(1859).

P. 252.—"Is not this Mozino's?"—No. For a further notice see p. 388.

P. 258.—"Kawichen and Tlaoquatch."—The Kawichen is nearer to the Nusdalum, Squallyamish, and Cathlascou than it is to the Tlaoquatch. This may be seen in Buschmann p. 649. At the same time it is more Tlaoquatch than Buschmann makes it.

P. 259.—"The Athabascan languages are undoubtedly Eskimo."—Between the notice contained in p. 299 and the paper which precedes it there is an interval of no less than five years. There is also one of three years between it and the paper which follows.

Now up to 1850 I gave the term Eskimo a power which I afterwards found reason to abandon. I gave it the power of a generic name for a class containing not only the Eskimo Proper, but the Athabascan, and the Kolooch. The genus, though in a modified form, I still believe to exist; I have ceased, however, to think that Eskimo is the best name for it. Hence, expressions like "the Athabascan languages are, undoubtedly, Eskimo—and the Kolooch languages are equally Eskimo with the Athabascan" must be read in the sense of the author as expressed in p. 265—"that the line of demarcation between the Eskimo and the Indian races of America was far too broad and trenchant."

Whether certain forms of speech were not connected with the Eskimo Proper—the Eskimo in the limited and specific meaning of the term—is another question. The Ugalents was so treated. The Kenay—until the publication of Sir T. Richardson's Loucheux specimens—was made both too Eskimo and too Kolooch. On the other hand, however, both the Eskimo and the Koluch were divisions of the same order. The actual value of the term Kolooch is even now uncertain.

P. 276.—"The Ahnenin etc."—A reference to the word Arrapahoes in Ludwig's Bibliotheca Glottica (both in the body of the work and the Addenda) suggests a doubt as to the accuracy of the form Ahnenin. Should it not be Atsina?

Turner remarks that "there is no evidence that Dr. Latham collated" Mackenzie's vocabulary—which, as far as the text of Ludwig goes, is true enough. I had, however, viv voce, informed Ludwig's Editor that I had done so. As Turner knew nothing of this his remark was a proper one. The main question, however, touches the form of the word. Is Ahnenin or Atsina right? I can not make out the later history of the MS. In my own part, I copied, collated, and returned it; and I imagine that it still be amongst either Prichard's or Gallatin's papers. I have the transcript before me at this moment; which runs thus. "The vocabularies of the Blackfeet, of the Crows or Upsarokas, and of the Grosventre, Rapid, or Fall Indians who call themselves Ahnenin; by D. M. M'Kenzie of the St Louis American Furr Comp. They appear to belong to three distinct families. But the Crows speak a dialect clearly belonging to the same language as that of the sedentary Minitares and Mandans, which is Sioux."

As the MS. was written with unusual clearness and distinctness I have no doubt as to Ahnenin having been the word. That Prichard read it so is evident; for the foregoing explanation has made it clear that he and I are independent witnesses. If error, then, exists it is in the MS.

The Blackfoot and Crow (which having also transcribed, I have by me) are as follows:—

The Italics are the present author's. They draw attention to either a coincidence between the two languages, or the compound character of the word.

II.—The Sioux group.—For a remark on the affinities between the Pawni and Caddo, see p. 400.

The following coincidences are the result of a very limited collation.

(1).
Cherokee and Caddo.

(2).
Cherokee and Iroquois.

IV. The Athabaskan group.—I find that the affinity between the Loucheux and the Kenay languages is given by Prichard, who, at the same time, separates both from the Athabaskan. "Mr. Gallatin says that the similarity of languages amongst all these" (i. e. the Athabaskan) "tribes is well-established. The Loucheux are excepted. This language does not appear to have any distinctly marked affinities except with that of the Kenay."—Vol. V. p. 377.

I believe that Dr. Prichard's informant on this point was the same as my own i. e. Mr. Isbister.

Scouler also suggests the same relationship.

That Buschmann has arrived at the results of his Athabaskische Sprachstamm through a series of independent researches I readily believe. Whether, after taking so little trouble to know what had been done by his predecessors, he is right is saying so much about his discoveries is another question.

That the Pinaleno is in the same category with the Navaho is shewn by Turner, who gives a vocabulary of the dialect.

V. The Kitunaha language.—The Kitunaha, KÚtani, or Cootanie vocabulary of Mr. Hall was obtained from a Cree Indian, and is not to be depended on. This being the case it is fortunate that it is not the only specimen of the language. There is an earlier one of Mr. Howse's, published in the Transactions of the Philological Society. It is as follows.

VI. The Atna group.—The numerous vocabularies that represent the dialects and sub-dialects of this large class are the following—Atna Proper or Shushwap, Kullelspelm (Pend d'oreilles), Spokan, Kettlefall dialects of the Selish; Okanagan; Skitsuish (Coeur d'alÈne); Piskwaus; Nusdalum; Squallyamish; Kawichen; Cathlascou; Cheeheeli; Tsihaili; Kwaintl; Kwenaiwitl; Kowelitz; Nsietshawus or Killamuk. To this, the present writer adds the BillechÚla.

XI. The query as the likelihood of the Straits of Fuca vocabulary having been Mozino's finds place here. The two are different: though both may have been collected by Mozino. Each is to be found in Buschmann, who, exaggerating the isolation of Wakash, NÚtka, and Tlaoquatch forms of speech, separates them too decidedly. Out of nineteen words compared nine are not only alike but admitted by him to be so.

The Billechula.—This lies intermediate to the Hailtsa and Atna groups; being (apparently) more akin to the latter than the former. Of the Atna dialects, it seems most to approach the Piskwaus.

The Chinuk.—The Chinuk of which the Watlala of Hale is variety is more like the Nsietashawus or Killamuk than aught else.

The Kalapuya.—The harshness of the Kalapuya is an inference from its orthography. It is said, however, to be soft and flowing i. e. more like the Sahaptin and Shoshoni in sound than the Chinuk, and Atna.

The Jakon.—This has affinities with the Chinuk on one side, and the Lutuami on the other; i. e. it is more like these two languages than any other. The likeness, however, is of the slightest.

Miscellaneous affinities.

The Sahaptin.—The Sahaptin, Shoshoni and Lutuami groups are more closely connected than the text makes them.

The Shoshoni (Paduca) group.—The best general name for this class is, in the mind of the present writer, Paduca; a name which was proposed by him soon after his notification of the affinity between the Shoshoni and the Comanch, in A.D. 1845. Until then, the two languages stood alone; i. e. there was no class at all. The Wihinast was shewn to be akin to the Shoshoni by Mr. Hale; the Wihinast vocabulary having been collected by that indefatigable philologue during the United States Exploring Expedition. In Gallatin's Report this affinity is put forward with due prominence; the Wihinast being spoken of as the Western Shoshoni.

In '50 the Report of the Secretary at War on the route from San Antonio to El Paso supplied an Utah vocabulary; which the paper of May '53 shews to be Paduca.

In the Report upon the Indian Tribes &c. of '55, we find the Chemehuevi, or the language of one of the Pah-utah bands "for the first time made public. It agrees" (writes Professor Turner) "with Simpson's Utah and Hale's East Shoshoni."

Carvalho (I quote from Buschmann) gives the numerals of the Piede (Pa-uta) of the Muddy River. They are nearly those of the Chemehuevi.

For the Cahuillo see below.

Is the Kioway Paduca? The only known Kioway vocabulary is one published by Professor Turner in the Report just alluded to. It is followed by the remark that "a comparison of this vocabulary with those of the Shoshoni stock does, it is true, show a greater degree of resemblance than is to be found in any other direction. The resemblance, however, is not sufficient to establish a radical affinity, but rather appears to be the consequence of long intercommunication."

For my own part I look upon the Kioway as Paduca—the value of the class being raised.

XIII. The Capistrano group.—Buschmann in his paper on the Netela and Kizh states, after Mofras, that the Juyubit, the Caguilla, and the Sibapot tribes belong to the Mission of St. Gabriel. Turner gives a Cahuillo, or Cawio, vocabulary. The district from which it was taken belonged to the St. Gabriel district. The Indian, however, who supplied it had lived with the priests of San Luis Rey, until the break-up of the Mission. Whether the form of speech he has given us be that of the Mission in which he lived or that of the true Cahuillo district is uncertain. Turner treats it as Cahuillo; at the same time he remarks, and shews, that it is more akin to the San Luis Rey dialect than to any other.

But it is also akin to the Chemeuevi, which with it is tabulated; a fact which favours the views of Hale respecting its San Capistrano affinities rather than those of Buschmann—Hale making them Paduca.

A vocabulary, however, of the unreclaimed Cahuillo tribes—the tribes of the mountains as opposed to the missions—is still wanted.

P. 353. Now comes the correction of a statement in p. 353—"the language of San Luis El Rey which is Yuma, is succeeded by that of San Luis Obispo, which is Capistrano."—This is an inaccuracy; apparently from inadversion. A reference to the Paternosters of pp. 304-305 shews that the San Luis Rey, and the San Juan Capistrano forms of speech are closely allied. Meanwhile, the San Fernando approaches the San Gabriel, i. e. the Kizh.

See also Turner, p. 77—where the name Kechi seems, word for word, to be Kizh. The Kizh, however is a San Gabriel form of speech.

XIV. The Yuma group.—Turner gives a Mojave, or Mohavi vocabulary; the first ever published. It is stated and shewn to be Yuma. The Yabipai, in the same paper, is inferred to be Yuma; containing, as it does, the word

The Mohave vocabulary gives the following extracts,

We leave California with the remark that in Ludwig's Literature of the American Aboriginal Languages Mr. Bartlett's vocabularies for California bear the following titles.

See Californians.

There is also a Piros vocabulary for the parts about El Paso: also a notice (under the word) that the Mutsunes Indians speak a dialect of the Soledad.

Old California.—As a general rule, translations of the Pater Noster shew difference rather than likeness: in other words, as a general rule, rude languages are more alike than then Pater Nosters make them. The reasons for this lie in the abstract nature of many of the ideas which it is necessary to express; but for the expression whereof the more barbarous forms of speech are insufficient.

This creates the necessity for circumlocutions and other expedients. In no part of the world is this more manifest than in Old California; a district for which our data are of the scantiest. I think, however, that they are sufficient to shew that the Northern forms of speech, at least, are Yuma.

The Pima group.—One of Mr. Bartlett's vocabularies is of the Opata form of speech. (Ludwig.)

Tequima, according to the same authority is another name for the same language: in which there is a vocabulary by Natal Lombardo; Mexico. 1702, as well as an Arte de la Lengua Tequima, vulgarmente llamada Opata.

A Vocabulario de las Lenguas Pima, Eudeve, y Seris is said, by De Souza, to have been written by Fr. Adamo Gilo a Jesuit missionary in California.—Ditto—v. Pima.

Exceptions, which the present writer overlooked, are taken in the Mithridates to the statement that the Opata and Eudeve Pater-nosters represent the Pima Proper. They agree with a third language from the Pima country—but this is not, necessarily, the Pima. Hence, what applies to the Pimerian may or may not apply to the Pima Proper.

Nevertheless, the Pima belongs to the same class—being, apparently, more especially akin to the Tarahumara. I have only before me the following Tarahumara words (i. e. the specimens in the Mithridates) through which the comparison can be made. They give, however, thus much in way of likeness and difference.

Buschmann connects the Pima with the Tepeguana.

Another complication.—In Turner's Extract from a MS. account of the Indians of the Northern Provinces of New Spain I find that Opa (Opata?) is another name for the Cocomaricopas whose language is that of the Yuma. This is true enough—but is the Opata more Yuma than the text (which connects it with the Hiaqui &c.) makes it?

The Pima, Hiaqui, Tubar, Tarahumara, and Cora as a class.—An exception to the text is indicated by the footnote of page 357. The Mithridates connects the Cora and Tarahumara with the Astek and with each other. The Astek elements of the Hiaqui, as indicated by Ribas are especially alluded to. So are the Tarahumara affinities of the Opata. All this is doing as much in the way of classification as is done by the present author—as much or more.

As much, or more, too is done by Buschmann; who out of the Cora, Tarahumara, Tepeguana and Cahita (the latter a representation of the section to which the Yaqui belongs) makes his Sonora Class—Sonorischer Sprachstamm. As a somewhat abnormal member of this he admits the Pima.

Of the Guazave there is a MS. Arte by P. Fernando Villapane—Ludwig.

That the data for the Tepeguana are better than the text makes them has already been suggested. Buschmann has used materials unknown to the present writer.

See Ludwig in voc. Tepeguana.

Pirinda and Tarasca.—The statement that there is a Pirinda grammar is inaccurate. There is one of the Tarasca; to which the reader is referred.

But this is not all. Under the title Pirinda in Ludwig we find that De Souza says of Fr. Juan Bravo, the author of a grammar of the Lengua Tarasca "fue maestro peritissimo de la lengua Pirinda llamada Tarasca." This makes the two languages much more alike than the present paper makes them. The present paper, however, rests on the Pater-nosters. How inconclusive they are has already been indicated.

? The following table, the result of a very limited collation gives some miscellaneous affinities for the Otomi.

The other two are as follows.

(2.)
The Otomi with the languages akin to the Chinese en masse.

(2.)
The Maya, with the languages akin to the Chinese en masse.

The Acoma.—Two vocabularies from a tribe from the Pueblo of San Domingo, calling themselves Kiwomi, and a third of the Cochitemi dialect, collected by Whipple, are compared, by Turner, with the Acoma, of which they are dialects. Turner proposes the names Keres for the group. Buschmann, writing after him, says, "I name this form of speech Quera"—"ich nenne dies Idiom Quera."

The notice of the "outward signs" is not so clear as it should be. It means that two of the languages, the Taos and Zuni, run into polysyllabic forms—probably (indeed almost certainly) from composition or inflexion; whereas the Tesuque (which is placed in contrast with the Zuni) has almost a monosyllabic appearance. This phenomenon appears elsewhere; e. g. in the Attacapa, as compared with the tongues of its neighbourhood. Upon the whole, the Zuni seems to be most aberrant of the group—saving the Moqui, which has decided Paduca affinities. They are all, however, mutually unintelligible; though the differences between them may easily be over-valued.

Texas.—p. 101.—"Ini and Tachi are expressly stated to be Caddo, &c. as it is from the name of the last that the word Texas is derived &c."—The name Teguas is a name (other than native) of the population which calls itself Kiwomi. Word for word, this may (or may not) be Taos. It is only necessary to remember the complication here indicated. The exact tribe which gave the name to Texas has yet to be determined.

The Witshita.—Allied to one another the Kechis and Wacos (Huecos) are, also, allied to the Witshita.—See Turner, p. 68.

Turner makes these three languages Pawni. In the present text the Witshita is made Caddo. It is made so on the strength of the numerals—perhaps overhastily.

That a language may be Pawni without ceasing to be Caddo, and Caddo without losing its place in the Pawni group is suggested in the beginning of the paper. Turner's table (p. 70), short as it is, encourages this view.

The truth is that the importance of the Caddos and Pawnis, from an ethnological point of view, is inordinately greater than their importance in any other respect. They are, however, but imperfectly known.

In Gallatin's first paper—the paper of the ArchÆologia Americana—there is a Caddo vocabulary and a Pawni vocabulary; and all that be said of them is that they are a little more like each other, than they are to the remaining specimens.

When the paper under notice was published the Riccaree was wholly unknown. But the Riccaree, when known, was shewn to be more Pawni than aught else. This made the Pawni a kind of nucleus for a class.

? Somewhat later the Caddo confederacy in Texas took prominence, and the Caddo became a nucleus also.

The true explanation of this lies in the highly probable fact that both the Caddo and Pawni are members of one and the same class. At the same time I am quite prepared to find that the Witshita (though compared with the Caddo by myself) is more particularly Pawni.

That the nearest congeners of the Caddo and Pawni class were the members of the Iroquois, Woccoon, Cherokee, and Chocta group I believed at an early period of my investigations; at a time (so to say) before the Riccarees, and the Californian populations were invented. If this doctrine were true, the Caddo (Pawni) affinities would run eastwards. They may do this, and run westwards also. That they run eastwards I still believe. But I have also seen Caddo and Pawni affinities in California. The Caddo numeral one = whiste; in Secumne and Cushna wikte, wiktem. Again the Caddo and Kichie for water = koko, kioksh. Meanwhile kik is a true Moquelumne form. This I get from a most cursory inspection; or rather from memory.

Upon the principle that truth comes out of error more easily than confusion I give the following notice of the distribution or want of distribution of the numerous Texian tribes.

1. *Coshattas—Unknown.
2. Towiach—Pawni (?).
3. Lipan—Athabaskan (?).
4. *Alish, or Eyish—Caddo (?).
5. *Acossesaw—Unknown.
6. Navaosos—Navahos (?).
7. *Mayes—Attacapa (?).
8. *Cances—Unknown.
9. Toncahuas—Are these the Tonkaways, amounting, according to Stem, to 1152 souls? If so, a specimen of their language should be obtained. Again—are they the Tancards? Are they the Tunicas? If so, they may speak Choctah.
10. Tuhuktukis—Are these the Topofkis, amounting to 200 souls? If so a specimen of their language, eo nomine, is attainable.
11. Unataquas, or Andarcos—They amount, according to Stem, to 202 souls. No vocabulary, eo nomine, known. Capable of being obtained.
12. Mascovie—Unknown.
13. Iawani or Ioni—Caddo? Amount to 113 souls. Specimen of language, eo nomine, capable of being obtained.
14. Waco—Wico?—Pawni.
15. *Avoyelle—Unknown.
16. 17. Washita—Kiche—Pawni.
17. *Xaramene—Unknown.
18. *Caicache—Unknown.
19. *Bidias—Unknown.
20. Caddo—Caddo.
21. Attacapa—Attacapa.
22. Adahi—Adahi.
23. Coke—Carackahua.
24. Carankahua—Attacapa (?).
25. Towacano—Numbering 141 souls. Is this Towiach?
26. Hitchi—Kichi (?).
27. *Nandako.—Caddo (?).
28. *Nabadaches.—Caddo (?).
29. *Yatassi.
30. *Natchitoches.—Adahi (?).
31. *Nacogdoches.—Adahi (?).
32. Keyes.—Adahi (?).

These last may belong as much to Louisiana as to Texas—as, indeed, may some of the others. Those marked * are apparently extinct. At any rate, they are not found in any of the recent notices.

Finally, Mr Burnett mentions the San Pedro Indians.

The previous list shews that the obliteration of the original tribes of Texas has been very great. It shews us this at the first view. But a little reflection tells us something more.

Like Kanzas and Nebraska, Texas seems to have scarcely any language that is peculiar to itself; in this respect standing in strong contrast to California. The Caddo belongs to the frontier. The Pawni forms of speech occur elsewhere. The Adahi is probably as much the property of Louisiana as of Texas. The Cumanch, Chocta &c. are decidedly intrusive. The nearest approach to a true Texian form of speech is the Attacapa. No wonder it is isolated.

The Adahi, is has, at least the following affinities.

Mexico-Guatemala.—The details of the languages of Mexico and Guatemala that are neither Mexican Proper (Astek) or Maya are difficult. Availing myself of the information afforded by my friend Mr. Squier, and the bibliographical learning of Ludwig, I am inclined to believe

1. That all the following forms of speech are Maya; viz. Chiapa, Tzendal (Celdal), Chorti, Mam, Pocoman (Poconchi), Populuca, Quiche, Kachiquel, Zutugil (Yutukil), Huasteca.

2. That the Zoque, Utlateca, and Lacondona may or may not be Maya.

3. That the Totanaca; and

4. The Mixteca are other than Maya.

5. That, if the statement of Hervas be correct, the Zapoteca, the Mazateca, the Chinansteca, and the Mixe are in the same category.

The Tlapaneka according to Humboldt is a peculiar language.—Ludwig in voc.

I have done, however, little or nothing, in the way of first hand work with the languages to the South of Sinaloa and the West of Texas. I therefore leave them—leave them with a reference to Ludwig's valuable Bibliotheca Glottica, for a correction of my statement respecting the non-existence of any Indian forms of speech in New Grenada. The notices under v. v. Andaquies, Coconucos, Correquajes, Guaques, Inganos, will shew that this is far from being the case.

The present paper has gone over so large a portion of North America that it is a pity not to go over the remainder. The ethnology of the Canada, and the British possessions akin to Canada contains little which is neither Eskimo or Algonkin, Iroquois or Athabaskan. Of new forms of speech like those of which Oregon and California have given so many instances it exhibits none. Everything belongs to one of the four above-named classes. The Bethuck of Newfoundland was Algonkin, and so were the Blackfoot, the Shyenne and Arrapaho. Indeed, as has been already stated, the Eskimo and Athabaskan stretch across the Continent. The Blackfoot touches the Rocky Mountains. Of the Sioux class the British possessions shew a sample. The Red River district is Assineboin; the Assineboins being Sioux. So are a few other British tribes.

Upon the whole, however, five well-known families give us all that belong to British America to the East of the Rocky Mountains. As the present paper is less upon the Algonkin, Sioux and like classes than upon the distribution of languages over the different areas of North America this is as much as need be said upon the subject.

For the Northern two-thirds of the United States, East of the Mississippi, the same rule applies. The Sioux area begins in the West. The Algonkin class, of which the most Northern branch belongs to Labrador, where it is conterminous with the Eskimo, and which on the west contains the Blackfoot reaches as far south as South Carolina—the Nottoways being Algonkin. The enormous extent of this area has been sufficiently enlarged on. Meanwhile, like islands in an Ocean, two Iroquois district shew themselves. To the north the Iroquois, Hurons and others touch the Lakes and the Canadians frontier, entirely separated from the Tuscaroras who give a separate and isolated area in California. Whether the Iroquois area, once continuous, has been broken-up by Algonkin encroachments, or whether the Iroquois &c. have been projected into the Algonkin area from the South, or, whether vice versa, the Tuscaroras are to be considered as offsets from the North is a matter for investigation. The present writer believes that south of N. L. 45. (there or there about) the Algonkins are intrusive.

N. L. 35. cuts the Cherokee, the Woccoon, the Catawba, and the Chocta area—to the west of which lies of the Mississippi.

Between the frontier of Texas, the aforesaid parallel, and the Ocean we have Florida, Alabama, Mississippi, and Louisiana.

Now here the displacement has been considerable. The part played by the Algonkins, Iroquois, and (it may be added) the Sioux is here played by the Cherokees, the Choctahs, and the Creeks. Whatever is other than Creek, Choctah, and Cherokee is in a fragmentary form. The details of what we know through vocabularies are as follows:—

1. The Woccon—extinct, and allied to——
2. The Catawba—also extinct. These belonged to the Carolinas. The Woccon and Catawba vocabularies are mentioned in the Mithridates.
3. The Tinqua—see Ludwig.
4. The Timuacuana—see p. 377.
5. The Uche—of this we find a specimen in the ArchÆologia Americana. The tribe belongs to the Creek confederacy and must be in a very fragmentary state.
6. The Natchez—on the Mississippi, facing the Caddos, Adahi.
7. The Chetimacha.—In Louisiana. Vocabulary in ArchÆologia Americana.

In the way of internal evidence (i.e. the evidence of specimens of language) this is all we have what may be called the fragmentary languages of the South Eastern portion of the United States. Of the Choctah, Creek, Chikkasah, and Cherokee we have an abundance, just as we have of the Algonkin and Eskimo. It is, however, the fragmentary tribes, the probable representatives of the aboriginal population, which we more especially seek.

As may be expected the fragmentary languages are (comparatively speaking) isolated. The Woccon and Catawba, indeed, are thrown into the same class in the Mithridates: but the Natchez and Uche are, by no means, closely akin. Why should they be? Such transitional forms as may once have existed have been obliterated. Nevertheless, both have miscellaneous affinities.

So much for the languages represented by specimens. In the way of external evidence I go no further than the Mithridates, and the ArchÆologia.

With the exception of the Woccons the Catawba and a few words from the Timuacana, the Mithridates, gives no specimens—save and except those of the Choctah, Cherokees, and Chikkasah. These two last it looks upon as the representative languages and calls them Mobilian from Mobile. Hence, the question which was put in Texas is, mutatis mutandis, put in Florida. What languages are Mobilian? What other than Mobilian?

The Woccons are either only or chiefly known through a work of Lawson's. They were conterminous with the Algonkin Pamticoughs (intrusive?), and the Cherokees.

The Catawba lay to the south of the Woccon. Their congeners are said to be

1. The Wataree;
2. The Eeno—Compare this name with the Texian Ini;
3. The Chowah, or Chowan;
4. The Congaree;
5. The Nachee—Compare with Natchez; word for word;
6. The Yamassee;
7. The Coosah—Compare (word for word) Coosada, and
8. Coshatta.

In the South lay the Timuacana—of which a few words beyond the numerals are given.

In West Florida and Alabama, the evidence (I still follow the Mithridates) of Dr. Pratz scarcely coincides with that of the account of Alvaz NuÑez de Vaca. This runs thus.

In the island of Malhado were spoken languages of

• 1. The Caoques;
• 2. The Han.

On the coast—

• 3. The Choruico—Cherokee?
• 4. The Doguenes.
• 5. The Mendica.
• 6. The Quevenes.
• 7. The Mariames.
• 8. The Gualciones.
• 9. The Yguaces.
• 10. The Atayos—Adahi? This seems to have been a native name—"die sich Atayos nennen."
• 11. The Acubadaos.
• 12. The Quitoles.
• 13. The Avavares—Avoyelles?
• 14. The Muliacone.
• 15. The Cutalchiche.
• 16. The Susola.
• 17. The Como.
• 18. The Camole.

Of migrants from the East to the West side of the Mississippi, the Mithridates gives—

1. The Pacana, conterminous with the Attacapas.
2. The Pascagula.
3. The Biluxi.
4. The Appalache.

The Taensa are stated to be a branch of the Natchez.

The Caouitas are, perhaps, word for word the Conchattas; also the Coosa, Coosada, Coshatta.

The Stincards are, word for word, the Tancards=Tuncas=Tunicas.

Dr. Sibley gives us Chetimacha as a name; along with specimens of the Chetimacha, Uche, Natchez, Adahi, and Attacapa as languages.

Word for word, Chetimacha seems to Checimeca; Appelusa, Apalach; Biluxi (perhaps the same); Pascagoula, Muscogulge. How, however, did Chichimeca get so far westwards?

We are scarcely, in the condition to speculate much concerning details of the kind. It is sufficient to repeat the notice that the native languages of the parts in question are in a fragmentary condition; the Uche being the chief representative of them. Whether it were Savaneric[49], or not, is uncertain. It is, certainly, not Shawanno, or Shawno, i. e. Algonkin. On the contrary it is, as is to be expected, from the encroachments and displacements of its neighbourhood a very isolated language—not, however without miscellaneous affinities—inter alia the following.

Such our sketch of the details. They give us more affinities than the current statements concerning the glossarial differences between the languages of the New World suggest. It is also to be added that they scarcely confirm the equally common doctrine respecting their grammatical likeness. Doing this, they encourage criticism, and invite research.

There is a considerable amount of affinity: but it is often of that miscellaneous character which baffles rather than promotes classification.

There is a considerable amount of affinity; but it does not, always, shew itself on the surface. I will give an instance.

One of the first series of words to which philologues who have only vocabularies to deal with have recourse, contains the numerals; which are, in many cases, the first of words that the philological collector makes it his business to bring home with him from rude countries. So generally is this case that it may safely be said that if we are without the numerals of a language we are, in nine cases out of ten, without any sample at all of it. Their value as samples for philological purposes has been noticed in more than one paper of the present writer's here and elsewhere; their value in the way of materials for a history of Arithmetic being evident—evidently high.

But the ordinary way in which the comparisons are made between the numerals gives us, very often, little or nothing but broad differences and strong contrasts. Take for instance the following tables.

No wonder that the tongues thus represented seem unlike.

But let us go farther—in the first place remembering that, in most cases, it is only as far as five that the ruder languages have distinct numerals; in other words that from six onwards they count upon the same principle as we do after ten, i. e. they join together some two, or more, of the previous numerals; even as we, by adding seven and ten, make seven-teen. The exact details, of course, differ; the general principle, however, is the same viz.: that after five the numerals become, more or less, compound, just as, with us, they become so after ten.

With this preliminary observation let us ask what will be the Kamskadale for seven when nittanu=two, and kumdas=five. The answer is either nittanu-kumdas or kumdas-nittanu. But the Kamskadale happens to have a separate word for six, viz. kiekoas. What then? The word for seven may be one of two things: it may be either = 6 + 1, or 5 + 2. The former being the case, and kemmis=one, the Kamskadale for seven should be either kemmis-kilkoas or kilkoas-kemmis. But it is neither one nor the other. It is ittakh-tenu. Now as eight=tshok-tenu we know this word to be compound. But what are its elements? We fail to find them amongst the simpler words expressive of one, two, three, four, five. We fail to find them amongst these if we look to the Kamskadale only—not, however, if we go farther. The Aleutian for one=attak-on; the Aleutian for six=attu-on. And what might be the Aleutian for seven? Even attakh-attun, little more than ittakh tenu in a broader form.

The Jukahiri gives a similar phenomenon.

Such is the notice of the care with which certain comparisons should be made before we venture to commit ourselves to negative statements.

There is an affinity amongst the American languages, and (there being this) there are also the elements of a classification. The majority, however, of the American languages must be classified according to types rather than definitions. Upon the nature of this difference, as well as upon the cause I have written more fully elsewhere. It is sufficient for present purposes to say that it applies to the languages of North America in general, and (of these) to those of the parts beyond the Rocky Mountains more especially. Eskimo characteristics appear in the Athabaskan, Athabaskan in the Koluch forms of speech. From these the Haidah leads to the Chimmesyan (which is, nevertheless, a very outlying form of speech) and the Hailtsa, akin to the Billechula, which, itself, leads to the Atna. By slightly raising the value of the class we bring in the Kutani, the Nutkan and the Chinuk.

In the Chinuk neighbourhood we move via the Jakon, Kalapuya, Sahaptin, Shoshoni, and Lutuami to the languages of California and the Pueblos; and thence southwards.

In American languages simple comparison does but little. We may test this in two ways. We may place, side by side, two languages known to be undoubtedly, but also known to be not very closely, allied. Such, for instance, are the German and Greek, the Latin and Russian, the English and Lithuanic, all of which are Indo-European, and all of which, when placed in simple juxta-position, by no means show themselves in any very palpable manner as such. This may be seen from the following table, which is far from being the first which the present writer has compiled; and that with the special view of ascertaining by induction (and not a priori) the value of comparisons of the kind in question.

Again—the process may be modified by taking two languages known to be closely allied, and asking how far a simple comparison of their vocabularies exhibits that alliance on the surface, e.g.:—

The difference is great: but the two forms of speech are mutually intelligible. On the other hand, the Cayuse and Willamet are more alike than the English and Latin.

Next to the details of our method, and the principles of our classification, the more important of the special questions command attention. Upon the relations of the Eskimo to the other languages of America I have long ago expressed my opinion. I now add the following remarks upon the prevalence of the doctrine which separated them.

Let us imagine an American or British ethnologist speculating on the origin and unity of the European populations and arriving, in the course of his investigations, at Finmark, or any of those northern parts of Scandinavia where the Norwegian and Laplander come in immediate geographical contact. What would be first? Even this—close geographical contact accompanied by a remarkable contrast in the way of the ethnology: difference in habits, difference in aptitudes, difference in civilisation, difference of creed, difference of physical form, difference of language.

But the different manner in which the southern tribes of Lapland comport themselves in respect to their nearest neighbours, according as they lie west or east, illustrates this view. On the side of Norway few contrasts are more definite and striking than that between the nomad Lap with his reindeer, and reindeer-skin habiliments and the industrial and highly civilized Norwegian. No similarity of habits is here; no affinity of language; little on intermixture, in the way of marriage. Their physical frames are as different as their moral dispositions no and social habits. Nor is this difficult to explain. The Norwegian is not only a member of another stock, but his original home was in a southern, or comparatively southern, climate. It was Germany rather Scandinavia; for Scandinavia was, originally, exclusively Lap or Fin. But the German family encroached northwards; and by displacement after displacement obliterated those members of the Lap stock whose occupancy was Southern and Central Scandinavia, until nothing was left but its extreme northern representatives in the most northern and least favored parts of the peninsula. By these means two strongly contrasted populations were brought in close geographical contact—this being the present condition all along the South Eastern, or Norwegian, boundary of Lapland.

But it is by no means the present condition of those parts of Russian Lapland where the Lap population touches that of Finland Proper.

Here, although the Lap and Fin differ, the difference lies within a far narrower limit than that which divides the Lap from the Norwegian or the Swede. The stature of the Lap is less than that of the Fin; though the Fin is more short than tall, and the Lap is far from being so stunted as books and pictures make him. The habits, too, differ. The reindeer goes with the Lap; the cow with the Fin. Other points differ also. On the whole, however, the Fin physiognomy is Lap, and the Lap Fin; and the languages are allied.

Furthermore—the Fin graduates into the Wotiak, the Zirianean, the Permian; the Permian into the Tsheremiss, the Mordvin &c. In other words, if we follow the Lap eastwards we come into a whole fancy of congeners. On the west, however, the further we went, the less Lap was everything. Instead of being Lap it was Norwegian, Swedish, Danish, or German. The last of those, however, would lead us into the Sarmatian family, and this would bring us round to the Fins of South Finland. The time, however, may come when Russia will have so encroached upon the Fin populations to the south of the Arctic Circle as for the Lap and Slave to come in immediate contact; and when this contact is effected there will be contrast also—contrast less strong, perhaps, than that between the Lap and Swede, but still contrast.

Mutatis mutandis—this seems to have been the case with the Eskimo and the North American Indians as they are popularly called—popularly but inaccurately; inasmuch as the present writer considers the Eskimo to be as truly American as any other occupants of the soil of America. On the East there has been encroachment, displacement, and, as an effect thereof, two strongly contrasted populations in close geographical contact—viz.: the Eskimos and the northern members of the Algonkin family. On the west, where the change has been less, the Athabaskans, the Kolutshes, and the Eskimos graduate to each other, coming under the same category, and forming part of one and the same class; that class being by no means a narrow, though not an inordinately, wide one.

Another special question is that concerning the origin of the Nahuatl, Astecs, or Mexicans. The maritime hypothesis I have abandoned. The doctrine that their civilisation was Maya I retain. I doubt, however, whether they originated anywhere. By this I mean that they are, though not quite in situ, nearly so. In the northermost parts of their area they may so entirely. When I refined on this—the common sense—view of them I was, like many others, misled by the peculiar phonesis. What it is may be better seen by an example than explained. Contrast the two following columns. How smoothly the words on the right run, how harshly sound (when they can be sounded) those of the left. Not, however, that they give us the actual sounds of the combination khl &c. All that this means is that there is some extraordinary sound to be expressed that no simple sign or no common combination will represent. In Mr. Hale's vocabularies it is represented by a single special sign.

Now if the Astec phonesis be more akin to the Selish and its congeners than to the Shoshoni and other interjacent forms of speech, we get an element of affinity which connects the more distant whilst it separates the nearer languages. Overvalue this, and you may be misled.

Now, not to mention the fact of this phonesis being an overvalued character, there is clear proof in the recent additions to the comparative philology of California that its distribution is, by no means, what it was, originally, supposed to be. This may be seen from the following lists.

From the North of California.

(1.)

(2.)

In the South of California.

I cannot conclude without an expression of regret that the great work of Adelung is still only in the condition of a second, or (at best) but a third edition. There is Vater's Supplement, and JÜlg's Supplement to Vater. But there is nothing that brings it up to the present time.

Much might be done by Buschmann and perhaps others. But this is not enough. It requires translation. The few French writers who treat on Ethnological Philology know nothing about it. The Italians and Spanish are, a fortiori, in outer darkness as to its contents. The Russians and Scandinavians know all about it—but the Russians and Scandinavians are not the scholars in whose hands the first hand information falls first. The Americans know it but imperfectly. If Turner has had easy access to it, Gallatin had not: whilst Hales, with great powers, has been (with the exception of his discovery of the Athabaskan affinities of the Umkwa and Tlatskanai, out of which Turner's fixation of the Apatch, Navaho, and Jecorilla, and, afterwards, my own of the Hoopah, seems to have been developed,) little more than a collector—a preeminent great collector—of raw materials. Nevertheless, the Atna class is his.

However, the Mithridates, for America at least, wants translation as well as revision. It is a work in which many weak points may be (and have been) discovered. Klaproth, himself a man who (though he has saved many an enquirer much trouble) has but few friends, has virulently attacked it. Its higher classifications are, undoubtedly, but low. Nevertheless, it is not only a great work, but the basis of all others. Should any one doubt its acumen let him read the part which, treating on the Chikkasah, demurrs to the identification of the Natchez with that and other forms of speech. Since it was written a specimen of the Natchez language has shewn its validity.

I think that the Natchez has yet to take its full importance. If the language of the Taensas it was, probably, the chief language of Tennessee. But the Creek, or Muscogulge, broke it up. Meanwhile the fragmentary Catawba, with which I believe that the Caddo was connected had its congeners far to westward.

I also think that the Uche represents the old language of Florida—the Cherokee being conterminous with the Catawba. If so, the doctrine of the fundamental affinity between the Pawni, Caddo, Catawba, and Cherokee gains ground.

The Uche demands special investigation. The Tinquin and Timuacana should be compared with it. Then why are they not? Few works are more inaccessible than a Spanish Arte, Diccionario, or Catecismo. The data for these enquiries, little known, are still less attainable. Without these, and without a minute study, of the first-hand authorities we can do but little but suggest. All that is suggested here is that the details of Florida (in its widest sense) and Louisiana must be treated under the doctrine that the aborigines are represented by the congeners of the Woccon, Catawba, Uche, Natchez, Tinquin, and Timuacana, inordinately displaced by the Cherokees and Creeks; who (for a great extent of their present area) must be considered as intrusive.

---