* tf IMAGE EVALUATION TEST TARGET (MT-3) 1.0 I.I [f ii^ IIM •^ 1^ 1112.2 1^ IIIIIM 1.8 1.25 1.4 1.6 .4 6" — ► m \^' Photographic Sciences Corporation 23 WEST MAIN STREET WEBSTER, N.Y. 14580 (716) 872-4503 L(P CIHM/ICMH Microfiche Series. CIHM/ICMH Collection de microfiches. Canadian Institute for Historical Microreproductions / Institut Canadian de microreproductions historiques Techrical and Bibliographic Notes/Notes techniques et bibliographiques The institute has attempted to obtein the best original copy available for filming. Features of this copy which may be bibliographically unique, which may alter any of the images in the reproduction, or which may significantly change the usual method of filming, are checked below. L'Institut a microfilm* ie meilieur exempiaire qu'il iui a 6tA possible de se procurer. Les details de cet exempiaire qui sont peut-Atre uniques du point de vue bibiiographique, qui peuvent modifier une image reproduite, ou qui peuvent exiger une modification dans la methods normale de filmage sont indiquAs ci-dessous. D Coloured covers/ Couverture de couleur □ Covers damaged/ Couverture endommagie r~| Coloured pages/ Pages de couleur Pages damaged/ Pages endommagdes □ I'.o its restored and/or laminated/ f iverture restaur^e et/ou peiiiculAe Cover title missing/ Le titre de couverture manque □ Pages restored and/or laminated/ Pages restaur^es et/ou pelliculdes □ Pages discoloured, stained or foxed/ Pages dicoiordes, tachetdes ou piqu^es □ Coloured maps/ Cartes giographiques en couleur □ Coloured init (i.e. other than biue or black)/ Encre de couleur (i.e. autre que bleue oi* noire) D Coloured plates and/or illustrations/ Planches et/ou illustrations en couleur □ Pages detached/ Pages ddtachdes 0Showthrough/ Transparence □ Quality of print varies/ Qualit^ in^gaie de I'impression Bound with other material/ Reii* avec d'autres documents B Tight binding may cause shadows or distortion along interior margin/ La rellure serr^e peut causer de I'ombre ou de la distortion le long de la marge intdrieure D Blank leaves added during restoration may appear within the t3xt. Whenever possible, these have been omitted from filming/ II se peut que certaines pages blanches ajouties lors d'une restauration apparaissent dans le texte, mais, lorsque cela itait possible, ces pages n'ont pas 6ti (meaning "CON- TINUED"), or the symbol V (meaning "END"), whichever applies. Maps, plates, charts, etc., may be filmed at different reduction rat'os. Those too large to be entirely included In oi « exposure are filmed beginning in the upper left hand corner, left to right and top to bottom, as many frames as required. The following diagrams illustrate the method: L'exemplaire filmA fut reprodult grice A la gtnArositA da: Metropolitan Toronto Library Literature Department Las images suivantas ont *t« reprodultes avac la plus grand soln, compta tenu de la condition at da la nettet* de l'exemplaire film*, et en conformity avac las conditions du contrat de filmage. Les exemplalres originaux dont la couvarture en papier est imprlmte sont filmAs en commen9ant par la premier plat at en termlnant soit par la darnlAre page qui comporte une empreinte d'impreasion ou d'illustratlon, soit par la second plat, salon la cas. Tous las autras exemplalres originaux sont flimte an commengant par la premlAre page qui comporte une empreinte d'impreasion ou d'illustration et en termlnant par la darnlAre page qui comporte une telle empreinte. Un des symboles suivants apparaltra sur la darnlAre image de cheque micro Jche. selon le cas: le symbole -^ signifie "A SUIVRE ", le symbole V signifie "FIN". Les cartes, planches, tableaux, etc., peuvent Atre fllmAs A des taux de reduction diff Arents. Lorsque le document est trop grand pour Atre reprodult en un seul cllchA, 11 est fllmA A partir de Tangle supArleur geuche, de gauche A droite, et de haut an bas, an prenant la nombre d'images nAcessaire. Les diagrammas suivants illustrant la mAthode. rata lelure, I A H 32X 1 1 2 3 1 2 3 4 5 6 .r;:^'^^i^ I w^fi- ■ '';■■'*«: '' ii^:7^- '■V-. "5 '^^-,.-- b-^ SOME LAWS OF PHONETIC CHANGE IN THE KHITAN LANGUA(iES. IIY .U)HN CAMIMJKU., M.A. I'riifr.-ixDr ill Hir I'cmhiilcrittti Collfi- , Mniitiv/'L m \' i M SOME LAWS OF PHONETIC CHANGE IN THE KHITAN LANGUAGES. BY JOHN CAMPBELL, MA. Professor tn the Preshyterian College, Montreal, In several published articles, some of which were read before the Caradian Institute, I have givju comparative vocabularies illus- trating the connection of the American languages with those of the Oltl World. Among ethnologists there is a strong prejudice against this mode of procedure, a prejudice arising partly from an unwilling- ness to undertake the labour necessary for an appreciation of the results obtained ; partly, it may be, from a suspicion that the vocabulist has been too anxious to prove his point to be scrupulous about the means ; and, in particular, from the possibility or prob- ability that the resemblances exhibited are nothing more than such chance coincidences as will appear more or less in comparing any two languages in the world. A similar prejudice nught have opposed, and in many minds probably did for a time oppose, the reception of the Indo-European family of languages, for the resem- blances presented in their vocabularies as compared among them- selves are not a whit more striking than those which characterize a comparison of the languages of north-eastern Asia with those of the principal native races of North and South America. This, howexer, distinguishes the two linguistic fields ; the Indo-European is infinitely better known. Now, speaking of that field. Professor Max Miiller tells us that, as far as etymological science is concerned, identity or similarity of sound or meaning is of no importance whatever, This, of course, is true when we ai*e dealing with individual words, but to apply such a rule in the case of a general comparison of vocabularies would be to remove the foundation on which the classification of languages has been laid and from which comparative etymology has sprung. As well go to the extreme at once, and, with Schleicher, 2 in SOME LAWS OF PHONETIC CHANGE assert that gramiiiiitical oonHtruction is the only test of linguistic affinity, as if no great changes liad taken phico in such construction, soul of language though it be, even within the period of modern history. Putting aside such extreme views, or i)('rhai)s, as it would be more just to term them, extrinne statements, and asking the philologist to suggest some valid criterion of i-elationship among languages which we deem to be connected and whose grannnatical systfinis are, to say the least, not discordant, he will probably invite us to discover among them such a process of phonetic change as has been illustrated in the case of the Indo-European languages by the N.'eli-known Grimm's law. Now it is precisely such a law, or a portion of such a law, that I profes., to have found, after a somewhat laborious and careful examination of those New and Old World languages which may constitute provisionally the Khitan family. The name requires explanation. About the middle of the tenth century, a foreign horde, whom the Chinese annals know as the Khitan, descending from the north, took i)OSsession of Mantchuria, and extended their sway over the whole of Northern China. For two centuries they maintained themselves as the rulers ot that empire, being recognized in Chinese history as the Liao Dynasty, and were then expelled to the north-east by the Nyuche, a supposed Mautchu tribe, wiio ruled in their place as the Dynasty of Kin. It was these Khitas or Khitan, for the final n is the Khita mark of the plural, who gave to the Celestial Empire its mediajval name Cathay. Some of the Chinese historians derive the Khitan from the desert of Kobi, b\it, farther to the north about the sources of the Yenisei, and throughout Southern Siberia according to Tartar tradition, their I'cmains are found. These are tumuli, similar to the mounds of this continent, containing arms and ornaments, and sculptured inscrip- tions upon adjoining rocks in an unknown hieroglyphic character. The Tartars call the tnmuli Li Katei, or the tombs of the Cathayans. Tumuli of the same character as those of Siberia, accompanied in many cases by cup shaped and other rude sculptures agreeing in out- line with those found in many parts of this continent, appear in India, where they are regarded as the work of a Turanian people, the Indo-Scyths of history. These raust have been none other than the Kathaei of Arrian and Sti-abo, whom Alexander the Great encountered at Sangala in the Punjatib. The very name Sangala is Khitan, for from the Songari River the Khitan are said to have IN THE KHITAN LANGUAGES. 5 (lescendod upon China ; to the country of Sii^^halien thoy retired ; and their presence fai-ther oast in Japan is marked by the straits of Sangar. Sangura again or Sagura was the name of a river in the country of the Kliita or Hittites, according to the Assyrian inscrip- tions, and its ethnical character is apparent in its use as tlie proper name of one of the greatest Ilittite raonarchs, Sangara of Carchemish. Several native references to the Indian Sangala, as well as that of Isidorus Characenus, make it plain that its population was not Aryan, but Turanian or Indo-Soythic. In the third century, A.D., these Indo-Scyths were expelled or subdued, and at that j)oint the migration northwards through Tartary to Southern Siberia mus;t have commenced. It is natural to suppose, in the want of definite information, that the Kathaei or Khitan reached the Punjaub from the west V>y skirting the northern boundary of the Persian empire, arriving in their Indian home at or before the fourih centuiy, B C, when Alexander found them there. The Persian chronicles class among the northern peoples of Touran the Khatai, and link them with Shankul, Prince of Hindustan, another Sagala or Sangala. The original cause of tlieir movement eastward was the capture of the Hittite capital Carchemish on the Eui)hrates by Sargon, King of Assyria, in 717 B.C., and the consequent dispersion of a brave and restless people unwilling to live under a foreign yoke. Many ti-ibes, as has been shown by Professor Sayce, Dr. Hyde Clarke, and othei-s, found their way into Asia Minor, where Hittite dynasties reigned down into the days of Rome's supremacy. Otheis, long ages before, when the Kheti invaded the land of the ancient Pharaohs, leaving their Syrian domain, planted colonies in northern Africa, and even penetrated into Europe. But the great bulk of the Hittite population took refuge in the Caucsisus, and from thence by dint of j)ressHre, internal and external, forced its eastward way along the route that has been traced in i-etrograde order, from the Caucasus to the Punjaub, from the Punjaub to the Yenisei, from the Yenisei to the Songari, and thence to Corea, Japan, the Kurile Islands, Kamtchatka, and, finally, as far as the Old World is concerned, to the Aleutian chain. They carried with them their practice of mound building, their peculiar hieroglyphic character, and their own geographical and tribal nomenclature. The mounds begin with the Tells of Syria, ate followed on the west by the Lydian and other similar tombs of Asiii Minor, on the east by the tumuli of the Caucasus, SOME LAWS OF PHONETIC CIIANOE TiiiHa, Tartary, Siberia and Japan, and on tliis continent give name to tiu'ir otluTwiso unknown architects, tlio Mound Builders. At Caivlicniish and Ifanuitli, in Plirygia and Lydia, tlie llittite liicro- glypliics stmngo and distinctive nMuain as niotiuinents of Kliitan empire and journeyings. The (Jy[)iioto sylhibic notation has bor- rowed largely from them; the Libyan and Kelt-Iberian alpliabets are their descendants. Some of the more characteristic symbols aj>pear on rudely sculpturelication of simi- lar laws to forms of sj)eech widely se|>anited geographically. Instead of setting foilli in this paper the whole of my com|Mvm- tive vocabulary of over 150 words in the various languages and dialects of the Khitan family, which would be more likely to con- fuse than to convince, I prefer for the present to restrict mys(df to an exhibition of some of the relations of one such language to its connecteil forms of speech. The language selected is the Hui'on- Iroquois in its various dialects, the Huron, Tuscarora, Nottoway, Mohawk, Oneida, Onondaga, Cayuga, Seneca, &c. This is one of the most i)ectdiar and difficult members of the family, differing from all the others known to me in this particular, that no one of its dialects possesses the labials b, p, v, f, or the liquid and labial w. The nearest appioach they can make to a labial sound is to, and where m cannot be similarly represented it must be replaced by another liquid, n. With the Huron-Iroquois language I compare first of all that m: mber of the family which, following the line of Khitan migration backwards, is the most remote from it, namely the Basque of northern Spain and south-western France. Grammatically the two languages agree, for it has been rightly said that the Basque is the most American of the Old World tongues known to philology. A better acquaintance than is at present possessed of the languages of north- eastern Asia woidd doubtless modify such a statement. Still it is well to be on a right footing with the grammarians, although one of them, M. Vinson, a distinguished Basque scholar, who, some time ago, pub- IN THE KHITAN LANGUAOKS. V liHliotl an article comparinj^ tlio n.iH(|iio with tlio Iroquois, failntl to fiiul the gminiiiiitical acconhinre of tlicj hiiigiiagcH borne out I a- th«' lexicon. This, however, arose frouj the fact tlint M. Vinnoii hail not nnule a 8|tecial study of the Inxiuois, and that he hail ucgh-cted the geographically intermediate languages which, in sonje reH|iect8, furnish the key to the common origin of the Iroquois and the Basque. I.— In a large number of isstancrs, althouuh thkre ark manv EXCEPTIONS, TUB IrOQUOIS KEI'LACES TUB BASQUE UyflDS / AND /' BY ANOTUER LiyUIli, H. Take, for example, tin; Iroquois word for tooth, honozsia, nnutchla. It is easy to perceive the relationship hetwecir these forms and the innotay, noti, of the Choctaw, the ente of the Natciirz, tlu^ twto of the Shoshonese, and even the neaa, nar/ha, of the I^enca. But where, it may be asked, is the similarity between these names for tooth and that of the Yunia, which is aredoche 1 The Basque dis- plays the relation. Its word for tooth is Ao?<^, ortz, or, in the plural, hortzdc, ortzac. The unaspirated ortn, somewhat drawn out as is generally the case in the pronunciation of iniciviiized man who haH ivbundance of time for his conversation, becomes, without any conso- nantal change worth noting, the Yuma aredodie. If, however, we apply the nde which transforms the Basque r into the Iroijuois n^ then ortz becomes ontz, and hortz, the aspii-ated Liivbourdin and Baa- Navarrais form of the word, hontz, thus furnishing us with abbrevi- ated but distinctly recogniaible equivalents of the Ii-oquois onotchia and honozzia. In the Kasi Kumuk dialect of the Lesghiau the Basque aspirate is strengthened into k, kertshl being its rendering of hortz. Indeed it may almost be said to be a rule that the Basque aspinite, as an initial letter at least, becomes the liCsghiau g^ttui'al. The Quichua of Peru follows the same rule, ivnd aurpas^ses the Lesghian in its attenuation of the vowel, by ohs^nging kertski to kii'V, Thus the two forms onotchia and kirn, which ai)peav to pveaont no feature in common, are found to have the same origin, A similar instance is that of the Iroquois kelan^uaw, which de- notes the moon, but also the sun. The Pueblo word for sun ia hoolenwah, with which the Yukahiri name for the same orb, yelonsJ/LOi, invites comparison. But in the Basque the equivalent for kdan- quaw, the moon, is hilargia; and, just as the Yunaa (tred^oche cor; 10 SOME LAWS OF PHONKTIC CHANGE resiionJed with tlie Basque ortz, so does the Yuma hullyar almost perfectly repi'0(hice the Basque hilargia. Let the Iroquois n be- come r, and kelarqiiaw is tlie Basque hilargia and the Ynma huUyar. The Quichua, still retaining its original guttural, changes hilargia and hullyar to coyllor, but employs the word to designate not the moon but a star. It is worthy of note that the Yukahiri of Siberia, ■which renders the sun as yelonsha, calls the moon kininsha, thus replacing the I as well as the r of hilargia by n, and preparing the way for the Aino kunezu and another Iroquois form, kdnaughquato. An Iroquois word for an axe or hatchet is ahdokmih, and this is the Koriak adaganu. Turning once more to the Yuma, the pheno- menon presented in aredoche and hullyar is repeated, for the Yuma word for an axe is atacarte. Here again we meet with the Basque, f ir atacarte is to aizkora as aredoche is to ortz. In Aino and Ja. panese the Bas(iue woi'd takes a prefix ni, and aizkora becomes masakari. The Yuma gives us kooruk for the adjective old, and the Iroquois, akaioji ; here also the Yuma and the Basque agree, for in the latter language old is agurea. But in the Lesghian both forms appear, for, while the Avar and three other dialects accord with the Basque and Yuma in herau, two, the Akush and Kubetsh, are in harmony with the Iroquois, ukna and okna being their respective renderings. In North America the Dacotah also gives two forms, that of the Sioux or Dacotah proper being kon, and that of the Upsarokas or Crows, karrahairea. The double form karrahaireu is itself far from singular. The Lesghian tribe of the Avars, besides herau, uses mirvara, which becomes noorkoor in Coi-ean, poimgur in Aino, and furuberu in Japanese. A remarkable word for egg is the Basque arraulizia. The appli- cation of the rule to r and I reduces arraultzia to annauntzia, which is almost the sound of the Iroquois word ovhonchia. The Quichua agrees with the Iroquois in changing the I to n, but retains the r, and removes the initial vowel ; thus arraultzia' becomes runto. A similar elision of the initial vowel takes place in Kamtschatdale, which furnishes the two forms — lilchatsh cori-esponding with the Basque, and nyhatch according with the Iroquois. In all the Khitan languages there is no radical distinction of ad- jective and verb. Indeed almost any word may become a verb. Taking the word dead, therefore, we find it represented by the IN THF KHITAN LANGUAGES, 11 Basque substantive bo called, erio, heriotce, aud the Iroquois adjec- tive kenha. But kenha is the same word as heriotce, for, while the "Lesghian tribes, Tohar and Kabutsh, render it by chana like the Iroquois, ih-i other Lesghian tribes, Dido and Unso, agree with the Basques in calling it hura'z. The Dacotah sides with the Basque in karrtisha, and the Peruvian Ayinara with the Iroquois in hinata. A road or street in Basque is kharrika, but in Iroquois chanhei/ens. The Dacotah, which the late Lewis Morgan proved to be of the same stock as the Irocjuois, furnishes the more appropriate form kanga, while the Lesghian reconciles the Basque and it by its duplicate renderings clmhlu and chuni. The Corean rejects the termination which appears in kharrika and chuldu and calls a road kir. The Koriak ennen, innaen, a fish is the Basque arran, arrain, and the same with the prefix of a guttural is the Iroquois kunjoon. So the Iroquois enia a finger is the Basque erhia, and the Basque oscofa, the bark of a tree, is the Iroquois askoonta. Again, the Qnichua rejects the initial vowel and calls bark kara. The t of askoonta which is not found in oscola is probably a euphonic addition merely, since it frequently appears, as in ourata, a leaf, the Basque orri, in ashiichtit, a hand, the Basque escua, and Dacotah sake, and in kihade, a river, the Kamtchatdale kiha. II. — The Iroquois replaces the Basque m by an, en, on; and the Basque b follows Tun same rule as m when it is the equivalent OF that letter in the Caucasian lasouages. One of the best known Iroquois words is onontes, a mountain, figui-atively employed to denote a governor or great personage, as onontio the beautiful mountain. Tbifi (ovm onontio probably explains the Hittite word mati in the Hamath Miscriptions, which I have translated " king." However, the Iroquois onontes is the Basque mewHa. In South America the Basque form is almost given back in the Araucanian mahnida, but the Cayubabas of north-eastern Bolivia, a people allied to the Qiiichuas, are Vasconibus Vasconiores and turn the Iroquois onontes iuxj iruretui. The word tongue in Basque is mia, mihia, the Lesghian mitz and mas. The application of the rule transforms vias to ennas, which is just ennasa, the Iroquois tongue. The Georgian form is ena. The Caucasian m is frequently represented in Basque by b. Thus the Lesghian mussur, muzul, the beard, is the Basque bizarra. 12 SOME LAWS OF PHONETIC CHANGE There is little doubt t!?'it the Les•». In the second we meet wiih the Oiicassian abra. The final ac of ileac and hiloac is the Basque mark of the j)lur{d, and is the same in origin and in function as the Iroquois ke. Although not entitled to rank as a law of phonetic change, it is worthy of note, as tending to simijlify the exhibition of the common origin of Basque and Iroquois, that the Iroquois frequently differs from tlie Basque by inserting a dental between the letters a and r,for purposes of euphony. Thus the Iroquois kanadra, bread, is the Basque jnnhari, janari , food ; for the initial j, as we learn from M. Lecluse, though pro- nounced as in French in the canton of Soule, and as in German in that of Labourt, assumes the power of the Spanish letter in Gui- puzcoa, and may be represented by kh. The verb to love in Basque is onerechi, oniritzi, in which it is easy to recognize the Yukahiri anoorak, and the Japanese noroke. In Yuma the word is awvonoorch. Three Iroquois forms are eniloorooh- quah, aindoorookwa and enorongwa. A lai'ge number of words in Basque and in Iroquois coir 'ide in sound and in signification, and for such coincidence I have so far been able to discover no law. Among these may be mentioned the Iroquois garioha, bird, which is the Basque choria, the Lesghian zur, the Aymara chiroti. The final ti of the Aymara has also appeared in hinata, dead, as compared with the Iroquois kenha and the Lesghian chana. The Iroquois white, which English missionaries write kearagea and the French kenraken, is the Basque churia, the Japanese kiroi, the Loo Choo shirusa, the Lesghian tchnlasa, and the Quichua yurac. An Iroquois word for dog is tschicrha, tlie Sho- shonese schari, the Mizjeji (Caucasian) tkari, the Georgian djogori, and the Basque zacurra. So the Basque hezurra, bone, is tlie Iroquois ohskereh, and the Cherokee ookolah ; and the Basque aztala, leg, is the Iroquois okotara, and the Lesghian uttur. While geree, an Iroquois word for ti*ee, agrees with the Basque chara and the Quichua kullu, meaning wood, another Iroquois form, kaeet, is the 1 9sghian hueta, guet, the Basque zuaitz, zuhaitz, and the Aztec i u SOME LAWS OF PHONETIC CHANGE qu'ihuit, quauitl. The Kliitan terms for thunder are like the Semitic gUyad. The Lesghian seems to furnish the type in tjurgur, which is ai)proached by tlie Basque curciria, ihurzuria, and aggravated in the Koriak urgirgerkla. Tlie Georgian ujodiHes the harsh sound by dropping one of the ?•'«, as in yurgin and kuchili, the latter of which corresponds with otlier Koriak foi-ms, kyhal, ki/igala, and with the Kamtchatdale Kf/chichlizen. The Choctaw has the two forms jyrajaa nnd hiluha ; t' Yuma stops short at aA;er; but the Iroquois furnishes a word kaivseras that agrees more perfectly with the Old World forms. I have already referred to the Yuma dialects (the Yuma or Cuchan, Maricopa, Mojeve, Dieguno), as valuable members of the Khitan family for comparative purposes. Two Yuma woi'ds for cold are xetchicr and hutaeelo. The former accords with the Dacotah hootsheere and the Iroquois otsorai, which the Basque changes to otubero, while in hutseelo we find the Lesghian chuatzala. The con- nection of the Iroquois oni/are, neck, with the Basque cinzurra might seem doubtful, as the Basque sibilant and guttural prefixes are generally more cons{)icuous by their absence than by their pre- sence. But the Yuma form heiinceil shows that it is the Iroquois which errs by default in this respect. The YvLk&\\\v\ jomuel restores that original form which would naturally have been looked for in the Basque, and leads the way to the Lenca ampshala. The Lesghian word gabar, which could naver be evolved out of cinzurra, naturally rises out ol'jomiiel. The Iroquois onuste, maize, and the Basque m'to, artho, have little in common. The Yuma tarrichte, however, dropping the initial t and applying the first rule as if it were a Basque word, becomes annichfe. Another Yuma form is terditch, with which may be com- pared the Lesghivn zorolo and the Circassian nartuch, and with these the relation of the Basque arto, artlio is easily perceived. Still another Yuma word meaning to speak is atchuhquerck. This is undeniably the Aino itakguru. But another Aino dialect gives idakawa, and this prepares us for the Iroquois XUakia and the Basque itzegln. The nearest word to the Yuma haweel, meaning a river, is the Aymara hahuiri or hawiri, and this is plainly the Lesghian uor, chgare, and the Basque uharre, ukarka. In chyare, by the appli- cation of the first rule, we detect the Iroquois kahionha. IN THE KHITAN LANGUAGES. 15 1- Iti some cases the Basque word, while agreeing with the Iroquois, differs from the Lesghian, so that both Irocjuois and Basqiie nnist be bi-o»iglit under the first rule, in which Lesghian must take tlie place of Basque. Thus the word for name is in Iroquois chinna and in Basfjue icena, while the Lesghian form is zound words of the Iroquois tongue, but in Koriak, in Kaintchatdale, and in Japanese, we discover, not indeed the precise wortls, for a few centuries may suffice to alter these, but some of the elements of which they are composed. Take, for instance, the Iroq>iois word for silver. It is hwichtan-oron. I am not sufficiently verseil in ancient Iroquois to know the meaning of its component parts, but there can be no doubt that the first of these, hwichtan, is the same as wifchiin in the Koriak word elnipel-wychtin, denoting the same metal. An Iroquois word for the colour yellow is cheenu-yuarle, and uarle is ai)parently the same word as karalh in the Kanit(;hat- dale dnchl-karallo, which means not yellow indeed but green, colours not always distinguishable by savages, for the Koriak uses the same term, ui/il-tshachain, for both. Another Iroquois word for yellow is hotfi'ikkwa-rognn, of which the latter member, rojon, corresponds with tjrachen in the distinctive Koriak term for yellow, nuutel- (jrucheM. We are on a surer foundation in regard to the Iroquois words for red, two of which are otquccJi tarokxi and ory gains no support from phih)logy. If in this paper I have not exliibited tlie rehition of the Iroquois dialects to those of all the divisions of the Khitan family, it is not from lack of material or in order to avoid any difficulty. I have pui-posely chosen for comparison languages the most remote in place and in time of separation from the original tongue, languages of peoples most unlike in present feature and character, whose sole con- necting link has been supposed to be the common possession of a complicatfid grammatical s stem marked by polysynthesis. That I have succeeded in showing the relation of these languages to one another and at least to some of the intermediate members of the Khitan family, will be granted, I doubt not, by all true philologists who do not shut their eyes at antecedent improbability. RULE I. The Iroquois replaces the Basque I and r bv n. Basquk. Rur.K Applied. Ikoquois. 1. oitz, liiirtz ontz, hontz 2. Iiilarjiia hilatigia :>. iiizkiira aizkona 4. a^iirea agunea 5. anaiiltzia aniiiiuntzia 0. lit'i'iii henio 7. klianika klianiiika a. 9. 10. nrniiu eihia oscula annain enliia oscona ODdtchia, lionozzia ki'laiiquaw ahdokenh Rkaion (iiihiinchia keiiha cliaiilieyens (Oai-otah, canga) kurijiioii eiiia askoonta Snoi.ish. tooth 111(1011 axe old <'KS dead road fl.sh finger bark Note. — Illustrations of the Basque forms. 1. ortz rytti, A'oriV/fc; arcdorlie, Yuma hortz kiixlsUi, I I'mjIiiilH ; k\ru, Quichua 2. hilargia=luillyar, I'«mn ; voyXlor, Quichua 3. aizkdia iiiasakari, Japiinese, Mno; ata- carte, Yumn 4. agurea = Iierau, Lesg/tian ; kooruk, Yuma; karraliairea, Daeotih 5. arraultzia HlchaLtsh,Kaiiilchatdale •,mnto, Quithita 0. herio, lieiiutce, = haratz, Lesghian ; carrasha, Dacotah 7. kharrika = shara, Georgian ; chuldu, Lesghian ; kir, Coreui, 8. arrain atlaii, Aztec !). erhia - kilish, Lesghian ; gelyhat, Koriak ; hal, Yetmei 10. oscola keiki, Georgian ; IclialjO'n, Koriak ; kara, Quiohua Rule in other Languages. Iroquois korms. unotchia ^-^ iniiotay, noti, Choctaw; cnte, iiit, Notches:: iioto, Shonhouese ; ncas, iiigii, iiaglia, Lenca kelanqiiaw —^ gailgen, Koriak ; yelunsha, Yukahiri ; hoolenwah, I'liMos. ahdokenh = adaganu, Koriak ; tlatecoui, Aztec. akaion = ukna, okna, Lesghian ; kon, Dacotah onhonchia, ^ iiyh&Uth, Kamtchatdale ; nanki, Shosho?iese kenha = ohana, Lesghian; hinata, Aymara chanheycns ^ o.hiini, hutii, Lesghian ; canga, Dacotah; hinah, Choctaw kunjoon -= ennen, innaeii, Koriak ; henn, Natchez; "kwnn, Ajimara euia '^ onkaiiah, Dacotah askoonta kani, Georgian (skin); cangha, chanha, Dacotah IN THE KHITAN LANGUAGES. 19 RULE II. The Iroquois rkplaces the Basque vi by an, en, oh, and the Basque I> BY the same when b IS THE EQUIVALENT OF THE CAUCASIAN m. Basque. 1. raendi 2. niia, mihiu :<. bizarra 4. burua 5. biar, bihar, bigar i>. bizkhar Caucasian. Ikoquoib. uita, (leorgian onontes suiitii, Leaghian initz, iiiaM, Lestjhmn onnasii ena, Georgian inuB8nr, inuziil, I.esghian onwskera inier, maar, J.enghian nnuwara mlfliar, Oenrgian enior-heno macho), Lenghian olinaken Enulish. mountain tonguo beard liead to-morrow back Note. — Illustrations of j Basqub and Caucasian forms. 1. mendia = nialmida, Araucanian ; pinujidtsh, Kamtchatilate •2. mia, mas = mutt, motto, Mitjeji ;i. bizarra, mussur = musur, Atacnmtno; muzul, (hair) 4. burua, mier, maar = mari, Cortan ; marshaa, Dacotah ; moola, Scmora ; abara-cama, Cayubaba 5. biar, michar = myongir, Corean ; mayyokal, Yuma; miecar, Yuma (morning) ; emukulas, Kamtcliatdak (morning) tJ. bizkhar. machol = ushiro, Japanese HE Rule in other lanciuaoks. iROliUOI.S FORMS, (inontcs = suntn, Lesghian; nelt, Korialr nenicliaha, Choetaw cnnasa ^^^ ena, Georgian; oiinor, Yukahiri ; neeiglijee, Dacotah ; yalinoligali, Cherokee ; lioninee, Pueblos ; anongiu, Shoshonese ; nenetl, Axtec ; ine, Cayubaba onwskera = hannockquell, Hhoshonese (I'liin) ; Iiiintur, A tacameno anuwnra = nahuar-acama, Cayubaba enier-hene == unhaiel, Yukahiri (morning) ; onnihile, Choctaw (morning); sunahfa, Cherokee (to-morrow and morning) ; NliinnakHlinre, Dacotah (to-morrow and morning) ; yannie, Shoshonese (morning) ohnnken = senaka, Japanese RULE III. The Iroquois replaces the Basque 6 by m> when 6 is the equivalent OF A labial in the CAUCASIAN LANOUAORS. Basque. Caucasian. Iroquois. Ekolish. I. zabala chvallal, chvallase, Lesghian kowanea great balachi, Georgian wennokera grass 2. belharra 3. biloac abra, Circassian wennokera ahwerochia hair RULE IV. The Iroquois inserts ▲ dental between the Basque » and r. Basque. Iroquois. Other Lanouaqes. English. 1. Janari.janhari kanadra kendo wan, Koriafc bread, food 2. onerechi, oniritzi endooroohquah anurak, Yukahiri to love aiudoorookwa noroke, Japanese awvonooreh, Yuma Basque. 1. choria 2. chunu 3. zacurra 4. hezurra 5. aztala 6. chara zuaitz, zuhaitz 7. curciria Roots coinciding in Basque and Iroquois. Iroquois. Other Lanouaoes. garioha zur, Lesghian ; chiroti, Aymara kearagea kiroi, Japanese ; shirusa, Loo Choo tchalasa, Lesghian ; yurac, Quiehua tschierha djogori, Georgian ; tkari, Mii^eji schari, Shoshonese ohskereh ookolah, Cherokee okotara uttur, Lesghian geree kuUu, QuxchiM kaeet hueta, guet, Lesghian qnahuit, quauitl, AzUe kawseras gurgur, Lesghian urgirgerkin, kyhal, kyigale, Koriak kyohichlizen, Kamtctiatdale guTg[in, kuchili, Georgian .iyrajaa, hiloha, Choctaw ; aker, Yuma English. bird wnite dog bone leg wood, tree thunder 20 HOMK LAWS OK PHONETIC CHANGE, ETC. I. ill TuK Yuma dialects as aids to l'omparison. IllOltltdlH. Yuma. UA.114UK. Otiikr Lanouaobo. Enoi.irh. 1, otsoriii xutuhur liutseulii (itHberu (^liuatzalu, Usghlau <:ulii rnE KEL.VTION OF COMPOUND WORDS IS luoyuols. Japank-sk-Kuriak. eirtinej-wyclitin, Koriak (^uuAt-karullo, Kamlchiitdnlc )iuu(e2-gracheii, Koriak nitshel-nxcUen, Koriak. cAi-durako, c/u'-darakerio, Japunesc Knulish. silver yellow yellow ri'(l I I'i *1 ,i (•