Other formats

    Adobe Portable Document Format file (facsimile images)   TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

The Pamphlet Collection of Sir Robert Stout: Volume 67

Art. XLVIL—The Aryo-Semitic Maori. — [A Reply.]

page 400

Art. XLVIL—The Aryo-Semitic Maori.

[A Reply.]

In the last volume of the "Transactions" (xix.) there appeared a paper under the title of "The Aryo-Semitic Maori," by Mr. A. S. Atkinson, of Nelson, on the subject of the origin of the Maori race, and my writings thereon. . . . I will attempt to answer, in as few sentences as possible, the main objections made by Mr. Atkinson.

First, as to the method, a method nicknamed "The Method of Insight." The writer says, in effect, that I claim that my system is a delightfully easy mode of derivation and interpretation, merely being the comparison of surface resemblances. If I stated that for the reader of my little book the result was easy to understand, it by no means followed that the work necessary to produce such results was easy in its process. It is averred by others, (who are followers of the "high and dry" school of philology, and who seem to think that all human knowledge has been digested and absorbed by themselves,) that any two languages may be compared phonetically and resemblances be found. If any one of these persons will take the 7,000 words in Williams' "New Zealand Dictionary" and compare them with, say Esquimaux, or Mexican, he will have no "delightfully easy" task. Better still, a language like the Tlatskanai (Athabascan), quoted by Canon Farrar, wherein kholsiakatatkhusin = tooth, and kholzotkhltzitzkhltsaha= tongue, when compared with Maori, would be a pursuit of no light character, although, as I said of my work, the result would be easy for the reader to follow.

Can it be proved that the phonetic method of comparison utterly fails? or that it fails at all? The more one learns, the more one reads, there comes one crushing dominating idea, the immense antiquity of the human race on the earth. Professor Sayce, who stands amongst the highest of authorities, says (in the last number of the "Journal of the Anthropological Society") that he once made a calculation as to the time since man had been a speaking animal, and he assigned forty thousand years. I must say I feel sympathy with another writer, who has said that those who are hunting for derivations in written records are but "scratching about on the surface" of human speech: words, as symbols of things, had their birth in ages compared with whose antiquity all books, all rock inscriptions, all alphabets and picture writings are but the work of yesterday. With deep reverence for the learned, devoted students of historical research, we must come page 401 to real philology at last, to the phonetic bases of the linguistic divisions into families—perhaps at last to the bases of linguistic unity. As a general rule, there can be little doubt but that languages which use the same sounds to express the same ideas are near akin, and from the same primal source. This has been the idea—and, in substance, the only idea—which has made philology possible: it was solely the likeness of sound and sense perceived between words of Hindostanee and words of European languages which wrought the discovery of the Aryan unity, although afterwards strengthened by other assistance, such as that of grammatical forms, etc. Further research has made it certain that many of these resemblances were not justifiable in comparison; nay, those persons who love paradox and exaggeration delight in stating that if two words resemble each other in sound and sense it is a proof that they are not connected. This would destroy the connection between the English "brother" and the Sanscrit bhratri, between our "stand" and the Sanscrit sthá—but such assertion really hardly needs denial. However we may track a word historically, we get to a dim twilight at last, in which we see the word being written down by an unknown scribe, in letters whose values differed according to the differing phonetic values assigned to them by this writer: briefly, this early penman, or rock-cutter, was then doing for his particular dialect yesterday what the missionary is doing in Polynesia to-day—i.e., writing by sound: and it cannot be doubted that comparison between words in such a similar stage is very fairly permissible. Mr. Atkinson states that I take Aryan words in any period of growth, and compare these with Maori: I answer that in many cases I do this intentionally; with this much of reason,—that many words have scarcely changed to any extent within the historic period, and it matters little at what stage comparison is made with these. If I do not (or, rather, did not) give the oldest form, it was because I did not wish to cloud the sense of the passage by carrying the reader through strings of derivations, not always clear without long explanation. Had I taken the oldest form of the word obtainable, it would always have been to the advantage of the Aryan-Maori theory. Thus, I compared the Maori hoko, "to barter" (modern, "to buy or sell"), with the English "hawker," one who buys and sells; but the Teutonic words (German, hoken, "to higgle;" Danish, höker, "a huckster"), which have kept the old form better than the English, are also nearer the Maori. So with the English word "hook," which I compared with the Maori hake, "crooked." A Maori is perfectly able to say huka, and does use the word in a different sense, but does not mean a "hook" thereby: his word hake (✓Hak), "bent," (compare ahaaka, "bent like a large hook," Colenso,) is akin to the word whence our modern form is derived, the Anglo-Saxon haaca, page 402 "a hook;" Dutch haak, Swedish hake, German haken. Again, I compared the Maori hau, "to chop," "to hew," with the English "hew;" but the German hauen (hau-en) is nearer to the Maori word. As to the derivations within the German itself the same cousinship appears. The German tau, "a rope," is (as I said) the Maori tau, "a rope;" but in Kluge's "Etymolo gisclies Worterbuch der Deutschen Sprache" the German tau is said to be connected with the English "tow," as a rope-making material; while the English tow, "to drag," has its Maori equivalent in the verb to, "to drag," "to haul," as a canoe, (to-anga, "towing;" toanga-waka, "a place where canoes are hauled up); the English and the Maori to be similarly spelt, if written down for the first time, to-day, by sound. I think that the accusation about "words of any period" being used to suit my convenience in comparison fails—it was to suit my convenience as to brevity that I took the familiar form.

Mr. Atkinson considers that I take too much latitude in regard to the letter sounds, instancing that I bring the Sanscrit ve, "to weave," into comparison with Maori syllables we, whe, and whi. There is great indecision in some of the Maori forms between w and wh. I could quote numberless instances where good Maori scholars (in past days) use waka for whaka (causative). I am often doubtful, in comparing Polynesian words, as to which is the oldest form, h, wh, f, v, w, etc.: in many cases the Maori appears to be wrong. Thus, ahi, "fire," should (by comparison with the Samoan afi) perhaps be written awhi; hoe, "a paddle," (in Samoan foe,) should perhaps be whoe-it being possible to distinguish the true h sounds because rendered by Samoan s: thus Maori hau, "wind" =sau: Maori hoa, "a friend"—soa; this question requires much consideration. That the Maori whenu, "the warp of cloth;" whiri, "to plait," Tongan fifi, "to plait cocoanut leaves;" Tahitian, firi, "to plait;" Hawaiian, hili, "to plait, to twist, to spin," may all be connected with a root vi or ve, "to spin, to weave," is very probable phonetically : the sound here seldom gets so far away from its radical as the European derivatives of the root ve or ri; examples given being wine, withy, osier, uitis, ferrule, willon, etc. (Skeat's "Etymological Dictionary"). As to the assertion that I represent Sanscrit d, dh, l, and r by the Maori r, that is also a fact: my plea being that the English sounds are thus represented in Maori. We translate the D of "David" by R of Rawiri; the l of "linen" by r of rinena : and both l and r in the word "glory" as kororia. What interpreters do in translating Aryan English into Maori letters is the only guide I have in thus comparing the Aryan Sanscrit with Maori sounds written in these letters.

The most amusing part of this objection appears when we consider its bearing on the Malay. Because a few words in page 403 Malay resemble Maori in sound and sense, therefore, say the "high and dry" people, this is good enough to prove connection between these peoples. There is no proof that Malay dua, "two," is the Maori rua, "two," except sound-likeness; but if the student, conceding relationship, goes further, and expresses the opinion that the Aryan dua, "two," is also a near connection there is a shriek of horror from the classical linguists. The grammarian cries "Impossible!"—quite ignoring the fact that rules of grammar mark mere stages in the progress of a language, and that modern English grammar is as tar away from Sanscrit or old Latin as the Malay grammar is from the Maori, and that is a "far cry." For every Malay word traceable in the Maori vocabulary, the student can find twenty Polynesian words in the German Dictionary.

The objection that I split up a Native word as I choose is one that I defend, if in doing so I can prove that the probable radix appears more distinctly in that form. A greater or less complexity may be allowed to roots; but the more simple the form we use, the more we surrender the meanings possible to be expressed. I do not consider that the division according to the Native mode of so dividing a word would be the best way in which to arrive at the primal significance. Thus, if we take the word patu, "to strike," I consider that it may be treated under three distinct radical forms: that is, either as the third root Pat, "to strike;" or, as the second, the more simple Pa "to touch; or perhaps its most primitive form, ✓A, "to urge," "to drive." Whatever may be the vowel of direction or modification used to close the syllable in Polynesian fashion the sound of Pat or Pak (the true dialectical interchange) carries the sense of "striking," "knocking, "pattering," "patting," "breaking," etc., in patu, pata, patoto, patiti, patōtō, pakakū, pakanga, pakaru, paketu, paki, pakini, pakunu, and pakuru The sister-words in the Islands are multitudinous, but need not be quoted. I think this argument shows that, in spite of the natural divisions of a Polynesian word into distinct syllables, each ending with a vowel, a radical sense may extend itself over words like pata, "to drip" like water; ripi, "to cut;' mano, "the heart" (mental, not physical), manawa, "the heart;" mana, "influence, authority;" maru, "bruised;" kite, "to see, perceive" which would allow them to be referred to similar roots to those given (Appendix, Skeat's "Etymological Dictionary") Aryan roots: such as Pat, "to fall," Rup, "to break Mar, "to grind," Man, "to think," Kit, "to perceive,"

Mr Atkinson scores one point against me fairly enough : that in which he shows that my comparison of kiri, "the skin," with our curry, "to dress hides," is wrong. He does this on the authority of Professor Skeat, who states that curry comes page 404 from courroyer, the old French being con-royer; thus showing [unclear: it] to be a compound word conveying the idea of "to set in [unclear: order.] I yield the point, pleading that, at the time I wrote the paper containing this comparison, I had no copy of Skeat's [unclear: "Etymo]. logical English Dictionary," and that I was relying upon [unclear: tw] other Etymological English Dictionaries as references. In [unclear: the] first, Richardson's "New Dictionary of the English [unclear: Language,] 1855, the derivation of curry is given as through French corroyer from the Latin corium, "a hide," with the example : i corium equi et dorsum fricare strigilli, "to rub the hide of a horse with a currycomb." The other authority was Ward and [unclear: Lock] "Standard Etymological Dictionary," 1880, in which cornoyer is given as a derivative of corium. There can be little [unclear: dou] that the French cuir, "leather," is connected with Latin corium "a hide," Lithuanian skura, "a hide," Sanscrit karma, "hide," Irish creat, "a hide:" all pointing to a common [unclear: ro] Kar or Kir (KR) as their source, and thus being akin the Maori kiri, "a hide." If we also consider the word curée, given by Brachet* as a hunting term for "pieces of skin, etc., [unclear: throw] to the hounds," I think it possible that in the vulgar or provincial idioms unknown to literary men, a word "curry," meaning something to do with hides, or the skin, did exist in English, even by the side of such base compounds as con-royer This may yet be found by English scholars to be the case, and the etymology reconsidered. From my correction by [unclear: M] Atkinson a very useful lesson may be learnt, that is [unclear: as] considering any Etymological Dictionary as a thing "made [unclear: la] by the Medes and Persians." If we compare our derivation to-day with those given only thirty years ago, note the discrepancies, and then picture our present works in the light of century hence, the notion that we know all about everything—even about the impossibility of the Aryan-Maori theory—may be shaken.

Apparently the most serious argument used by Mr. Atkinson is that wherein he urges the Semitic side of the question, and uses my method in discussing it. Perhaps I am very dull and dead to fine raillery in not supposing it to be all pure wit, [unclear: but] will take it for granted that Mr. Atkinson has not been mocking the Society too utterly by sending in a paper for publication without any seriousness in it at all. He first discusses the word "Maori," which he says is the same as "Mauri" [unclear: (] and asserts that the meaning is probably "living, not [unclear: dead.] He then quotes from Codrington's "Melanesian [unclear: Languages'] in support of this. I think that most Maori student will agree with me in declining to consider that Mauri and Maori are one word, or that the meaning of Maori [unclear: is]

* "French Ety. Diet.," 1878.

page 405 "living, not dead," even if backed up by many Melanesians.* Williams' "New Zealand Dictionary" says that Maori means "native;" the "Tahitian Dictionary" gives maori, "indigenous, not foreign;" the "Hawaiian Dictionary" also gives maoli, "indigenous, native;" there is not the slightest reason for supposing the word refers to the Mauri or Moors of ancient history. Mr. Atkinson seems to have taken up (without acknowledgment) the late Rev. Mr. Taylor's idea as to the Moors and the Mauri, although Mr. Taylor's derivation of mauri (i.e., ma + uri, as being the "black blood of the heart") would scarcely coincide with the sense of "indigenous, native," and would be rejected with scorn by the Maoris, if "Moor" and "blackfellow" be offered as the explanation of the national name of all the fair Polynesians. There is, however, a race of people whose ancient language, called Sanscrit, does contain the word, or one very like it. In Monier Williams' "Sanscrit Dictionary" we find maulika, "original," "radical;" maulya, "being at the root;" maula, "living from olden times or for generations in any country," "indigenous." This word thus agrees in sense, and nearly in sound, with the "maol" of the Polynesian maoli; "Moor" does not so agree, either in sound or meaning.
Mr. Taylor is a very curious guide for Mr. Atkinson to follow. I do not by this expression mean to decry Mr. Taylor's ability—it would be well if there were a few more men in New Zealand actuated by a spirit of inquiry and love of knowledge akin to his. But, as Mr. Atkinson says he has been a student of the Maori language for a long time, he must know that Mr. Taylor is a very doubtful authority. Mr. Atkinson says that one proof of the Maoris having known the kava root, as chewed for intoxicant, is that Mr. Taylor so considers, on the strength of a place named "Kawaranga!" This must be jest: the name of the place is correctly spelt "Kauwaeranga." In Mr. Codrington's book there is much that is of very great value, but it is not all of equal value. As a grammar, etc., of the Melanesian languages it is quite unique, and has won so much of praise for its author that it can afford a few points of dissent being raised without suffering. But when the grammar, etc., has been considered, and the author then theorizes outside the technical part, we are at liberty to totally dissent from that theory and the conclusions. The theory is that the whole of the Oceanic languages are (in base) one : Malay Archipelago, Madagascar, Papua, Polynesia, Melanesia, Micronesia—all the

* Mr. Atkinson's example of mauri meaning "living" is the "God save you" expression, used when one sneezes : Tihe, mauri ora! ("Sneeze, living soul!") Only, unfortunately for Mr. Atkinson, the word for "living" in this sentence is ora, so that as an example of mauri meaning "living" it is very weak.

page 406 natives of these places speak one tongue. There are too many points of argument in the question to meet here, but I doubt [unclear: i] Polynesian scholars will accept any such theory. There [unclear: i] great persistence and no great difference in most vital Polynesian words : rakau, "a tree," for instance, is lakau, [unclear: ra] laau in almost all the islands. But if we are to study Polynesian through Melanesian; if we are to find the word [unclear: rak] (honestly) in hayu, ai, ei, kayu, diwal, pasil, ie, etc., as Mr. Codrington says, we shall require "more light" than a single book can afford us. There can be little reason for studying; Maori through such corrupt and degraded channels as the Melanesian speech; it would be about as reasonable as [unclear: t] study the English language through the "slave-blobber" o the American Negro. It seems to me that Mr. Codrington's efforts are used to make those among whom he labours be considered equal to any of the other islanders. He says (p. 12): "The Melanesian people have the misfortune to be black, to be much darker, at least, than either Malays or Polynesians;" (at p. 13) "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, [unclear: o] their more civilized and stronger neighbours;" (at p. 35) [unclear: "t] the Polynesian, who is shocked at being claimed as a relation by a much blacker man than himself, it is answered that be speaks a language very like the Melanesian, but not so complete and full." The gist of these remarks seems to be, "My black-fellow is as good as (if not better than) your brown-[unclear: fellow.] I can only say that there is no moro reluctance among the Polynesians to acknowledge kinship with the Melanesians [unclear: tha] there is among Europeans to acknowledge kinship with the light races of the South Seas. By the accounts of the early explorers, they again and again mistook Polynesians [unclear: f] Europeans. Ethnologically, I should think that the distance between the straight-haired, light-brown, Polynesian and the blue-black, woolly-haired Melanesian was very great, [unclear: i] type, although there may be many intermediate links in the islands, made by persistent "crossings" of the [unclear: strains] blood.

The Malagasy speech-family is a very difficult subject [unclear: t] treat of: as to many words being kindred (if the sound and [unclear: so] resemblance is acknowledged as proof of unity) with the Malay, the fact is indisputable, although I believe many of the [unclear: coincidenes] are fallacious. But the Polynesian words to be found are [unclear: fe] indeed : the words used in Madagascar have been but too [unclear: ofter] compared with the corrupt and abraded forms of Eastern Polynesia, wherein, by the dropping of important letters, all [unclear: origins] form of the word has been lost, and become worthless for purposes of comparison. And the grammar! He must [unclear: be] page 407 "grammarian with a theory," indeed, who finds Malagasy, Papuan, Malay, and Maori grammars identical. The Malagasy is so full of words adopted from English, French, Portuguese, Arabs, and the neighbouring African tribes (as well as Malays), that comparison is infinitely dangerous and difficult.

Mr. Atkinson made a very long quotation from Professor Whitney's "Life and Growth of Language," with the object of overwhelming me by an authoritative statement as to the only manner in which etymologies of differing languages may be compared. The main point in the quotation is, that "whereas a close verbal resemblance between the nearly-related tongues has the balance of probabilities in its favour, one between only distantly-related tongues, or those regarded as unrelated, has the probabilities against it." Who, then, are to be the judges as to the languages to be considered related or unrelated, before they have been compared? For centuries the classical languages of Europe and the classical language of Hindustan were supposed to be unrelated, and it is only in our own generation that the claim, fiercely contested, of the Celtic-speaking peoples to be admitted into the Aryan family was acknowledged. Plenty of sarcasm and ridicule (now forgotten), plenty of loud, frothy denial was poured upon the advocates of the Aryo-Celtic theory. To use the name of Professor Whitney is to charm with the wand of one whose name is respected by every educated Englishman, and a two-fold measure of this respect is due from those who are students of language; but can anyone believe that Professor Whitney advocates the method followed in European linguistics being applied to the study of Polynesian? That is to say: that the literary and historic method should be applied to the study of races having no literature and no historical records? An idea so brilliant cannot be Professor Whitney's—it was reserved for an Antipodean writer to evolve this spasm of genius. Must we be contented never to compare the Polynesian language with any other until we obtain their literary records? I decline to do this, and I will quote the words of one greater than Professor Whitney concerning this question. Professor Max Müller, in his "Introduction to the Science of Religion," p. 97, says:—

"My chief object in publishing, more than twenty years ago, my letter to Bunsen 'On the Turanian Languages,' in which these views were first put forward, was to counteract the dangerous dogmatic scepticism which at that time threatened to stop all freedom of research, and all progress in the Science of Language. No method was then considered legitimate for a comparative analysis of languages except that which was, no doubt, the only legitimate method in treating, for instance, the Romance languages, but was not, therefore, the only possible method for a scientific treatment of all other languages. No page 408 proofs of relationship were then admitted even for languages outside the pale of the Aryan and Semitic families, except those which had been found applicable for establishing the relationship between the various members of these two great families of speech. My object was to show that, during an earlier phase in the development of language, no such proofs ought ever to be demanded, because, from the nature of the case, they could not exist, while yet their absence would in no way justify us in denying the possibility of a more distant relationship."

It is precisely this point I wish to establish : my contention being that Polynesian is "Aryan in the agglutinative stage"—the more valuable because in the agglutinative stage language is comparatively "transparent," and therefore etymologies are knowable; while in the later inflected stage (when the inflections consist of dead and forgotten agglutinations, as in case-endings) the primitive sense can only be guessed at and quarrelled over. From European philologists we hear what they infer the pre-inflectional Aryan language must have been, when "the flections had not yet been evolved, and when the relations of grammar were expressed by the close amalgamation of flection-less stems in a single sentence-word;" when "there was as yet no distinction between noun and verb," and "the accusative and genitive relations of after-days did not exist;" "when as yet an Aryan verb did not exist, when, in fact, the primitive Aryan conception, of the sentence was much the same as that of the modern Dyak;" when, "apart from the imperative, the verb of the undivided Aryan community possessed no other tenses and moods;" when "the parent-Aryan was once itself without assigns of gender." It is to the Aryan tongue in its crude but more vital childhood that I wish to compare the Maori language.* If there was the slightest historical probability that the Maoris had received the words which compare in sound and sense with European and Asiatic ones, either by conquest, by commercial intercourse, or by religious teaching, then those words would be easily separable : as easily as the later Sanscrit can be detected in those words of Malay introduced by Brahmin and Buddhist priests from India. It is not possible even to pretend the probability of such late intercourse between Asia and the Maori: the Maori tongue gives evidence of being a primitive form of speech, not a decayed dialect of a nobler language. It is only through negligence, or want of acquaintance with the Polynesian tongues, that the most remarkable of these "coincidences" have not been investigated before. Such men as Bopp and Humboldt (both of whom saw the sure affinity) failed from sheer want of material. To rely upon such information as the Tongan of Mariner, and the Maori of Lee and

* See Fornander, "Polynesian Race," vol. iii. p. 12.

page 409 Nicholas, was to possess a most unfortunate basis whereon to erect any kind of superstructure. As an example of later incapacity in regard to Polynesian, I may quote from Maxwell's valuable "Manual of the Malay Language" (p. 10): "A long list might be made of common words not included in any of the following groups, which are almost pure Sanscrit, such as bawa, 'to bring,'" etc. Now, here we have a Malay word, bawa, which, if allied to the Sancrit root vah, "to carry," is certainly not so near as the Polynesian vaha, "to carry." Yet in the same work (p. 3) is given a quotation from Crawford's "Malay Gram-mar," in which the writer says: "An approximation to the proportions of Sanscrit existing in some of the principal languages will show that the amount constantly diminishes as we recede from Java and Sumatra, until all vestiges of it disappear in the dialects of Polynesia." The writer must have known more about Malayan than Polynesian.

Concerning those Semitic languages which Mr. Atkinson brings forward as evidence how my method can be applied to them, he at once frankly confesses that his knowledge of Arabic is of such a quality that he has not even taken the trouble to acquire the ability to read the written character. I have a greater charge to bring against him even than indolence: he does not seem to have understood the A, B, C of modern philology, which separates the root-formation of the Aryan and Semitic branches of language into two distinct systems : "The root of Aryan verbs is all but invariably monosyllabic, consisting of a consonant followed by a vowel, as in da, 'give,' or sta, 'stand:' but the root of the Semitic verb is always trilateral, or rather triconsonantic, and therefore necessarily dissyllabic—i.e., instead of being, as in Aryan, an open syllable, it is always close (as in qtl, 'to kill;' dbr, 'to speak;' ktb, 'to write.') . . . Thus, in Greek, γράμμα is 'a writing,' γραφϵúς 'a writer,' and ϵγραψϵ, 'he wrote;' whereas in Hebrew SeePheR is 'a book,' SoPHeeR is 'a writer,' and SâPHaR, 'he wrote.' Again, in Greek, βασιλϵρς is 'a king,' and greek text 'he reigned;' but in Hebrew MeLeK is 'a king,' and the same word with other vowels, MâLaK, 'he reigned.' Thus it is as if in Hebrew the trilateral consonants—which were the only things which appeared in writing at all, the vowels being left absolutely unrepresented—were things too sacred to touch."*

My contention is that the Maori language is founded on the Aryan, and not on the Semitic root-system of trilateral consonants; that Maori is formed, fundamentally, on open syllables of a consonant followed by a vowel, as Farrar and Max Müller both state that the Aryan languages are formed. If Klaproth's theory of a primitive universal language can

* Canon Farrar, "Language and Languages," pp. 354-56.

page 412 Mr. Atkinson makes this assertion, (satirically, of course,) not I, because I know, as everybody does, that the modern Maori meaning of po is "night," and "Hades." But that a bird should be named after a bull is not without precedent, the English bird, the bittern (Botaurus stellaris), being named doubly after the bull—the derivation of bittern being supposed to be from bos and taurus—perhaps from its booming, bellowing cry; but it seems almost as ridiculous, at first, to associate together the idea of a bull and a wading-bird as a bull and a parrot. In this paper I will not enter into the question of the Maori having been acquainted with cattle, because I hope next session to deal with the subject exhaustively, on account of information gathered from all parts of Polynesia and from ancient writings; but I will in a few words put the argument into such a shape as will deprive it of any idea of improbability. If the Maoris migrated to Polynesia, they probably came from one of the great continents. If Mr. Atkinson and his party reject the evidence of tradition, and its universal consensus in the South Seas as to the migration hitherwards, I have no more to say. But, if they came from some other place, where was that place in which they had no knowledge of cattle? In Europe, the haunt of the aurochs and urus? In Asia, the home of the pastoral tribes and cattle-tending nomads? In Africa, the dwelling of the buffalo? or in America, the land of the bison? With the single exception of the island-continent of Australia, (a place there is no tittle of evidence to show that they visited,) wherever the Maoris came from, they must have known cattle. From the most eastern extremity of Asia, where the Chinese call the cow ngau, (k = ng, kau—ngau; and I compared the English gnaw with the Maori ngau, "to bite," the idea being "ruminant," "cud-chewing,") right through all changes acknowledged by philologists—kau, coo, go, gau, bos, βovc, etc.,—to the extreme west, in Ireland, where the word is bo, the name stretches right across Europe and Asia. If the Maoris came from either of these two great continents, and knew cattle at all, it was probably by one of these two forms of the word "bo" (Tongan bo, Maori po) or "kau." that the animals were called. Thus I, who believe that the Maori race had its cradle on the lofty plateaus of Central Asia, cannot see anything ludicrous in the idea that their language retains some trace of words which they must have used, if they used any.

The "nay a" theory I shall not defend at any length, but I decline to accept Mr. Atkinson's play upon nga as conclusive in any way. The Maori, who knows not the snake in New Zealand, uses for "snail, slug," etc., the word ngata, which in Polynesia means the "snake," and is in Hawaiian naka. Whether the Hawaiian form is nearest of kindred to the naga serpent of Sanscrit, the naga, Malay, "a dragon," snaca, Anglo- page 413 Saxon, and the Icelandic snakr, I do not know; but Hawaiian forms are nearer to European in many words than Maori forms are—perhaps because, as I said in "The Aryan Maori," the Hawaiian may be a later migation. It is curious, too, to notice that "snail" (Maori ngata,) is (snægl) from the same root as "snake," whilst "adder" (properly a "nadder") is the German natter. I had spoken in "The Aryan Maori" of the "footed serpent," the lizard, being regarded with awe; it is certain that nya in Polynesia was used for the "lizard." The Marquesans call the large house-lizard nga-nga, a curious word if it is merely the duplicated article. Constantly, in Marquesan, ng changes with k, (as it does in Maori—and in Latin,)* the representative of this word in New Zealand being kakariki, the green lizard, for nganga, ngaka, or ngata-riki. Thus it seems highly probable that the important part of the word ngata, "a snake," is nga. If it was really the case that the Maoris, like other Polynesians, knew the snake as ngata, nata, or naka, then the missionaries in giving them the word naka for "snake" were unintentionally (even pathetically) giving them back their own word lost for centuries, as I feel certain they unsuspectingly did in a hundred other cases, where words supposed to be pakeha-maori, or corrupted English, may be found in songs and incantations ages old. Be that as it may, anyone who considers that where nga is used in composition as a prefix it is but the plural article "the," can scarcely have examined the subject at all. Setting aside the direct words nga, "to breathe," nganga, "a stone," etc., there are many words prefixed with nga, in which nga evidently has some direct bearing on the sense of the word that no conception of it as merely a prefixed article agglutinated will explain. Whether nga has ever meant naga or not, it seems possible, from the genius of the Polynesian language making a vowel follow a simple consonant, that this double consonant sound ng may once have had a vowel between n and g = na-ga.

* Marquesan, ikoa, for ingoa; hoki, for hongi, etc. Latin, pingo and pictum, tango and tactum.

See "Aryan Maori," p. 26.