PHILOLOGY 779 inland Galla ; the first two lying along the Red Sea north of Semitic Abyssinia, the others south of it, to the equator. By some authorities (Lepsius, Bleek) there is added to the Hamitic family as a fourth division a group from extreme southern Africa, the Hottentot and Bushman languages. The ground of this classifica tion is the possession by the Hottentot of the distinction of gram matical gender, and even its designation by signs closely corre sponding to those used in the Ethiopia division. Others deny the sufficiency of this evidence, and rank the Hottentot as a separate group of African dialects, adding to it provisionally the Bushman, until better knowledge of the hitter shall show whether it is or is not a group by itself. If the Hottentot be Hamitic, we shall have to suppose it cut off at a very remote period from the rest of the family, and forced gradually southward, while all the time suffering mixture both of speech and of blood with the negro races, until the physical constitution of its speakers has become completely metamorphosed, and of its original speech no signs are left save those referred to above ; and while such exceptional phon etic peculiarities have been worked out as the use of the clicks or clucking sounds (see HOTTENTOTS) : and this must be regarded as at least extremely difficult. South- 4. Monosyllabic or South-eastern Asiatic Family. This body of eastern languages may well enough be the next taken up ; and here again Asiatic, (as was the case with the preceding family) on account of the prominent importance of one of its dialects and of the people speaking it -the Chinese people and language. The territory of the family includes the whole south-eastern corner of Asia : China on the north-east, Farther India in the south, and the high plateau of Tibet, with the neighbouring Himalayan regions, to the west ward. The ultimate unity of all these languages rests chiefly upon the evidence of their form, as being all alike essentially mono syllabic and isolating, or destitute of formal structure ; the material correspondences among them, of accordant words, are not sufficient to prove them related. The Chinese itself can be followed up, in contemporary records, to a period not far from 2000 B.C., and the language, the people, and their institutions, are then already in the main what they have ever since continued to be (see CHINA) ; the other leading tongues come into view much later, as they receive culture and religion from China on the one hand (the Anamites), or from India on the other (the Tibetans, Burmese, Siamese) ; and the territory includes great numbers of wild tribes unknown until our own times, whose race-relations and language- relations are as yet very obscure. Current opinion tends to regard the Anainites, Peguaus, and Cambodians as forming a more nearly related group or division of the family, and as having been the earlier population of Farther India, in part dispossessed and driven forward by the later intrusion from the north of Siamese and Burmese, of whom the former are more nearly related to the Chinese, and the latter to the Tibetans ; but these group ings rest as yet upon too slender evidence to be accepted with confidence. The character of the languages of this family, especially as in stanced by its most important member, the Chinese, has been pretty fully set forth in the general discussions above. They are languages of roots : that is to say, there is not demonstrable in any of their words a formative part, limiting the word, along with others simi- larl y characterized, to a certain office or set of offices in the formation of the sentence. That the words are ultimate roots, come down from the first period of language-making, we have no reason what ever to believe ; and they may possibly have passed through pro cesses of growth which equipped them with some scanty supply of forms ; but no evidence to that effect has yet been produced. The indications relied on to show an earlier polysyllabism in the family (though already in Chinese reduced to monosyllabism before the earliest historical appearance of the language, some 4000 years ago) are the comparatively recent loss of certain final mutes in Chinese words, and the presence on a considerable scale in Tibetan spelling of added initial and final consonants, now silent in the literary dialect, but claimed to be still uttered in some parts of the country. If the theory connecting these phenomena be established, the Tibetan will approve itself to be by far the most primitive of the dialects of the family, furnishing the key to the history of the rest. For further details respecting the various tongues of the mono syllabic family, the articles on the different divisions of its territory (BuRMAH, CHINA, SIAM, TIBET, &c.) may be consulted. The lan guages all alike show an addition to the resources of distinction possessed by languages in general, in the use of tones : that is to say, .words of which the alphabetic elements are the same differ in meaning according as they are uttered in a higher or a lower tone, with the rising or the falling inflexion, and so on. By this means, for example, the monosyllabic elements of the literary Chinese, numbering but 500 as we should write them, are raised to the number of about 1500 words. Ural- 5. Ural-All,aic (Scythian, Turanian] Family. China and Tibet Altaic, are bordered on the north and west by the eastern branches of another immense family, which stretches through central and northern Asia into Europe, overlapping the European border in Turkey, and reaching across it in Russia and Scandinavia to the very shore of the Atlantic. Usage has not so definitely determined as in the case of most other families by what name it shall be called ; Turanian is perhaps the commonest appellation, but also the most objectionable. Five principal branches are generally reckoned as composing the family. The two easternmost are the Tungusian, with the Manchu for its principal division, and the Mongol (see MONGOLS). Of these two the language is exceedingly simple in structure, being raised but little above the formlessness of the Chinese. The three others are : the Turkish or Tatar, the dia lects of which reach from the mouth of the Lena (Yakut) to Turkey in Europe ; the Samoyed, from the Altai down to the arctic shore of Asia, and along this to the White Sea an unimportant congeries of barbarous tribes ; and the Finno-Hungarian, including the tongues of the two cultivated peoples from which it takes its name, and also those of a great part of the population of northern and central Russia, to beyond the Ural Mountains, and finally the Lappish, of northern Scandinavia. The nearer relation of the Samoyed is with the Finno-Hungarian. The Turkish is a type of a well-developed language of purely agglutinative structure : that is, lacking that higher degree of integration which issues in internal change. Whether this degree is wholly wanting in Finnish and Hungarian is made a question ; at any rate, the languages named have no reason to envy the tongues technically called " inflective." Of a value not inferior to that of inflective characteristics is one that belongs to all the Ural-Altaic tongues, in varying measure and form, and helps to bind them together into a single family the harmonic sequence of vowels, namely, as between root and endings, or a modification of the vowels of the endings to agree with that of the root or its final syllable. While the physical race-characteristics known as Mongolian are wanting in the speakers of the western dialects of this family, they are conspicuously present in the people of Japan and Corea ; and hence the tendency of scholars to endeavour to connect the languages of the two latter countries, since they also are of agglutinative struc ture (see JAPAN and COREA), with the family now under treatment, as also with one another. Neither connexion, however, can at present be regarded as proved. Other languages of north-eastern Asia, too little known to group, and too unimportant to treat as separate families, may be mentioned here by way of appendix to their neighbours of the most diversified and widespread Asiatic family. They are the Aino, of Yezo and the Kurile Islands with part of the neighbouring coast ; the Kamchatkan ; and the Yukagir and Tchuktchi, of the extreme north-east. The opinion was recently held by many scholars that the agglutin ative dialects Accadian, Sumirian, &c. of the presumed founders of Mesopotamia!! culture and teachers of the Assyrian Semites (see BABYLONIA) belonged to the Ural-Altaic family, and specifically to its Finno-Hungarian branch ; but it is believed to be now generally abandoned. The mere possession of an agglutinative structure cannot be taken as proving anything in the way of relationship. 6. Dravidian or South-Tndian Family. This is an important Dravi- body of nearly and clearly related tongues, spoken by about clian. 50,000,000 people, doubtless representing the main population of all India at the time when the intrusive Aryan tribes broke in from the north-west, and still filling most of the southern peninsula, the Deccan, together with part of Ceylon. In an earlier article (see INDIA) the names of the dialects have been given, with indica tion of their locality and relative importance, and with some account of their leading features. They are languages of a high grade of structure, and of great power and euphony ; and the principal ones have enjoyed a long cultivation, founded on that of the Sanskrit. As they obviously have no Aryan affinities, the attempt has been made to connect them also with the Ural-Altaic or Turanian family, but altogether without success, although there is nothing in their style of structure that should make such connexion impossible. Not all the tribes that make iip the non-Aryan population of India speak Dravidian dialects. The Santals and certain other wild tribes appear to be of another lineage, and are now generally known as Kolarian. 7. Malay-Polynesian Family. The islands, greater and smaller, Malay lying off the south-eastern coast of Asia and those scattered over Poly- the Pacific, all the way from Madagascar to Easter Island, are nesian filled with their own peculiar families of languages, standing in no known relationship with those of the mainland. The principal one among them is the great Malay-Polynesian family. It falls into two principal divisions, Malayan and Polynesian. The Malayan includes, besides the Malay proper (see MALAYS), which occupies the Malaccan peninsula (yet doubtless not as original home of the division, but by immigration from the islands), the languages also of Sumatra, Java, Borneo, &c., of the Philippine Islands, of Formosa, and of Madagascar, together with the coasts of Celebes and other islands occupied in the interior by Papuans. The Polynesian division includes most of the tongues of the remaining scattered groups of
