 4. Interrogative Pronouns. Ihe ? he 1 who % Feminine irohe ? Plural irhe ? Na hav ? what 1 a Noun. Examples. — Se me hel ? natga, naghow mo Jcor who stole it ? No, a rat ate it; irhe goll who are these? irohe ma van mel what woman came here? For 'which,' 'whether,' the Adverb ' where ' is used as in Mota : nali naJchi a Del which do you like ? your heart desires where ? 5. Indefinite Pronouns do not appear; ratkelgi some persons, [ra tahelei Mota) a Noun with the Plural ra ; but hav is something, somewhat, as well as 'whatl' and he indefinite some one. The Distributive vel is * each,' vel he, every one, each, velvet kimi each of you. To express an English Relative; ave net nik mi tig he where is the man whom you saw ? you saw him ; net nen nik mo ivoh he that is the man whom you struck, you struck him. V. POSSESSIVES. These are no of general relation, mw with rather the sense of origination, ga of close relation, ma of things to drink. 1. no. — This is never used with the First Singular Sufl&x (see VII. (4) ) : nih mino the ' bow with me ' is my bow. The Suffix m is used in the Second Singular; in the Plural, First and Second, n is introduced. Singular: i. — ; 2. noia thy; 3. non his, her, its. Plural: i. nonged, nonmem; 2. nonmi\ 3. nor. With the Article nonom., nonon, &c. 2. m«. — Tlie meaning is, more distinctly than that of Mota mo, that of a thing proceeding from, rather than possessed by, the person; no me fe iwuh I saw it myself, na mun he mi tin whose making was it. Tlie Suffix m is not used. Singular: i. muk; 2. mu; 3. m«». Plural: i. munged, munmem; 2. munfni; 3. mur. With the Article numith, &c. 3. ga. — In the First Singular this is not used at all ; na kis is used : be his Wotlav. Possessives, Adjectives, Verbs. 315 hinag for my food. In the Second Singular the Suffix m is used, and the vowel becomes o. Singular: i. — ; 2. gom; 3. gan. Plural; i. ganged- ganmem; 2. ganmi; 3. gar. With the Article nagom, nagan, &c. 4. ma. — In the First and Second Singular the vowel is e; mek, me; in the other Persons it is ma ; man, manged, &c. With the Article nemek, neme, naman, &c. The M.oi9,pulai is hlege; nahleh qo my pig, nahle thy, nablan his property. There is not the Mota anai ; net mino the man with me, my man. VI. Adjectives. There are pure Adjectives; et liivo big man, nevsx su small house ; but Adjectives are generally in verbal form ; net nilwo, nem nusu, a man (that) is big, house (that) is small. The Comparative is made by the Prepwsition den : no qo nilwo den nagTioio a pig is bigger than a rat. \\Tien the comparison is of number hev is used : kemem na madeg hev ker we are more than they. For 'fond of man, mansem fond of money; mere possessed of, meresetn rich ; diminutive, depreciatory, var, varsu, rather small ; na mati et a good man, na mali et a bad man. The Terminations g, r, as Mota ga, ra. The Prefixes ta and ma shift the vowel before a consonant, or elide, togolgol straight, netgolgol is straight, ntowor, nemwor broken, malaklah happy. VII. Verbs. The Tem2)oral Particles are n-, m-, t-, with shifting vowel. There is no Modal Particle of the same kind. 1. «-. — This is the Indefinite, na hng sits, ni tig stands, ne het is bad, no goh is sick, the vowel shifting with the vowel of the Verb. When the Verb has two or more syllables the vowel of the first is elided and is represented in the Particle, nebros from boras, being, as in Nouns, shortened also. The Particle coalesces with a Verb that begins with a vowel ; nin rather than ni in drinks, net sees. 2. m-. — The Past Particle is similar in all respects ; ma hag sat, me ten cried, mo tog abode, mi tin made, mu wuh struck ; meslon put lengthways, salon ; met saw, et. 3. t-. — The Future Particle is similar ; ta hag will sit, te gen will eat, to sok will seek, tu rmo will plant; si he tes nek, nek ta mat ae if it should pierce, es, you, you will die of it. It follows from the shifting of the vowel after t- that there is no distinction of te and fi in sense as in Mota ; there is only the future t-, none of continu- ance ; nagveg teticag the Malay apple flowers, Mota gaviga ti tawaga ; na ro . lak ti nin te her lokse the banyan leaf falls, (and) buds again. Besides these regular Particles there are two others used only with the first and third Persons Singular, k, and ni. 4. k. — After the Pronoun this has the deceptive appearance of the Possessive nolc, which, perhaps on this account, is never used ; nok hag I sit, noh tav I 3i6 Melauesian Grammars. Bpeak, nok boros I like; but the Ureparapara form ha makes it plain that it is a Verbal Particle used only with the First Person Singular, and with a present signification. 5. lit. — This again has the appearance of the Indefinite n-, but is shown not to be so because the vowel i remains before the other vowels in the Verb, and does not shift; Ice ni liag he sits. This, being used only with the Third Singular, appears like the Mota Pronoun ni, but must be said to be, what it is also in that language, a Verbal Particle. 6. e. — With we, good, e is used ; ne tenge e we gol hih this tree is good for a bow. The use of it in other languages makes the use of it here likely, but not with one word only. 7. In Conditional or Potential clauses the Future t- is used, or rather it is from the shifting vowel impossible to distinguish a separate Particle like Mota ta ; no tet ke, no te lev hir he if I should see him I will give (it) to him. The Particle wo is com- monly used, but is not a Verbal Particle, so no wo met he if I should see him, or noh wo et he. 8. The Particle which marks a Pluperfect is to ; he me lev loh me no boh he mevtavtah to alon he brought back the book he had been reading in. The same Particle signifies something remaining; misu lev to there is still a little. The same also is used in a civil way of speaking ; van to me, van me to, pray come here. 9. Without a Verbal Particle the Present or Indefinite tense is used in all but the First and Third Singular, in which h and ni are used ; nek hag thou sittest, ged vav we speak, inclusive, kemem sok we seek, exclusive, kimi in you drink, her gen they eat. 10. In the Im2)erative the Verb has no Paiiicle; van m,e come here ; or the Number and Person are marked by a form of the Pronoun ; neh van Second Singular ; ami van Second Plural ; amru van Second Dual. The Particle ni is used, as in Mota, ke ni van let him come. 11. A sort of Infinitive is the bare Verb; neh ma van me so mugumugu you have come here to work. 12. Suffixes to Verbs. Consonantal — g, maneg to convey mana ; n, salon to put length- ways ; r, vetgir to set on end, va-tig-ir. Syllabic — heg, borheg laugh at ; teg, 7'onteg hear ; geg, evevgeg throw away ; veg, matveg die with. The last example is of the separable rag of ^lota, no qohig mitrvej ne he the cask is full of water ; but there is another termination not of this character, though of similar force, ter: ne wet tenge memlaticr he the branch of the tree broke with him. See Volow, Pak. yiotlav. Verbs, Adverbs- 317 13. Prefixes. — i. Causative, va, ve; uacA make live, re^jriV make stand. 2. Of Condition as mentioned with Adjectives, m-, t-; nemwur is broken, netgolgol is straight. 3. Spontaneity, tav ; ma tatavser has come undone, (Mota me tavasaru, reduplicated) 4. Reciprocal, ver; vei^vav talk together. The auxiliary Verbs as they may be called appear as Prefixes ; he (Mota so) henege set the face (Mota sonago) hatig stand up ; tig (Mota ti) tigkele turn the back. 14. Imiiersonal Verbs as they may be called do much of the work of Passives, (see Mota) ; ne tenge meslon wolwol avwo be the tree has been laid lengthways across the water, me vetgir netrag hem the ladder has been set up against the house. But the sense need not be Passive ; tak na hav ae ? tak na gargar ae, ta 7na.1v ae to do what with it 1 to do the skin disease with it, that it may Ileal with it ; t- the future Verbal Particle. 15. Reflective Verbs — lok back is used; mu ivuJi he ? si ke mak mat lok ke ? was he killed ? (impersonal) or did he kill, (make dead) himself? 16. Negative Verbs. — The Negative Particles with Verbs are two, et and te, which come before and after the Verb ; no et horos te I don't like, no tit moros te I shall not like ; no et et te ke, ne tenge ni tig gor I did not see him, the tree stood against. The first Pai-ticle goes with the Present and Past, as the Mota ga ; in the Future the Verbal Particle t- precedes it. It may be doubted whether e in et is not itself a Verbal Particle. The Cautionary or Dehortative Particle is tog ; tog hagliag ho nen don't sit down there ; ni tog let it not be, is in form a Verb. 17. Reduplication is simple because the love of short forms and elisions only allows of the repetition of a syllable, haghag, tenten. The notion of repetition and continuation of the action goes with the circumstances. The Verb without the Prefix is reduplicated metegteg, fear, takut Malay. VIII. Adverbs. 1 . Adverbs of Place — igok, inen, here, there, are the Pronouns this, that ; me ' hither ' makes with re ' where,' the equivalents of ' whence ' and ' whither;' luk back, nok lit lok I return back ; hir, ho (Mota siwo), down, van in a certain direction, are pure Adverbs. Others are Compound Adverbs, Nouns with Prepositions; ae there, ave where, amag before ; some Nouns without Prepositions vawo, wo, above ; some in origin Verbs, beten near. 3i8 Melajtesian Grammars. Examples. — Nek met Tee a ve? Van gin Where did you see him Over there; nek me are? whence are you? lit. you hither at where; namtehal gol van are 1 where does this path go to ? rav Mr ho na gaban pull down the sail ; ko lu ne lew pull out the tooth ; he mevher lu ne tern in ? who threw a stone through the glass ? no lo alge the sun above, a lege ; mok aslil put it outside, a selil ; ke na hag aslil he sits without, in the space outside ; evergeg aqut throw it away at the back of the house, a qut ; hag leten me sit near here. The Adverb ae is used as 'thereby,' 'therewith,' 'thereat:' mi ti nu hul, nem memreren ae a candle was lighted, the house became light thereby ; ave naqrin nek mu wuh ke ae ? where is the club you struck him with ? therewith ; lev me ne qet lot, nok bus no lot ae give me the pestle, I will pound the pudding therewith ; ke ten ae he cried thereat. '- 2. Adverbs of Time — qirig, a qirig to-day, righagoh now, {rig little, hagoh this,) anor yesterday, anereh day before yesterday, aneh of past and future time, {nor, ereh, neh, Nouns,) talow to-morrow, oreh day after to-morrow; ne fe formerly, Verb, Mota we tuai; rigrig soon, to still, teqe yet, qoro thereupon. Examples. — Terse ker neA Jem? when will they be paid for the house? ke ma van me aneh ? anor, when did he come here ? yesterday ; na ma vav vagsiso vatag I have often said so already; no met ke to ne te a M-ot I had seen him formerly at Mota. 3. Adverbs of Manner — qele as, qelegoJi, qelenen, thus, so, as this, as that, qele ave ? how 1 as where 1 Jiethet badly, reduplicated Ad- jective ; galsi well, hoqur without due cause or consideration ; ae ' there ' is used like Mota apena ' on that account ; ' the Noun ma- nege, cause, makes ' why ' 1 and ' because.' Examples. — Ker mo ho a qirig na manege beg they paddled (went out in a canoe) to-day on account offish, literally, the cause about fish; na manege na hav nek makf why did you do it? nek mu wuh bel net ba havl why, about what, did you murder the man ? no tu rutv geteg ne tenge ke ? how shall I plant this tree ? ker mak qele ave ne qen ke ? how do they make this net ? Ara na hag lolha den na T'no Lav qele so Mot a Vlow Ara lies (sits) far from Vanua Lav as if Mota from (at) Volow ; ker ma hag bat lo toti tenge they sat in hiding, out of sight, by the stem of the tree ; mo mu tcuh ke, ke qoro tenten ae I beat him, he thereupon cried because of it. 4. Negative. — This is not an Adverb but a Noun ; natga or netga No, i.e. na taga the nought ; nek mo tog to ale aqirig ? Nataga have you been at the beach to-day 1 No. ' Nothing' is nat- or net- hav, na Article, t- Negative Particle, hav somewhat; nek me rev neg veveh ? nat hav, how many fish have you caught ? None. A word, veh, by which 'can' is expressed, must probably be ranked as an Adverb, no et rav veh te I cannot write; it is the Mota wesu, to reach com- pletion. yiotlav. Prepositions. 319 IX. Pkepositions. These are Simple and Compound. 1 . Sim'ple. — Locative, a, I- with shifting Vowel ; of Motion and Dative, hir ; of Motion from, den ; Motion against, gov ; Instru- mental and Dative, mi, ge ; of Relation in Place, t- with shifting Vowel ; mi of Persons ; of general Relation, h- with shifting Vowel. Examples. — i. a. — This has been shown in Adverbs, are, aqirirf, aslil, and with names of places, a Mot at Mota, a Ma. In accordance with the idiom explained pp. i6o, 292, a is translated by 'from,' ke mi sis gor uvivo em he fell from on the house. 2. l: — This is no doubt in origin a Noun meaning inside ; it either coalesces with a Noun beginning with a vowel, lem in the house, or shifts its vowel to match that of the Noun when it begins with a consonant, lo tot tange at the tree trunk ; or it takes the vowel which the Noun casts out of its first syllable, generally shortened or modified as e for 0, levno = l-vono in the place ; nu hav letber'] what (is there) in the dish? Ice ma Jcal le ternin he climbed up into the window ; neg ma hat' le qen the fish drew into the net ; nek lin ne le letno sitgsttg pour the water into the washing-tub {tano the place of anything, receptacle) ; ke a ve ? ke lem where is he ? he (is) in the house. 3. hir. — The Mota sur with wider application. Dative, nek lav hir ke give (it) to him ; Motion, van hir ke go to him. 4. den. — Simply 'from;' lev naqrin den ke take the club from him; na Vno Lav et te wiwi te den na 'M.ot Vanua Lava is not far from Mota, ke mi sis gor den nem. he fell from the house. 5. gor. — The Mota goro, explained in that Grammar, p. 293 ; kemem ma Tiag gor nange we sit (have come to sit) before your face. It is used Adverbially as above, sis gor fall against something, tig gor stand against. 6. mi. — The difficulty felt in Mota as to the identity of this word in origin with me is equally or more felt in ilfotlav. Dative, le me mino give (it) to me. Instrumental, mu iciih ke mi na hav ? he was struck with what ? ke ma kar ke mi nih he shot him with a bow, ker ma kar ni siso mi nih nonor they shot many with their bows; no qolug mur mi ne he the cask was full with water. me, probably the same as this, is used at the end of a sentence like Mota nia ; na hav nek mak me ? what have you done with it ? 7. mi. — Relation with regard only to persons. In this way it serves with the First Singular Pronoun instead of a Possessive, 'with me' for 'my;' le me, mi no ke give (it) here, it is mine ; tog mi no stay with me ; ke me gel me no la lantanu he was angry with me about the bees. In nok tar mi no ih I will shape myself a bow, mi no may be equally translated ' for me ' or ' my.' 8. ge. — To these must be added ge, though natives are not clear in their account of it; le me, nok tar mino ih ge give it here, I will cut for myself a bow with it. 9. t-. — Eolation in respect of belonging to a place ; with shifting, assimi- lated vowel, or coalescing ; to Mot of Mota, ta Muetvo of Maewo, Torbarbar of Ureparapara. 10. b: — The Mota^e, but never used, as is the case with I- and mi, with a 320 Melanesian Grammars. preceding; shifting to ha, he, hi, bo, bu, or coalescing; kimi hu me ha qaqare ok paddle here to the side of the ship ; nek ma van me ba hav ? ha maUah, be helmlsmis, what have you come here for? for clothes, for knives ; vah ne tenge hi higi gear plant the trees by the side of the fence ; ke ma hag tab bemtem. he sits leaning back by the door ; mok no ro menmen bekletber {he kule taber) put the cloth behind the dish ; hem, beg, as above h- em, b- eg. 2. Com2}ound Prepositions are those in which a Noun with a Simple Preposition has a sense equivalent to that of a Preposition. Example : avwo he, a xaivo, over the water ; avtvok, on me, shows vawo a Noun. These Nouns are often used as Prepositions by themselves ; ivo, namlig womram a cloud above the earth ; wor na tahge tcorlih spread out the mat to dry on the grass, ke ave ? too ner where is he ? on the ner tree, reh kal womtig climb up on the cocoa-nut tree. The same is seen in lalnek beneath me, my neath ; naqran gohow lalne em the rat's hole is under the house. In leretne the Preposition I- is seen, tiet ni tig levetne ner nabte the man stands between the almond and the breadfruit trees ; ke na hag levetnar ro he sits between them two, levetnanged in the midst of us. Others are, ne met no tog dilnet nem. the wood is round about the house ; nek mo mol dilnet na Vlotvl We, have you gone round Volow? Yes; ke a vel ke ne res taval tenge ho where is he ? he is lying beyond the tree down there ; Norbarbar taval Roiv ho Ureparapara is beyond Eowa westwards ; silt ev beside the fire. Some are Verbs in origin : sureg lege nem throw it over across the house ; hag beten no sit near me ; nek ma ran hog a ve ? No ma va7i hog Melioo, qoro mol me, where did you arrive at ? I came up to Melwo, then came here. X. Conjunctions. The common Copulative is wa, and. The Adversative is ha but ; ke met ke ha memtegteg ae he saw him but was afraid of him; it is sometimes, however, not much more than copulative. One Conjunction si, is Disjunctive, or, Con- ditional, if, Declarative, that, and Illative, that. There are two forms, si and so, but the Vowel does not shift. Sometimes si and so are used together : si so Suwa tet ger gor te ke wa ke ma mat if Sawa had not swum after him he would have died, literally, if Sawa shall not swim and he was dead ; nek me ter naqrin ba havl nok so wuh no qo ae what were you holding the club for ? that I might strike the pig with it ; ave na malsab so nok herl where is the garment, that I may put it on ? Another Conditional Conjunction is wo; nek wo kur ta mat if you eat it you will die. Others are tevJe, tele, lest ; et gor tevle leh take care lest it get dirty ; tevle or tele sisgor (take care) lest it fall ; qoro the Mota qara ; goh until, hag goli den hel sit till the bell I'ings, The Noun manige because of, 'As' is qeJe ; ke ma van qele nek ma vav to he went as you had told him. Mot lav. Nimicrals. 321 The use of a Noun, meaning companion, in a way that must be translated 'and' is the same as in Mota, mefek my companion = and I ; ino metek ithik my brother and I ; matan he he and who besides ? matan he moH who will go with him ? The ma or me here is probably the same word as the Preposition mi. See Maewo. XI. Numerals. 1. Cardinals. — One vitwaw' 'ur sit with ; Motlav ter. 4. Prefixes, i. Causative, va, ve; vaes make to live, ve^jgn make to sit. 2. Recij)rocal ver ; vervus fight, vervav converse. 3. Of Condition, m-; ser to tear, meser torn, wor apart, mowor come apart, le' break, mele' broken : 'a before a, '0 before 0, and so on, 'awag come open. Spontaneity, 'av ; 'avroro hang down of itself. 5. A Reflective Verb is made with kel back ; ten ta ma' kel te he kills himself. 6. Negative Verbs. The Negative Particle is '«; na va 'a I don't go, lok 'oron 'a I don't like, lok 'oron 'a mauas I shall not like hereafter, literally, my heart desires not, le 'a it cannot be done. The Dehortative is no 'og (Motlav rd tog) ; but 'a also is used, 'a tnta makane don't do so. VIII. Adverbs. 1. Oi Place; tite here, tino there; kct, va, ekava ? M'here ? te ka va ? where is he 1 nik va va ? where are you going ? 2. Time; qere te to-day, now, lo qere te to-day past; manas hereafter, lo nas heretofore; aloq to-morrow, lo nonor yesterday, ires the day after to-morrow, lo nores the day before yesterday : qere, was, loq, nor, res, being Nouns with Prepositions. Other Adverbs, ma hither, at outwards ; malcane thus, so ; res, Mota wesu, used as Motlav re/i completely, quite, na va ves 'a I will not go at all, le ves 'a can't be at all ; qal, to strike, in the same sense, na va qal'a I shall not go by any means ; mak of immediate consequence, te mak roix he hears thereupon, or for the first time. The Negative is e 'aga, no. IX. Pkepositions. Locative, a, ^ ; Motion to, sir; Motion from, ew ; Motion against, gar; Dative me; Instrumental, men, me; Relation, general pe; to Persons, ma, me; to Places, 'a. I. a is not common, but it appears in Adverbs, as aloq. 2. lo in ; tan pii' lo en he sits in the house, tern va pa' lo en he has gone {pata inwards, out of sight) into the house. 3. sir to persons, va sir te go to him. 4. e»i from, no doubt Mota nan, Motlav rfen, must have been ten and so 'e« ; le en te take from him, tine vono me te, te va en that is his place (that) he comes from. 5. gor; ta nor gar 'eqe make fence to protect the garden, ten et gor let him look after it, va gor go after. 6. me ; le me te give it to him, same as 9 below. 7. me, men with ; vus te men qoron strike him with a club, tite no qoron tern viis te 'i me this is the club he struck him with ; the same with 6 and 9 below. 8. pe at, about, because of ; pe sav why ? on account of what ? te pe me'e en he Leon and Sasar, Vanua Lava. 2)Z1 is at the door, pe umt because of rain. 9. ma, me, with, near, by, persons : in this language not to be distinguished from men the instrumental. The possessive use has been mentioned, vono ma na my land ; it is ma or me ; ma na, me vili, me te ; mek'nia with us, melc'mi with you ; tem 'o ma na he stayed with me. 10. 'a belonging to a place ; irge 'a Moi' the Mota people : combines with lo and me, these being in fact Nouns, 'alo en belonging in the house, 'a mck'ma belonging to the things with us. Prepositions which are plainly Nouns are ro^o (Mota rawo) on, I'o^fo en on the house ; lalne en under the house ; peri ev beside the fire ; lo lie in the midst ; ne represents tine, 'ine. X. Conjunctions. Copulative, wa and; Adversative, pa but, not fctvong; Condi- tional, if, si, wa ; si na wa et te nak vav me te if I should see him I will speak to him, na wa et 'a te if I should not see him ; Illative and Declarative, that, si, wa ; nam tot tua ten wat I thought that it would rain, 7iam vav me te si ten va I told him that he was to go, nam vav si nak va I said that I would go ; Disjunctive, or, si ; ten wat si e'aga ? will it rain or not ? ' Lest' tva, with en, from ; et gov en wa vsxes look out lest it fall. The Noun of companionship is me\, ma'a, Motlav mata ; ina me'eJc ^isik I and my brother, ite ma' an ^esen he and his brother, inik ?ne'em Hsim you and your, irge me'er ^eser they and their, brothers {es and is unaccountably changing). XI. NUMEEALS. 1 . Cardinals. — One vuival, two vum, three vu'wl, four vuve', five 'evelem, six leve'a, seven leveru, eight leve'ol, nine leveve', ten sanwnl ; twenty samoul ru ; thirteen sawioul xval demei vuul ; a hundred meltol, four hundred and nine meltol vagve venegi leveve'; a thousand 'ar. The Verbal Particle vu is prefixed to the true Numerals, as shown in twenty and thirteen. The unit above ten demei, sum above a hundred, venegi, Nouns. The Interrogative, Indefinite, res. 2. Ordinals end in gi ; vaegi first ; keregi, lologi. 3. MultijMcatives formed with va or vag; vawal once, vagru, vagal ; vagves ? how many times 1 vagve four times. 5. Leon and Sasar, Vanua Lava. Leon and Sasar are close together, and about half way between Pak and Vuras ; it is natural therefore that the dialects of the two should be much alike, and that both should show a connection z 1 -^ 38 Melanesian Grammars. with the dialects on either side. The two are here combined ; the language of Sasar being added when it differs from that of Leon. The word Leon means 'on the sand;' one dialect is spoken on the beach and on the point iV^arames. That there is considerable difference between the Vocabulary of Leon and Sasar is shown by the words for ' house,' en and qeqeh, ' fish,' mes and r)ianat. I. Alphabet. L Vowels.— m, long and short, e, e, i, 0, 0, u. There are no Diphthongs. 2. Consonants. — k, g ; t ; p, v, w ; q ; m, m, u, n ; 1, r ; s, h ? at Leon. The dropping of t, so characteristic of Pak, still prevails, but is not so com- plete ; yet less so in Leon, where tit is 'star,' vi at Sasar ; it comes in for n as at Pek. Before r at the beginning of words d is sounded. The sound of q is peculiar. There may be a doubt whether h is not heard at Leon at the end of a word where s would occur ; e. g. in Mota %is, a bow, is at Sasar vus, but at Leon vuli or viC, Sasar ge is Leon geh, Mota gese ; Sasar vi, Leon rih, Mota visa. But the breathing is not so strong as to demand a symbol, and it is better, probably, to write vu\ vi'. II. Articles. 1 . The Demonstrative Article is n ; nen a house, nah a canoe. It is remarkable that this Article never seems to be used except when a Noun begins with a vowel, and not then when t has been cast out ; it cannot be nenge with ^enge a tree : when Nouns also have the Sufiixed Pronoun they have no Article ; ^arpeh my body. 2. The Personal Article is e ; ge a thing, e gene the person. In Leon the feminine particle is used, e ro before a woman's name. III. Nouns. \. Some at least of the Nouns that take the suffixed Pronouns undergo a change when they take them ; ak is a canoe, with the Article naJc, but ' my canoe ' is kaJc, ' thine ' kam, ' his ' kan. No Article is prefixed and ak becomes ka-k. This, however, is not clear. 2. Verbal Substantives have the termination a ; me', Sasar ma, to die, ma'a death. 3. The termiuation of Indej^endent Nouns is gl; jnnigi hand, qu'gi head, 'olgi egg, Leon nagi, Sasar sesegi, name. This termina- tion does not of course appear in Coinposition ; ^«?ii 'amar a man's Leon and Sasar. Pronouns. 339 hand, qii qo pig's head, 'ol to fowl's egg, na amar Leon, sese emar Sasar, man's name. Observe change of a to e in the latter. 4. There is no Plural sign. For houses they would say vol en at Leon, that is every house, or en ge ; at Sasar qeqek ge, taking in all the houses and nothing besides ; en, or qeqek, tol the whole house or all the houses. There is no proper Plural ; but Reduplication gives number ; tirtirigi many legs, Leon ; malmalpegi many feet, Sasar. IV. Pronouns. 1. Personal Pronouns. Leon. Singular, i.no; 2. nik; 3. te. Plural, I. incl. get, excl. kamam; 2. kimi; 3. ter. Dual, I. incl. gotro, excl. kamaro; 2. komro; 3. tor ro. Sasab. Singula/r, i. no; 2. nek; 3. te. Plural, I. incl. gat, excl. kemam ; 2. kimi; 3 tar. Dual. I. incl. gatro, excl. kemaro; 2. komro; 3. ter ro. In both Dialects for the Trial 'ol, three, is added to the Plural. Observations. — i. It does not appear that the Prefix i is used with the Pro- nouns, "which are substantially the same as those of Pak. 2. In the Dual in Leon, and to a less extent in Sasar, the vowels of the Pronouns are modified by the sufiixing of ro ; got, kom, tor, ter, for c/et, Icim, ter, tar. 3. There is no difi'erence between the Pronouns as Subject or Object. 2. Suffixed Pronouns. Singular, i. k; 2.711; 3. n. Plural, I. incl. t, excl. mam; 2. mi; 3. r. These are sufiixed, as in the other languages, to certain Nouns only, as ' my,' 'thy,' &c. ; 'arpek my body, ^arpeva thine, ^arpen his. 3. Demonstrative Pronouns. Leon, ties, ekes, this. Sasak, tegen, genegen, this. tine, ekene, that. tene, genene, that. In Leon the Demonstrative Particles are cs and 7ie this and that ; tl, the stem, is probably the same with the Pronoun te ; eJce represents rather some definite object, ties this, ekes this thing. In Sasar, while ne is the Particle that points to ' that,' gen, very different to es, points to 'this.' In this dialect gene is 'thing' (which in Leon is ge); genegen and genene are therefore this and that thing ; and vol gene is ' these ' or 'those' when definite objects are pointed to, Mota o ike, inc, nan. The Demonstrative which comes from the Vocative ge ! is in Leon eraga, raga, in Sasar rege ; both ' you people ! ' and ' those people/ Vocative and Demonstrative. z 3 340 Melanesian Grammars, 4. Interrogative Pronouns are the same in both ; ene ? who ? nane ? what ? n- being the Article. In Leon ' who ? ' of a w^oman, is roene ? and ' who 1 ' plural, is reene ? the common feminine and plural personal joarticles ro and ra being used. In both another plural, enepa, is found ; jpa, a plural sign in Lakona. These Interrogatives are no doubt also used as Indefinite. 5. In both a Noun ^owogi is used, like tuaniu in Mota, for ' some ;' with lilc also, which cannot be explained. In Leon 'o!«o ane, 'owe ane lik some things; in Sasar '«ma?- ^owoiji lik some men, ''owngi ane lik some things. The Distributive Particle, not a Pronoun, is vol. V. POSSESSIVES, I. General Relation, ro, dro; 2. as belonging to, because pro- ceeding from, mo; 3. of close relation as food, Leon ga, Sasar ^0; 4. of drink, Leon ma, Sasar mo. I. In Leon ro follows the Noun, vono ron his country. In Sasar the Pre- position ma, as in Pak and Motlav, is used with some words as a Possessive ; ■vono me te his country. For pi'operty such as a pig, jmla in Leon, ^;o?o in Sasar ; 2^ulak, 2)oloJc, qo my pig. VI. Adjectives. Though Adjectives are commonly used in Verbal form with ga, they some of them stand as pure Adjectives after the Noun ; 'amar lotvo a big man, new wogrig the small house. Comparison is made by a Preposition 'from;' qo ga loivo ten gosoto (Sasar en) a pig is bigger than a rat, kamam ga mew ten ter we are more than they. The Prefix ma and Termination g characteristic of Adjectives are seen in melegleg black, muJumlum soft, qag white. In Leon ma 'amar is good man ; in both mal 'axaar bad man ; but in Sasar me 'emar is for some reason correct, not 'aiwar. VII. Verbs. 1. Verbal Particles ai'e comparatively little used in either dia- lect, and there is no difference in those that are used. There are two ; ga indefinite, and -m past. 1. jra is used with Adjectives, with words that declare condition ; ga we good, ya se bad, ga mav heavy ; but also with ordinary Verbs ; no ga va7i alow I go to-morrow. 2. -m is used only with and after no, te, and hinn of the Personal Pronouns, those, that is, that end in a vowel ; nom, fern, kiniim ; nam ran I went, tern ta va'a he has already done it, nom van et te ine I went (and) saw him not. Leon and Sasar. Verbs. 341 In Leon this m seems to be a part of the word man, which is used as a mark of past time ; no man fa I did, neA' man van you went, te man van, &c. The First and Third Pronouns Singular take after them, before a Verb, k and n respectively, but without any temporal force : nok ran I shall go, ten van alow he will go to-morrow, noh po" ravrav I am sitting writing (Sasar), ten fo" he is sitting. This k and n might be taken for Suffixed Pronouns, were it not that Ureparapara (compare Motlav) has ka and Mota nl as Verbal Particles. This whole matter is unsatisfactory, from want of examples in which the Subject in the sentence is not a Personal Pronoun. In Sasar there is a Particle of continuance te, ; merag te 'atves lo rar the Malay apple flowers in the winter; in Leon there is none, vegig 'awes lo rar, 2. The Plujyerfect Particle is 'i ; Leox, tem 'tir kel ma belel te gen 'i Man ; Sasar, te 'tcr kel me belel tem gen 'i Ion, he brought back the dish he had eaten in. The same Particle signifies that something remains, as elsewhere ; Leon tiktik va H, Sasae ga ivogrig lav 'i, there is a little left there. 3. Verbs without Particles appear to be the rule, and without any sign of tense ; in Leon, gotro poe siksik we two are sitting and seeking ; nik van, get van, ter van, you, we, they, will go ; vegig 'awes, te gen, as above ; Sasar, nek j^o ravrav you sit writing, get van we went 3 te 'ur as above. All Verbs are thus used. 4. In the Imjjerative a modification of the Pronouns is used in the Plural ; emi van go ye, omro van go ye two, 'ol van go ye three ; in the Singular the Verb is as the Indicative, nek van, ten van, go thou, let him go; or without a Pronoun, van gasem go tell. 5. A Conditional sentence is expressed either, as in Leon, as Indicative, nok eso nok asem me te (should) I see him, I (will) tell to him ; or with a Conjunction, as in Sasar, si no wo et te no mak aram »Jie te if I should see him I will thereupon tell to him. 6. Suffixes. — I. Consonantal transitive Suffixes are n, sonon to put together, n, vepen to seat, v, vaqev to cover over. 2. Syllabic, ve, sirve shear close ; me, ulme annoy ; re, se^;re throw away ; se, Sasar, porse mock ; 'e, Leon, x>ore mock (Mota porosag, and, as it might be, ])orotag') ; ge, vange go with, ulge fall with, ulge fall down with, kalge rak kel get uj) again with. This last is no doubt the Mota gag, if it be right to spell it ge, for it is inseparable from the Verb. But it has been also written o. It is clear that it is not the Motlav ter, for that word, as in Pak, is 'ur. 7. Prefixes. — i. Causative, va, ve; va 'aka to hang up, vaqev to cover over, cause to turn down, vejyen cause to sit, va-pute-rx. 342 Melanesian Grammars. 2, Rec'xinocal, ver ; tor ro ver vu they two beat one another. 3, Of Conditio7i, m- ; meser torn, vioivor come apart, mele broken, with shifting vowel ; 'a ; 'aioag to come open. 4. Sjiontaneity, 'am. ; 'amol come loose of itself as a rope. 8. Reflective Verbs are made with kel back ; te (a me' kel te he killed himself. 9. Negative Verbs. — The Negative Particle is 'e repeated before and after the Verb ; no 'e et 'e I do not see, te 'e van 'e me he has not come here, no 'e van 'e I shall not go. There is no sign of Tense. This Particle is also used with the Dehoi'tative 'og ; 'og 'e 2>el- pel don't steal, 'og 'e ta 'amne don't do so. 10. Bedujdi cation signifying repetition and continuance, j9o' ^;o' go on sitting, luklukun, Leon, vasvasgo, Sasar, count. VIII. Adverbs. In both dialects the Adverbs of direction hither and outwards are me and nar, the latter a strange word. 1. Place. — Leon, kes here, etarti there, ekene there; te jw' ekene he is sitting there ; ava where, nek van ekeva 1 where are you going to 1 ten van va ? where is he going to 1 See Demonstra- tives. Sasar, aken here, etarne there ; ge tagen '0 aken this thing stays here, ge tene '0 etarne that thing stays there ; pan and akne there, ava and lokva where ? ten van lokva ? where is he going ? In these there appear the Prepositions a, e, lo, and the demon- strative ne. 2. Time. — In both, kire now, to-day, alow to-morrow, mana hereafter ; in Leon, nanor yesterday, aris day after to-morrow, naris day before yesterday ; in Sasar, lova'an to-day, of past time, etegine now, lo mana ? at what past time 1 ras the day after to- morrow, lo naras the day before yesterday, lo nonor yesterday. 3. Manner. — In both, 'am as, like ; Leon, 'ames thus, 'amene so, like this, that; 'am va (Mota torn avea) how'? Sasar, 'amgin thus, 'amne so, 'ava ava how ? The Negative in Leon is 'iga (^lota tagai), in Sasar ine. IX. Prepositions. 1. Locative, a at; a Pak at Pak; seen in Adverbs, ava where alxj/w to-morroAV ; lo, le, in, is more commonly used, lo gemel in the club house ; Leon, le en in the house, hon on the sand ; and in the Leon and Sasar. Prepositions. 343 Sasar Adverbs lo naras, lo nonor ; Sasar, le vene in the middle, lo lolo qpqek inside the house. 2. Motion to a person ; Leon, ir ; van ir te go to him ; Sasar van gir in ; both strange if forms of hir, sur. 3. Ifotion from, Leon ten, Sasar en; used at the end of a sentence, Leon, tine vono ron te van me ten, Sasar, tine vono me te te van me en, that is his place he has come from. 4. Motion against, gov ; Leon, et gor look out after, van gor go after; Sasar, ta nar gor qo make a fence against pigs. 5. Dative, me ; Leon, 'or me te, Sasar, 'or me te, give to him. 6. Instrumental, men ; Leon, nam vu te men ker, Sasar, no7n vus te men ker I struck him with a club. But at the end of a sentence me ; Leon, ties ker nom vu te me, this is the club I struck him with, Sasar, tigen ker nom vus te 'i me, I had struck him with. 7. Relation, general, ^^e ; jpe a?ie ? about what 1 why 1 pe ma'aen at the door. 8. Relation to Persons, me ; the same with the Dative and In- strumental ; Leon, me no, me nek, me te, men kamam, with me, thee, him, us, te 'og meno he stays with me ; Sasar, 7no ?io, me nek, me te, me get, men kemam, men kimi, me ter. The ter- mination n in men with kemam (compare Mota), shows the word a Noun. 9. Relation in Place, 'a, belonging to ; 'amar 'a Molav man of Motlav, '«mar 'a ve ? man of what place 1 'amar 'amen kamam man of our place, belonging to us. On these it may be observed that there is no great use of locative Preposi- tions; the Noun stands alone: leon PaJc on the sandy beach at Pak. The forms ir and gir may be explained if both are supposed to be hir (of Motlav and elsewhere) borrowed by people who cannot say h, though it would be natural that they should make it sir, as at Pak. Similarly ten and en are no doubt the same word, the Mota nan, Motlav den, properly represented by ten with the characteristic change of n to t ; but en must be supposed ^en by the loss of t, and this difference between the two dialects suggests that the Pre- position came into them from different quarters. Nouns used as Prepositions ; Leon, lalana en under the house, Sasar, lalane qeqek ; Leon, venigi the middle, Sasar, le vene vono in the middle of the place ; Leon, te po'e melno ev, Sasar, te pd an pensi ev, he sits by the fire. In the last example an shows what is in fact again a Noun used as a Preposition, and with the suffixed Pronoun n. The Mota vawo is loow in Leon, vogo in Sasar ; a Noun not used here as as a Preposition but as the Noun ' top ; ' an wow en on the house, on top of the house ; ve' ul an xuok a stone 344 Melanesian Grammars. fell on top of me, Leon ; iu Sasae, an voyo qeqek on the house, an vogok ou me : see also Alo Teqel. X. Conjunctions. The common Cojiulative is wa, but often omitted ; nom van et ten mav I went (and) saw that he was away. Conditional, si if ; si na 'ar if there be a calm. Disjunctive, si, or ; ga we si 'iga ? is it good or not ? Declarative and Illative, si that ; tem tek si nok van he said that I would go, or was going. * Lest ' is wa, but not used without ten or en the Preposition ' from,' which also is used alone with the sense of ' lest ;' et gor ten nek wa ul look out lest you fall, et gor ten ul look out lest it fall ; et gor n^k en ul, Sasae. The Noun whicli is translated by ' and ' is 'a, in Leon ; ''ak 'isiJc I and my brother; in Sasar, ok isik, 'am 'isim. you and your brother, ^an 'isin he and his. XL NUMEEALS. L Cardinals. — Leon, one vowel, two voru, three ve'ol, four veve', five 'eveliva., six leve'a, seven liviru, eight livi'ol, nine leveva', ten sanwul. Sasae, one voival, two voro, three ve'ol, four veve', five 'eveleia, six leve'a, seven liviro, eight livi'ol, nine loveve', ten sanwul. In both the sum above ten is teme ; thirteen sanwul wal, or ivel, feme ve'ol. Hundred vaeltol. In Leon the sum above a hundred is still teme, in Sasae virnegi ; a hundred and forty, Leon, meltol vawel teme samvul ve , in Sasar, meltol vaioal virnegi sanwul ve. A thousand is in Leon 'er, in Sasae 'ar. The Interrogative and Indefinite iu Leon ve vi, in Sasae vo ve. With the Cardinals is the Verbal Particle v- with shifting vowel ; the true Numeral being shown in sanwul ival, or ve'. After meltol the Leon vatuel, and Sasar vawal, do not show the Verbal Particle va, but the Prefix va = vaga. 2. Ordinals.— ^econA, Leon vorunagi, Sasar voronegi ; third, Leon veo/nagi, Sasae veolnegi ; with Prefix va and shifting Vowel, and nagi = Mota anai. First, in Leon meagi, Sasae 'ow'oiv, Mota towotowo. 3. Multiplicatives with Prefix vag except befoi'e w ; once vaivel, vawal, twice vagru, vagro, three times vag'ol ; how many times ? vagvi ? XII. Exclamations. Vocative, ge f Affirmative, Leon tvugi, Sasae irgi, the Motlav irive. Negative, Leon iga, Sasae ine. Vtcras. AlpJuibet, Articles, Nouns. 345 6. VuRAs, VoNO Lav. The district called by its inhabitants Vuras, at Mota Vureas, or Avreas, lies between Mosina and Leon on the coast. The same dialect is spoken in the bay of Vuras and on the promontory iV^erepot, Mota iVuspaut. It is the part of Vaniia Lava which has most communication with Santa JMaria. I. Alphabet. L Vowels. — a, e, e, i, o, o, u, u. The tendency is to shorten the Vowels, which are sometimes difficult to dis- tinguish : e represents generally the a of Mota, netui — natui. There is a sharp short e heard in mcs a parrot, as distinct from mes a fish. The only Diphthong is ai. 2. Consonants. — k, g ; t, d ; p, v, w ; q ; m, m, n, n; r, 1 ; s. d = nd often represents the Mota n. Though p is rather to be written, there is a tendency to sound mb ; this, however, does not affect q, which is kpw. II. Aeticles. L Demonstrative Articles, 0, na ; the latter used only with Nouns which have a suffixed Pronoun; qotui a head, na qotun his head. 2. The Personal Article i ; feminine iro, plural ira ; vo a, thing, i vo the person. III. Nouns. There are the two classes, as elsewhere, those which take and do not take a suffixed Pronoun ; and again there are those which have a special termination as Nouns. 1. Verbal Substantives. — The terminations are g and e ; vano to go, vanog a going, meat to die, mate death. Verbal Nouns are often reduplicated verbs, tern to love, temtem love ; Mota tape. 2. Independent Substantives have the terminations i and gi, connected with the stem ending in a consonant sometimes by u ; qotui head, turgl body. 3. In Composition with a second Noun, which stands in a genitive relation to it, a Noun has the fonn of the stem to which the Independent termination is affixed ; qotui a head, qot qo a pig's head, doi a leaf, do tenge leaf of a tree, ueregi a snout, Nerepo^ locust's snout. But the final a of a stem becomes e, matai an eye, mate govur a door, mateqersal a road. See Mota. 34^ Melanesian G^^ammars. 4. Plural. — There is no Plural sign but tare, a Noun meaning an assemblage ; tare yovur houses, a group of houses ; clol signifies totality, vono dot the whole land, haniam dol ta Vuras veles we are all belonging to Vuras only. Reduplication gives the notion of plurality and size ; ranranoi many, or large, legs. IV. Pkonouns. 1. Persoiml. Singidar, i. ino, no, ua ; 2. inik, nik; 3. ine, ne, ni. Plural, I. i7icl. inin, nin, excl. ikamam, kamam; 2. ikemi, kemi ; 3. iner, ner. Dual, I . incl. cloro, excl. komorok ; 2. komoron ; 3. roro. Trial, I. incl. nindol, excl. kamaktol ; 2. kemitol; 3. nertol. Ohservations. — In the Singular, na and ni are not used as objects of the Verb. The forms of the Dual are remarkable : doro is a combination of the Numeral ro with the Pronoun which in the Plural is nina, here probably na, of which the vowel has shifted to to match ro, and n has been attracted by r to d. In the terminations of komorok and komoron there are, no doubt, the Suffixed Pronouns k, n, of the First and Second Persons Singular ; the vowels oi kamam and kemi have shifted to match ro, kamam ro, kemi ro, = komoro. It may be supposed, then, that k and n distinguish komoro-k, my komoro, the two of whom I am one, from komoron thy komoro. In roro, r, which is the true element of the Pronoun in ner, has become ro before ro two. The Trial is more simple : nindol has d rather than t because of n before it ; in kamak- tol, k, though not wanted for distinction, must be taken fcr the Suffixed Pronoun. 2. Suffixed Pronouns. Singular, i. k; 2.«; 3. n. Plural, I. excl. mam ; 2. mi ; 3. r. In the Singular Second Person n is a change from m. In the Plural, in the inclusive, the common nin is used. 3. Demonstratives. — Ol:o, iloko this ; le/ce that ; gine, that thing, is the same with Mota gi^ia. The Plural ra makes a Demonstrative; ra ta Mot the Mota people. The Plural Demonstrative which comes from the Vocative is rege. In calling to a person they use vo, a thing, vo standing in place of a name ; but a common vocative Exclamation is i ka ; esi ika, Mota esi gai, I don't know ; i in this being the Personal Article. 4. Interrogatives ; ise, feminine irose, plural irase who 1 so ? what ? i the Personal Article. These are also Indefinite. Vuras. Possessives, Adjectives, Verds. 347 V. POSSESSIVKS. The Possessive Nouns taking suffixed Pronouns are no, mugu, ge, me ; with the same significations as in the neighbouring lan- guages ; nok my, mug^iu thy, of thy doing, gan his for eating, men nin ours for drinking. For a pig, &c., bula a property. VI. Adjectives. These are generally in Verbal form with ga, but there are i'>ure Adjectives ; govur liiivo a big house, govur nctui, or menet, small house. The Comjiarative is made by the Preposition den, from. Adjectival terminations are g and r; wuiouag dusty, lootwotor lumpy. VII. Veebs. 1. Verbal Particles have shifting Vowels as in Motlav, accord- ing to first Vowel of the Verb. The Temporal Particles are g- Indefinite and m- Past ; there is no Future Particle. 1. ff-: the vowel shifts according to that of the Verb, but does not always become identical with it : thus with Adjectives in the Verbal form, cfi tise bad, ga mame red, ge dcderes sweet, go korJcor black ; but gro toe good, go tatun hot. With Verbs, na gu mul ma I come, na gi stag I sit. 2. m-: na ma van me il ne I came (and) saw him, 7ia mi stag I sat, wo mo tog ai I stayed there, ni mo tur he stood. 2. The Particle ti of continuance and consequence ; o^;a^- ti nun, ni qoqet sag the banian sheds its leaves (and) buds again. The same Particle with a Pluperfect sense ; 7ii me le kel me teper me gengen ti lohn he brought back the dish he had eaten in. Signifying remainder ; ge menet ti there is still a little. 3. The Verb is used without a Particle sometimes in the Present, 710 qaqaq I write ; and in the Future, na van I shall go, 7ia gagneg I shall tell. The Verb with a, which must be taken as the Preposition, is often used in the Present and the Future ; oiin a dag we sit, ner a tur a tentew they stand (and) cry, win tar a halo mes if it is calm we shall catch fish. It is used with an optative meaning, as in Mota ; nin a van let us go, komorok a van let us two go, nertol a van let them three go. 4. In the Imperative there is no Verbal Particle, but a modifica- tion of the Pronoun ; nik he van go thou, ni ne van let him go, ru van go ye two, used also in speaking to many. 5. Svffixes. — These, Consonantal and Syllabic, are the same as 34^ Melanesian Grammars, in Mota ; meteg to eye, wonot to confine, hokor to keep, ronteg to hear, tegleg throw away, -porseg to scoflf at, unmeg to annoy. The suffix corresponding to tlie Mota tak is tek, not a suffix of the same kind as the foregoing teg ; vantek go with a thing, mastek fall with, kal raha tek climb up with. 6. Prefixes. — Causative, va, vi ; vateqev to hang up, vies to save, make live, vasgir put on the fire. Conditional, m~ ; melet broken, meser torn, moivor come apart, mamarseg pitiful; ta ; tanenig straight, tawag come open, tatewilwil roll over and over. Spon- taneity, tamo ; tamoul come undone as a rope, tamoras fall of itself. Reciprocity, ver ; vervuvus fight, beat one another. 7. Reflective Verbs, with kel ; ne me lig meat kel ne he strangled himself to death, no mo vus kel no I struck myself. 8. Negative Verbs ; the negative Particle is te, used either with- out a Verbal Particle, no te He o so I don't see anything, or with one, no me te He ai I did not see (him) there. The Dehortative is ni tog as at Motlav, or mmve, the Mota mawia it is enough ; mawe palpal don't steal any more. The word of prohibition is koro ; Jcoro palpal don't steal, Jcoro dada tiniealcgine don't do like that ; to many persons rii koro, literally, don't you two. 9. Auxiliary words as in Mota, ti, so ; tikule turn the back, sonogi set the face against. 10. Reduplication ; dada do often, siagsiag sit as a habit, sisisiag go on sitting. VIII. Adverbs. 1. Place; loko, here, = fo oko, Preposition and Demonstrative, in this, gine, demonstrative, there, ai there, indefinite, ave where. Of direction, me hither, et outwards. 2. Time ; okoi now, garqe now, to-day, whether present or past, tolow to-morrow, ares day after to-morrow, nares day before yesterday, wais hereafter, nanor yesterday. 3. Manner ; tiiweak as, tivneak gine, like that, so ; timeak ave, how, in what manner, literally as a, here 1 gese, gese ai, how 1 by what means 1 4. The Negative is tege, sometimes as a Verb ga tege. IX. Prepositions. I. Locative, a, at; as in ave where? at the place where; lo in. 2. Motion to, sur ; of persons, van sur ne go to him; 3. Motion from, den ivom. 4. Against ^or. 5. Dative, min; la mi}i ne give Vtcras. Conjunctions, Numerals, Exclamations. 349 it to him. 6. Instrumental, min ; min o so ? min o ker with what It with a cluh. 7. Ai at the end of a sentence, the same word with ai there, compare ^lotlav ; iloko ker ne mo vus no te ai here is the club he struck me with. 8. Relation in general, h- with sliifting Vowel ; ho so why 1 about what ? ho qotuna at his head, he mate- govur by the door. 9. Of Persons, me; me no with me, me nik with thee. 10. Oi Places, ta ; ra ta ^lot the Mota people. 11. A Pre- position the same as Ureparapara ri, on ; li toio on the hill. It mekek on me, i. e. on the top of me. Compound Prepositions ; the last examj^le shows meke a Noun ; so alalue yovur under the house, a laJnak under me, at my underside, hersi ev, without a, beside the fire. X. CONJUXCTIONS. Copulative iva ; Adversative pa, nava ; Disjunctive si ; Illative si ; Conditional tvin ; na win il ne na gagneg min ne if I shall see him I will tell him ; probably the Mota wun. The Preposition den is used for ' lest,' il gor ten ni mas look out lest it fall. The Noun ta is used, tak I and-, tan, tan, &c. XI. NUMEKAXS. 1. Cardinals. — One tuwel, two niro, three nitol, four nivat, five tevelem, six levete, seven lovoro, eight lovotol, nine levevat, ten saimd, a hundred raeldol, a thousand tar. The unit above ten deme ; thirteen samul tuivel deme nitol ; the sum above a hundred vivi, a hundred and forty-two melnol vagaro vivi samul vat deme niro. How many, so many, ni vis. The Numerals ro, tol, vat, have the Verbal Prefix ni. The vowels in levete, lovoro, Icrvotol, change with the numeral stem : samul should, probably, be samwul for samvtd. 2. Ordinals, formed by suffixing ne to the true Numeral ; second ro')ie, third tolne, fourth value, tenth samulne ; first is moai. 3. Hhdtiplicatives, with va or vag prefixed ; valewal once, vagoro twice, vagtol, vagvat, vagsamul ; vagvis 1 how often 1 XII. Exclamations. The Affirmative is xi^e ! ; ika ! esi ! have been mentioned. 350 Melanesian Grammars. 7. MosiN, Vanua Lava. Mosin is the part of Vanua Lava nearest to Mota of those districts the dialects of which are given here, lying not far from Qakea where INIota is spoken. It will be seen that the language is more like Mota than the rest. The Mota name is Mosina. I. Alphabet. L Voiods. — a, e, i, o, u. These are all sometimes sharp and short. There are no Diphtliongs ; the Mota Jau, maur, tauive, au, sau, tursao, lai, are in Mosina Id, mur, toiv, a, so, metesd, le. The Mota u, when there is a change, is generally represented by 6. 2. Consonants. — k, g ; t ; p, v, w ; q ; m, m, n, n; r, 1 ; s. II. Articles. 1. The Demonstrative Article is o ; na is remarkably absent. 2. The Personal Article e; e Masre; feminine ero, plural era, feminine eraro. It personifies ; e gale the deceiver, gale to deceive ; va & thing, e va a person. III. Nouns. There is the double division of Nouns into those that do and do not take suffixed Pronouns, and those that have and have not a termination marking them as Nouns. 1. Verbal Substantives have the terminations r, ea, a; toga to abide, togar behaviour", nonom to think, nonomea thought, mat to die, mata death. As in Mota mata takes the Suffixed Pronoun, mutak my death. A re- duplicated Verb makes a Noun ; tap to love, tap tap love. 2. Inde2)endent forms of Nouns end in gi ; ^^^nigi, qatiuji, qetgi, ulsigi. 3. In Coni2)osition the true word takes a Vowel after a final Consonant ; o pent qo a pig's shoulder, o qatu ig a fish's head, o ulsi gd the end of a line, o qetgi motu the stem of a cocoa-nut. The last example requires explanation, for the independent form is qetgi. The Mota qetegiu shows the true word to be qeteg and qetgi to be in fact qeteg-gi. The words metesa a landing place, matesala a road, show e as the termination in composition of Nouns which in Mota end in a. 4. Plural. — The Noun tare means an assemblage; o tare im houses ; there is no other Plural sign, except ra with persons. Mosin. Pronouns, Possessives, Adjectives. 351 Totality is expressed by nol ; (jese, is replaced by vcles only, Mota vires. 5. lieduplication gives notion of number and size ; ronronogi many or large legs. IV. Pkonouns. 1. Personal. Singular, i. enS, no; 2. enik, nik ; 3. eni, ni. Plural, I. incl. enin, nin, excl. kemem ; 2. kemi ; 3. enir, nir. The Dual is the Plural with ro, for ru, suffixed ; except kemuro for Jcemi ro. The Trial is the Plural with tol three ; except also Jcemtol for kemi tol. 2. Siiffixed Pronouns. Singular, i. k; 2. m; 3. n. Plural, I. mem; 2. mi; 3, r. In the inclusive First Person nin is suffixed. With some Nouns there are vowel changes when the Pronoun is suffixed, e. g. i. o 2^itiik my hand, 2. piniva, ^.pencn; Plural, i. incl. penenin, excl. ^n'wmt'm, 2. pinimi, 1. pener ; the in- dependent form being penigi, and the stem no doubt pen. In the Dual pc- nenro, the hand of them two, shows the introduction, as in Mota, of n. 3. Demonstratives ; le this, 710 that ; va le this thing, va no that thing ; He this, ilno that, tare va le, tare va no these, those, things. Another is ti, with the Plural ra, ra ti those persons ; vet row mun ra ti tell those people. Ra ta Mo^ the Mota people. This coiTesponds to Mota ragai, but is not made with the Vocative va ! which means 'thing,' and is used for the person's name. 4. Interrogative ; esei, Plural erasei, erosei, erarosei feminine, who 1 sav ? what 1 all also indefinitive. The Distributive Particle is val ; val sei each one, val nir each of them. V. POSSESSIVES. Nouns with suffixed Pronouns, i. no; nok,nom, nan, &c ; 2. mugu; muguk, rather in the sense of done by me; 3. ga, of food; 4. ma, of drink. With the Article noli, vauguk, o gak, mak, mine, a thing of mine. A piece of property such as a pig is 2>olak. VI. Adjectives. Adjectives generally are used in Verbal form, ga liwo ; but some are used as pure Adjectives, im liwo a large house, im. manle a small house. Com2')arison is made by the Preposition nen from ; qo ga liwo nen gusuw a pig is bigger than a rat. The Adverb anan makes a 352 Melanesian Grammars. Superlative ; o kau ga liwo anan a cow is very large, is the largest. Greater number is shown with salo ; keinem ga marag solo we are more. As in Mota, mansom is fond of money, raercsom possessed of much money ; mal is bad, inal matesala a bad road ; tir matesala, Mota tur, the right sort of road. Adjectival teiToinations are g, r ; silig black, leulenir fluid. VII. Veebs. 1. Verbal Particles. — The Temporal Particles are, i.ga, indefinite and Present, no ga sag, no ga ravrav I sit, write; 2. me, Past, no me sag I sat ; ^. te, Future, taloiv nin te so to-morrow we shall paddle. To make the Past unmistakable the Adverb vctag is added; no me nag vetag I have already done it. 4. ti, of continuance, consequence ; gevig ti taw le rar the gaviga flowers in the winter; used in narrative, ti vet mini speaks to him. 5. The Pluperfect ti ; ni me le hel me pok ni me vosvosog ti lolon he brought back the book he had been reading in. The same Particle signifies remaining ; manle vog ti, there is still a little ; and le me ti is a civil way of asking, just give it here. The Conditional Particle is pe, Mota qe ; na ^:>e il ni should I see him. The Imperative may be the simj)le Verb ; mul, vet row mi ni go tell him ; or modifications of Personal Pronouns are prefixed ; ka van ka vet, go, say, ru mul go ye two, tur mul go ye ; ni mul let him go. 2. Suffixes. — These are much the same as in !Mota ; Consonantal, meteg, h-om. mete eye, sonon, vasager make to sit, horn, sag ; Syllabic, ronte, savre, sarve, matve, Mota ronotag, savrag, sarovag, matevag. The suffix tek, IMota tak, is not of this character, multek go with, mastek fall down with, 7iiul gaplot tek go quickly with. 3. Prefixes. — i. Causative, va ; es to live, vaes save, sag to sit, vasager make to sit. 2. Eeciprocal, tier ; nir ro ga vergat they two are talking one to another. 3. Of Condition, me ; meser torn ; mo, molumlum soft ; ta, reduplicated, tatawilwil roll over and over. 4. Of Spontaneity, tav, and tam. ; tavror to go down of itself, tata- mul, reduplicated, to come undone, ul of itself. 4. Im2)ersonal Verbs ; me viivml no it has tired me, I am tired of it, ga momos no it pains me. Mo sin. Adverbs, Prepositions. 353 5. Tlie little auxiliaries ti and so are used as in Mota ; tihul to turn the back, sonag to set the face, 6. Beflective A'^erbs with kel back ; ni vie nag mamat hel ni he lulled, did to death, himself. 7. Xcgatire Verbs. The negative Particle is te, used without a Verbal Particle; no te nag ves I have not done it at all ; but te is also ete and mote ; no ete ihnan I don't desire, no mote ilman ; there is no distinction of meaning or of tense between te, ete, mote. Dehortative as Motlav, ni tog ; ino ni tog let it not be I. 8. Reduplication ; sagsag sit often, sasasasasag sit on continu- ously, the number of reduplications conveying the notion of the length of continuance. VIII. Adverbs. 1. Place; He here, ^;en there, eve where ; of direction hither me, outwards nat. 2. Of Time ; garqe to-day, whether present or past, lenor yesterday, talow to-morrow, we ris day after to-morrow, noris day before yesterday, ones when, of future time, nanes when, of past time ; vog still, toivo, still, vatag already, qarak thereupon, for the first time, ves at all, the Mota wesu to arrive at completion. IX. Peepositions. Locative, i. a at, and e; ave, eve where : 2. ^e in ; ni ga sag le im. he sits in the house, ni me mul le im he went into the house. 3. Motion to a person, sur; mul sur ni go to him. 4. Motion from, nen; also at the end of a sentence, lele vanan ni me to vatag nen, that is his country that he has already gone aAvay from. 5. Mo- tion against, gor ; ar gor viitus fence in a garden, ar gar qo fence against })igs,sargormun sfo^; dress, clothe over the body, with clothes, il gor ! look out ! look after, van gor go after, fetch. 6. Dative, mi\ le mi ni give to him. 7. Instrumental, mun; different it should be observed from mi ; used also like ]Mota mun ; me le mi no mun polak given to me for my property. 8. Of general Relation, pe; ])e sav ? why 1 The same with the Mota j)e, but never used as a com- pound Preposition. It is shown a Noun by ^>en, the Adverb ' there,' which is used also as an Instrumental Preposition ; le ker no me vus ni fen this is the club I struck him with ; compare Motlav ai. 9. Of Relation to persons, vie ; komoru me tog me no you two have stayed with me. This is shown to be a Noun by its being men before kamem and kimi. 10. Oi Relation to Place, ta; ra ta Mot the Mota people, ta Mosin ; also ti, combining with ^^e A a 354 Melanesian Grammars. and le and making, in fact, comjjound Pi'epositions; ti pen le belong- ing to this, ti le lam belonging to the sea. I\founsave used as Prepositions; vogo tm on the house, the Mota vawo ; lalne im under the house. X. Conjunctions. Copulative, tva. Adversative, pa ; strong, like Mota nava, vakvo. Connective in narration oucg. Disjunctive, si. Conditional, mu; mu tar ga so if it be calm can (paddle) go on a voyage ; mu wen mote mul if it rain cannot go ; na mu te il ni if I should not see him. Illative, Declaratory, ta ; no me vet mi ni ta ni mul I said to him that he should go, told him to go ; no me vet ta no mul I said that I should go. This Declarative ta comes after wa the sign of Quotation ; ni me gat mi no wa ta ni van me he said to me that he was coming. For ' lest ' the Preposition nen, from, is used ; nen ni mas lest it fall: 'until' is gin; gin mate till death: 'as' is taxaa, like. The Noun of accompaniment is ^> of place 'a, to i>er- sons me. Nouns used as Prepositions ; lalne en, or qeqek, under the house, vogo the top, ran vogo, atop, upon, ra a Noun also used as Prepo- sition with suffixed Pronoun n ; ran vogo ve'e on a stone, ran vogok on me, i. e. on my top. See Leon and Ureparapara, an and re. IX. Conjunctions. Copulative, wa. Adversative, ^;a. Disjunctive, si ; ge we si na'ager ? is it good or not ? Conditional, si. Declarative, si ; tern tek si 'emen/e he said that it was so. For 'lest' en, away from, is used ; et gor en mos look after it lest it fall ; but the Mota wa is used as Cautionary ; loa enek out of the way with you, take care of yourself. The Noun is mo^o ; eno mo'ok 'isik I and my brother. X. Numerals. 1. Cardinals. — One vo'owal, two varo, three vo'ol, four ve ve'e, five 'evelem, six livi'e, seven liviro, eight livi'ol, nine livive'e, ten sonwul : thirteen sonwul 'awal temegi vool : a hundred and forty meltol sonwul ve'e ran (upon it): a thousand 'er: how many? veves ? 2. Ordinals. — First 'oui'ow, second vorogi, third vo'olgi. 3. Multiplicatives. — Vagve'e four times, vagves? how often? XL Exclamations. Yes we ; No iine ; na'ager nothing, no, a Noun. Vocative, e ge ! ge a thing, standing for the man's name. 9. J/erlav, Star Island. il/erlav, Star Island, the nearest of the Banks' Islands to the New Hebrides, shows a little approach-in lang-uage to Maewo. 358 Melanesian Grammars. The language of the people of the leeward side appears to the Mota people to be ' thick ; ' those, however, who speak it say that the natives of the windward side speak ' heavy.' There is a way of almost cutting off a final a ; vanov' for vanova. I. Alphabet. 1. Vowels. — a, e, i, 0, u. No true Diphthong. The Mota lau becomes lou, lai becomes lei. 2. Consonants. — k, g ; t, d = nd; b, v, w; q; m, m, n, ?i; r, 1; s. The sound of b is mb, but tends towards mp ; q in consequence is rather kmpw, or kmbw. There is no li. II. Articles. 1. The Demonstrative Article is na. This goes with names of places ; na Vun Lav nu lav dan na Islot Vanua Lava is larger than Mota. 2. The Personal Article ?', feminine iro, plural ira. It personi- fies ; i gale the liar ; na vei a thing, i vei the person, iro vei, ira vei ; see Mota. in. Nouns. The class taking the suffixed Pronoun is of course present ; not, however, always the same words with Mota, e. g. nok vus my bow. 1. Verbal Nouns. — The terminations are va, ne, ia, a; vanova, muleva a going, togne behaviour, matea death, vatgoa teaching, clo- domia thought. 2. The termination of Independent Nouns is gi and i ; and the Vowels of a dissyllabic stem are in some cases modified when the termination is suffixed, appearing in their ti'ue form when a Pro- noun is suffixed. Thus sesei a name, from sasa, nasasak my name; teqei stomach, from taqa, livnei hand, from lima ; qoii knee, from qou; qati head from qatu; while daloi, neck, dalok; misui, lip, nusuk; qotogi, beginning, qotogina; show no change. ' 3. When two Nouns are compounded together in a genitive relation, and the former ends in a, it changes the final a to e, and the foregoing vowel also may be shortened, as in Volow ; na sese tadun a man's name, sasa ; na sinsine aloa ma sin lue le Itirodid, the sunshine shone through at the window. (i) When the genitive of a person is signified, when the second word is a Personal name, or a Pronoun, the Suffixed Pronoun is used, na liman Wenag Wenag's hand ; yet the lighter ending is used, and i is also sometimes inserted, na limei sei whose hand ? In na ak iseil na aJc i Tf'oqas, whose canoe? Woqas' yierlav. Pronouus. 359 canoe, it may rather be thought that / is the J^ersonal Article. The house of the two j'ln i rarua, by the hand of the two ni lime rarua. (2) Words ending in a Consonant, which in the Independent form have the ending t or ffi, are compounded in their true form ; na qat qoe a pig's head, na utol ov a hen's egg, kesin nu sag be hag hei he sits by the water side. 4. Prefixes. — An instrumental Prefix is ya; gahala tongs, hala to take up witli crossed ends of sticks, gabulut glue, bulut to make to stick, gasva, ga sava, how. A roller for a canoe is geilau, in which ga is ajoparently conihined with the Prefix i of the Mota and Fiji. 5. The Plural sign is ges ; na im ges the houses, ges having the same radical signification as the Mota gese. A word meaning a company is also used ; na tore tadun. Dol is ' all,' with the sense of totality ; fagu, an Adverb, signifies comple- tion ; na ?m dol all the house, na mi ges fagu all the houses. A Plural Prefix, with terms of relationship, &c., ra. IV. Pkonouns. 1. Personal Pronouns. Singular, i. ino, no, na, 0; 2. iniko, nik, ko, k; 3. kisin, a. Plural, I. incl. igida, da, excl. ikamam; 2. ikamiu ; 3. ikera, ra. Dual, I. incl. durua, excl. kamarua ; 2. kamrua; 3. rarua. The Trial has no distinct form, the numeral is added to the Plural Pronoun ; igida bultol, we three. Observations. — i. Tlie Prefix ( is used or not with more or less of emphasis. 2. The Third Singular ^isiu is evidently not the original Pronoun, but is the same with the Gaua demonstrative k'osen ; the original was probably ke. 3. In the Dual inclusive da has become du by the influence of n in riia. The longer forms of these Pronouns are regularly used as the subject, but may be the object of a Verb. 2. Personal Pronouns as the object of a A^ei'b, or after Preposi- tions, have a form so far different from that which they have when the subject, that it is desirable to exhibit them separately. Singular, i. o; 2. k; 3. a. Plural, i. incl. da; 3. ra, r. After a Consonant i is introduced before the Pronoun ; vus to strike, vusio strike me ; dan from, danik from tliee. Tliese forms are the true Pronouns without the prefixed parts of the Pro- nouns used as subject in a sentence: a via na tadun nik ma maianial where is the man, you saw him? nik sa surmeleir nes be an ? when will you pay them for the house? kisin ma gon mi he was angry with me. The Pronouns in these forms are generally written in one with the Verbs, or Prepositions which precede them are taken as Suffixed. 3. Pronoiuis suffixed to Nouns. J 60 Melanesian Grammars. Singular, i. k ; 2. n; 3. na. Plural, i. incl. ([&, excl. mam; 2. miu; 3. ra. In the Second Singular n represents the more common m. It is remarkable that in one case, at least, the form of Pronoun used as suflBxed to a Verb occurs where, in other languages, the form suffixed to Nouns is employed. In Mota itiau magesek is I by myself, alone, magesevaa thou alone, and so on, magasei being in fact a Noun ; in Merlav it is ino gaso, iniko gasek, kisin gasea, gasegida, gaskamam, gaskamiu, gasera, as if ^a* were a Preposition. The Dual is not formed simply with the ordinary Dual Piououu after the Noun (with e if the termination be a as given above), except in the 3rd Person : e. g. na livnadrua, inclusive, na lima- marua, exclusive, hands of you two. In the 3rd Person na lime rarua. The hands of us three na limada hultol. 4. Demonstratives. Ke, Jcei, kekei, this, kekei ges these ; ne, nia, that, ges nia those. The Demonstrative formed from the Vocative is vatlumer ; and another rava=:Mota. ragai. A Vocative ' you people ! ' is semiu ; vatlumer and }-ava can be explained by the Gaua exclamation vae ! ( = Mota gai) and me?* ( = Mota vaera), a common word for boy ; vatlumer is then vae ! tolu iwer ! ' three,' being used often in addressing a number of persons. Examples. — Na tankei nu tvi'a ikei be vus the tree is good this for a bow; na ak isei ke ? whose canoe is that ? non isei na tevtev keke ? whose is this knife? na, tadun ne kisin ma bal na mavid that man, he stole the glass (bottle), literally, the obsidian ; na sava le tamber nia ? what is in that dish ? rava ta Qaur ma tor gasva na iuiara ? how do those Bauro people build their houses ? avia nik nu maros ? keke gina, which (where) do you wish ? this, to be sure. 5. Inierrogatives. I sei ? feminine, iro set ? plural, ira sei ? who ? Sava ? what : ' which ' is expressed by ' where.' 6. Indefinitive. ' Sei and sava are indefinite as well as interrogative, some one and some thing. There is also tia any ; taga tia not any, not at all. The Relative in English is represented by a Demonstrative, or has nothing to represent it : le ma na ve na ma toarek open give me the thing I spoke to you about it (observe icare to 'speak to' has the object Pronoun k) ; na taduti ma le mino na gagav ma mat vita the man (who) gave me the garment is already dead ; kisin na tadun nik ma vusia he is the man you struck (him). V. POSSESSIVES. I. no, general relation; 2. mugu, relation of proceeding from the person whose the thing is ; 3. ga, of closer relation, as of food ; 4. ma, of drink. Mer/av. Adjectives, Verbs. 3O1 1. no, with the Article, na nok mine, na non thine, i.e. a thing of nunc, &c. ; le ma, nanol; give it here, (it is) mine. Tlie word is never ano. 2. muffu, or n\u(/ ; 7ia m.u»e be gatogok come here behind me ; Third, ker ge mul let them go. To three persons arti mul bultol, which shows that aru cannot be, as it would seem, the Mota lira, a Dual. 3. There are no doubt other words besides kere, which is not easy to explain. In one sense it gives reason or ground, as in Mota : na ma viisia sur kisin kere da nok bok I struck him just because he damaged my book ; but also, as in Motlav, it has a negative sense ; si Saiv kere gar goroa mar kisin ma mat if Sawa had not swum after him it was as if he (would have) died. 3. Svffixes. The Consonantal Suffixes as tliey ai*e used in Mota, n, r, t, appear in matan, kokor, wonot, but g is absent ; the place of g is taken by a lengthened a ; mata an eye, matd to eye a person, man influence, mand to convey influence. The Syllabic Suffixes are those common in the Banks' Islands but without g ; va, vanva to convey, ra, vilra to distribute, la, sibla to hang up, na, hesna to lean against. The Sufiix van represents the separal)le vag, besides va. Examples. — Barta be gub im. throw it at the back of the house ; ma risra na gatogon make vat his back was broken on a stone ; na wot tankei ma malatvania ti soe suwo the bi-anch of the tree broke with him he falls down ; van gui va na taber go through with a dish, va separable. 4. Prefixes. 1. The Causative is va; vatabu to make holy ; but it is not com- monly used, the Verb da to make, taking its jilace ; na mug isei ma da sar na gagav ? who tore the garment 1 whose doing was it? 2. The Reciprocal is var ; rarva ma varvus rarua, they two beat one another, na ak irua sua vanlin two canoes paddled to meet one another. 3. The Conditioual ma and ta ; loor asunder, mawor broken ; hit to break, malat ; taavav to be unsteady. yierlav. Adverbs. Ci^O 4. The Prefix of S|)ontaneity is tuwa\ na hotn ma tawadan (Mota tavanawa) make qatuna, na matan ma qel a breadfruit came off on his head, his eyes were blinded. 5. Tlie Imjyersonal use of Verbs is shown by examples ; na Jua ma salna he beresin the arrow was put lengthway on the wall ; nu da gasva na revrev how is writing done 1 sa da sava nia 1 sa da na gatgat nia, ti maiv ben, to do (future) what with it 1 to do the sore with it, it heals with it. 6. Negative Verbs are preceded and followed by the particles ti and tia ; na ti maros tia I don't like. In the Present and Past ti alone is used before the Verb : na ti matania tia he tanhei nu ttir goroa I do not see him because a tree stands before him, or I did not see him; na ma valgira iniko amoa nik ti manas tia I knew you at first, before you had spoken, literally, you spoke not ; no Verbal Particle. In the Future hi (see Lepers' Island and Araga) is added before ti : no hi ti maros tia I shall not wish, si kisin hi ti maros tia if he should not be willing. To this fit also is added, *(' na hi ti vit maros tia if I should not like; vit, however, has much the appearance of hi ti. 7. What were called in Mota compound Verbs are thus shown, nik ma din area 1 na ma mid din a le qil, na ni kel me, where did you get to ? I went-to-get to the pool, then came back hither. VIII. Adveebs. These are either simple Adverbs or Nouns with Prepositions ; they can be sho\\ii by Examjjles. 1. Place. — "With demonstrative Particles, kalke here, kalne there, avia kisin? kalke le mi, where is he? here in the house; nik ma matania avia ? ka vano ne, where did you see him 1 there, up that way. Direction hither me, outwards at ; up kalo, sag ; isei nc ma kal sag 1 who is that who has climbed up? down suivo, sur ; na madun sage, na velen suwo, your nose above, your mouth below; rev sur suwo na gahan haul down the sail; out lue ; u-as lue na liwoi pull out the tooth. With the Preposition a and ia, aia there, nik ma mtil aia he sava J Be gagav, what did you go there for? for clothes. The place where, via, with Preposition a at, avia 1 where ? with me hither, whence, kamiu me avia whence are you ? kamiu ma mul me avia ? where have you come from ? with i, to, whither, na metsal kei nu mul i via ? nu mul i Veverau, where does this path go to ? to Veverau ; amoa first, Woqas amoa, Woher be gatogon, Woqas first, Wober behind him ; nik wo mul amoa, ino tagur, go thou before, I after. 2. Time. — Qari