nian and Turanian branches. We beg to call this definitively the Japhetic race. In many parts we know that the Turanian race has preceded the Iranian: its language certainly represents an anterior step or preceding degree of development. In some parts we found that the Turanian race succeeded to a still older native element.

We now return to the precursor of Iranism in western Asia and in Egypt. Semitism appeared to us as a one-sided progress of Chamism. Canaan (the ante-judaic inhabitants of Palestine and Phoenicia) is not literally a child of Cham, that is to say, a scion of the Egyptian stock: perhaps geographically, but not genealogically. But Canaan is called in Genesis the son of Cham, because the Canaanites, in the Abrahamitic period, as again in the Mosaic ime, left lower Egypt and occupied Palestine and Tyre. Canaan came out of Cham; certainly the Semitic idiom is in itself a child of Cham, by being a levelopment of that primitive (Asiatic) Chamism, which became fixed in the Egyptian. For the Egyptian language is as certainly the primitive formaion of the Euphrates and Tigris territory, fixed in Africa and preserved by he Egyptians, as the Icelandic is the old Norse fixed in that island. All this ollows out of the facts furnished by Egyptology, if the principles above estalished are applied to them.

The Semitic formation itself occupied Abyssinia; and the Berber language elongs evidently to the same stock. But what can we say of the rest of Africa?

Here late researches have opened a new and great field of the most intersting character. We allude in particular to the labours of Tutschek, and the nalytical inquiries of Von Gablenz and Ewald; but above all to the giganc and truly admirable labours of that indefatigable German Messenger of he Church-Missionary Society of England, the Rev. John Lewis Krapf, whose ompared manuscript grammar and dictionary of the Sawahili language and te cognate dialects of the Wanicka and Wakamba tribes, with introductions nd numerous translations, have been entrusted to me by the enlightened ecretary of that Society, the Rev. Henry Venn. These, and similar works bout the south-eastern languages of Africa, have entirely destroyed those nfounded notions of an infinite number of rude and poor tongues. We now now, that dialects of the Galla language, which in the North joins the Abyssinian, a very fine specimen of grammatical structure and euphonic ormation, are spoken, at least as far as the fifth degree south of the equator; hat it penetrates deeply into the continent along the eastern coast of Africa; hat it is joined by the noble Caffre idioms, which also enter far into the inerior; and that the Congo idioms on the western coast, if not cognate, are t least very analogous in structure, as the Galla and Caffre languages are ecidedly among themselves*.

They besides all bear on them vestiges of primitive affinity, according to ur principles, with the great tripartite stock. But if we are asked, do these nguages belong to Chamism, or do they stand on the degree represented by emitism? we are obliged to answer, neither the one nor the other. On the ontrary, applying to them the principles we have endeavoured to establish s the general principles of development, we must confess that they stand on aphetic ground. The primitive state of Chamism, exhibiting the germ both f Semitism and of Japhetism, is evidently left behind in those advanced

* At the moment that we are carrying these sheets through the press (April 26, 1848) e receive the first and second number of the Zeitschrift der deutschen morgenländischen Gesellschaft,' and find in it Prof. Pott's learned article on the languages of the Caffre and Congo tribes. We beg particularly to refer our readers to the ingenious and acute observaons of Prof. Schott, brought forward in this article.

formations. There is a further development, but that development does not run in the Semitic line. In the Semitic formations, the copula is constantly expressed by the pronominal form (he), whereas the Iranian, as well as the Turanian, therefore all Japhetic languages, have already the more abstract and therefore more advanced verbal form (to be). In this decisive characteristic those African tongues side with the Japhetic. And so they do in the whole system of conjugation in opposition to the Semitic conjugation, as explained above. As the American and, in a certain manner, all Turanian languages are distinguished by their system of incorporation, and in particular of agglutination of words, together with that of postposition; thus these African idioms bear the type of prefixes and indicate the congruence of the parts of speech by changes in the initials of the words. Lepsius' preliminary observations respecting the two languages of the Upper Nile, discovered and analysed by him, lead to the supposition, that they too represent a decidedly greater advancement than the Egyptian.

The ulterior question then will be, have those languages (in form and in matter) passed through the Chamistic formation? And if so, is their primary formation reducible directly to Asiatic Chamism, or have they passed through the Egyptian?

We feel unable to answer these questions. The combined progress of the study of the language of Egypt and of those of central and southern Africa, will, perhaps in a few years, enable us to see clearly on this point.

Nor do we undertake to answer the question whether that wreck of the primitive language, that great monument of inorganic structure, the Chinese, can be linked by any scientific method to the other families of human speech, and thus, directly or indirectly, connected with the great tripartite civilizing family of mankind. But we add, there is no scientific proof that it cannot. Chinese philology, from a general point of view, is in its infancy. Morrison's merit consists in having given us a tonic dictionary, that is to say, a dictionary which does indeed deserve that name, an alphabetic collection of sounds, not a system of signs. But the execution of this laudable plan is very defective. The object of real philology must be to classify, with the necessary regard for the accent, the numberless significations of a full root or syllable, in such a manner, that the primitive significations may be discovered; for, as is the case still in the Egyptian, one sound comprises generally many roots now apparently identical, but originally different. The ancient style ought to be consulted for this purpose, if not exclusively, at least most particularly. Treating in this manner for instance, the roots ngò (the pronoun J), and the roots for father and mother (foo and moo), the original substantial meaning of the two last words will be easily ascertained, and the signification of recite, speak, speaking for ngò, lead to the natural origin of the pronominal signification. Nor is it less important to discover the original pronunciation and the phonetic rules of that language. Endlicher has been the first, in his Chinese grammar, to consult the language on this point. Finally, we cannot help thinking, that a system of transcriptions in Latin characters ought to be introduced in the tonic dictionary as well as in the grammar, and the ancient texts published in the same manner. The philological as well as the historical treasury of Chinese literature would thus become accessible to the philosophical and comparative study of that most interesting language. It is only by being taken up by general scholars in this way, that we may hope to gain a basis for the comparison of roots; although we are far from denying that the historical study of the signs by the professional Chinese scholar will also contribute much to the real understanding of that peculiar formation. The study of the Tibetan or Bhotiya language, and

that of the Burmese, would probably offer the nearest link between the Chinese and the more recent formations; but even the comparison with Sanscrit roots will not be without results.

It would be presumptuous to anticipate the issue of such well-prepared and sifted comparisons; but we have no hesitation in saying, that we incline to believe it will be in favour of the existence of a primitive connexion. There is a gap between that formation and all others; and that gap corresponds probably to that caused in the general development of the human race by great destructive floods, which separate the history of our race from its primordial origines. In this sense the Chinese may be called the monument of antediluvian speech. Indeed the first emigration from the cradle of mankind is said in Genesis to have gone eastward.

But whatever be the result, there is only one method of arriving at it, and that is a combination of accurate philological observation and analysis with philosophical principles, and with the collateral researches of history and of physiology. It is only by such a combination of researches that we can hope to fix definitively the place of the Chinese language in the general history of human speech, and to pronounce with historical certainty on the great questions connected with that problem. The difficulties are immense; but greater ones have been overcome in the last thirty years, and we believe that our method of distinguishing between primary and secondary formation, and of determining the succession of the phenomena of development, and thus of languages, will not be found entirely useless in the pursuit of those ulterior researches. At all events, we flatter ourselves that we have made good our assertion, that the Egyptologic discoveries are most intimately connected with the great question of the primeval language and civilisation of mankind, both in Asia and Africa, and that they give a considerable support to the opinion of the high, but not indefinite antiquity of human history, and to the hypothesis of the original unity of mankind and of a common origin of all languages of the globe.

London, 27 April, 1848.

J. BUNSEN.

On the Importance of the Study of the Celtic Language as exhibited by the Modern Celtic Dialects still extant. By Dr. CHARLES MEYER. THE subject on which I have ventured to address this meeting, is one which appears to me particularly well-adapted for the purposes of the British Association, as it deserves attention in a national as well as a general point of view, being connected not less with the special history of this country, both political and intellectual, than with the universal history of human civilization and intellect. I shall endeavour to point out the high importance, both historical and philological, of the study of the Celtic language, as exhibited by the modern Celtic dialects still extant, and shall lay before my hearers some of the new facts and views pertaining partly to general history, partly to the history of human language in particular, which, I believe, I have, in the course of my study of the Celtic language and literature, succeeded in discovering.

Modern Europe possesses two great dialects or languages, each composed of three separate idioms, which exhibit what we may call the modern Celtic. The word Celtic I use as a generic name for all the different idioms and dialects, evidently united amongst themselves by a systematic family-likeness of grammatical features, once spoken by the different nations and tribes, which in the Greek and Latin records of ancient history are usually designated under the general name of Kéλraι (Keλroí) and Celtæ *, and still spoken by their descendants. The two great dialects of modern Celtic may be seen, each with its three subdivisions, only one of which is actually extinct, on the following table:

1. The Gallic or British, comprehending

a. The Cymric or Welsh.

b. The Cornish (extinct).

c. The Armorican or dialect of Brittany (Bas Breton).

2. The Gaelic (Gadhelic) or Erse, comprehending―

a. The Fenic or Irish.

b. The Highland Scottish (Gaelic).

c. The Manx f.

It appears from this table, that five of the modern Celtic dialects, and four of those still extant, belong to this country, while the sixth, the Armorican or the dialect of Brittany, belongs to a district which, although situated in a foreign country, yet is British by its population, since it was in the fourth and fifth century of our era entirely colonized by British settlers, and named by them after their motherland, the latter becoming henceforth

* Uckert's Geography, vol. ii. p. 186.

†This table is on the whole the same as that given by Dr. Prichard in his Eastern Origin of the Celtic Nation.' I have only added the names Gallic and Fenic, both of which are of too frequent occurrence, and of too significative import in the ancient national records (particularly the Irish), not to find a place in a pedigree of the Celtic. As for the etymology of the two principal words of this pedigree, I derive the word Gadhel, Gael (Irish: Gaodhal, Gaoidhal, Gaedhil) from an old Celtic root gwydh sequi, comitari,—preserved (with the regular change of gw into ƒ) in the Irish words fuidh-im sequor, comitor; feadhan comitatus, clientela; feadha patronus; feidhil cliens-so as to give to the word Gadhel, Gael the signification follower, with reference either to the nomadic propensities and practices of the whole tribe, or to their habit of living in clanships. The name Gall (Gallus, Gaul), although it is used by the Irish writers in direct opposition to that of Gael, so much as to have acquired the general signification of foreigner, yet I am inclined to consider as another more mutilated form of the same word, a contraction namely of Gwadhal or Gwodhal. (cf. the name of S. Vodoalus.)