pact Dra Their off shoots beyond sea (?) The com subdued by the higher civilisation of the Aryan race which vidians of pressed in among them, were never thus broken into fragments.1 Southern Their pure descendants consist, indeed, of small and scattered India.; tribes; but they have given their language to 28 millions of people in Southern India. A theory has been started that some of the islands in the distant Pacific Ocean were peopled either from the Dravidian settlements in India, or from an earlier common source. Bishop Caldwell points out that the aboriginal tribes in Southern and Western Australia use almost the same words for I, thou, he, we, you, etc., as the Dravidian fishermen on the Madras coast; and resemble in other ways the Madras hill tribes, as in the use of their national weapon, the boomerang. The civilisation and literature which the Dravidians developed in Southern India will be described in a later chapter on the Indian vernaculars. oblique form or genitive or other suffix. They agree with the Dravidian in having inclusive and exclusive forms for the plural of the first personal pronoun, in using a relative participle instead of a relative pronoun, in the position of the governing word, and in the possession of a true causal form of the verb. They have a dual, which the Dravidians have not, but they have no negative voice. Counting is by twenties, instead of by tens, as in the Dravidian. The Santáli verb, according to Skrefsrud, has 23 tenses, and for every tense two forms of the participle and a gerund.' - 1 Bishop Caldwell recognises twelve distinct Dravidian languages :(1) Tamil, (2) Malayalam, (3) Telugu, (4) Kanarese, (5) Tulu, (6) Kudugu, (7) Toda, (8) Kota, (9) Gond, (10) Kandh, (11) Uráon, (12) Rájmahal. In the Dravidian group,' writes Mr. Brandreth, there is a rational and an irrational gender of the nouns, which is distinguished in the plural of the nouns, and sometimes in the singular also, by affixes which appear to be fragmentary pronouns, by corresponding pronouns, and by the agree ment of the verb with the noun, the gender of the verb being expressed by the pronominal suffixes. To give an instance of verbal gender, we have in Tamil, from the root sey, "to do," seyd-an, "he (rational) did;" seyd-ál,"she (rational) did;" seyd-adu, "it (irrational) did;" seyd-ar, "they (the rationals) did;" seyd-a, "they (the irrationals) did ;" the full pronouns being avan, “he;" aval, “she;" adu, "it;" avar, "they; avei, "they." This distinction of gender, though it exists in most of the Dravidian languages, is not always carried out to the extent that it is in Tamil. In Telugu, Gond, and Kandh, it is preserved in the plural, but in the singular the feminine rational is merged in the irrational gender. In Gond, the gender is further marked by the noun in the genitive relation taking a different suffix, according to the number and gender of the noun on which it depends. In Uráon, the feminine rational is entirely merged in the irrational gender, with the exception of the pronoun, which preserves the distinction between rationals and irrationals in the plural; thus, as, "he," referring to a god or a man; ad, "she" or "it," referring to a woman or an irrational object; but ar, "they," applies to both men and women; abra, "they," to irrationals only. The rational gender, besides human beings, includes the celestial and infernal deities; and it is further LIST OF 142 NON-ARYAN TRIBES. 67 The following is a list of 142 of the principal non-Aryan List of languages and dialects, prepared by Mr. Brandreth for the Royal nonAryan Asiatic Society in 1877, and classified according to their gram- lanmatical structure. Mr. Robert Cust has also arranged them in guages. another convenient form, according to their geographical habitat. sub-divided, in some of the languages, but in the singular only, into masculine and feminine. The grammatical relations in the Dravidian are generally expressed by suffixes. Many nouns have an oblique form, which is a remarkable characteristic of the Dravidian group; still, with the majority of nouns, the post-positions are added directly to the nominative form. Other features of this group are-the frequent use of formatives to specialize the meaning of the root; the absence of relative pronouns and the use instead of a relative participle, which is usually formed from the ordinary participle by the same suffix as that which Dr. Caldwell considers as the oldest sign of the genitive relation; the adjective preceding the substantive; of two substantives, the determining preceding the determined; and the verb being the last member of the sentence. There is no true dual in the Dravidian languages. In the Dravidian languages there are two forms of the plural of the pronoun of the first person, one including, the other excluding, the person addressed. As regards the verbs, there is a negative voice, but no passive voice, and there is a causal form.' Bishop Caldwell's second edition of his great work, the Comparative Grammar of the Dravidian Languages (Trübner, 1875), forms in itself an epoch in that department of human knowledge. Mr. Beames' Comparative Grammar of the Modern Aryan Languages of India (Trübner, 1872) has laid the foundation for the accurate study of North Indian speech. Colonel Dalton's Ethnology of Bengal (Calcutta, 1872), and Sir George Campbell's Specimens of the Languages of India (Bengal Secretariat Press, 1874), have also shed new and valuable light on the questions involved. 1 Brackets refer to dialects that are very closely related; † to languages beyond the circle of the Indian languages. (See list above and on next page.) NON-ARYAN CENSUS OF INDIA. 69 lation We discern, therefore, long before the dawn of history, Recapitumasses of men moving uneasily over India, and violently the nonpushing in among still earlier tribes. They crossed the snows Aryan of the Himalayas, and plunged into the tropical forests in races. search of new homes. Of these ancient races, fragments now exist almost in exactly the same stage of human progress as they were described by Vedic poets more than 3000 years ago. Some are dying out, such as the Andaman islanders, among whom in 1869 only one family had as many as three children. Others are increasing like the Santáls, who have doubled themselves under British rule. But they all require special and anxious care in adapting our complex administration to their primitive condition and needs. Taken as a whole, and including certain half-Hinduized branches, they numbered 17,627,758 in 1872, then about equal to three-quarters of the population of England and Wales. But while the bolder or more isolated of the aboriginal races have thus kept themselves apart, by far the greater portion submitted in ancient times to the Aryan invaders, and now make up the mass of the Hindus. tion of The following table shows the distribution of the aboriginal Distributribes throughout British India in 1872. But many live in aborigines Native States, not included in this enumeration; and the in India Madras Census of 1872 did not distinguish aborigines from in 1872. low-caste Hindus. Their total number throughout all India (British and Feudatory) probably exceeded 20 millions in 1872. Aboriginal Tribes and Semi-Hinduized Aborigines in 1872. (Madras Presidency and the Feudatory States not included.) in 1881. As already stated, the Census of 1881 adopted a classification Aborigines which fails to clearly distinguish the aboriginal elements in the Indian population. In the North-Western Provinces, Oudh, Not and the Punjab, which returned an aggregate of nearly 1 millions of aboriginal or non-Aryan castes or tribes in 1872, no separate return of the aboriginal or non-Aryan element was made in 1881. It is merged by the enumerators in the returns separately of the Hindu low-castes. The same process has affected the returned. returns of other Provinces. In Madras, for example, 27 castes formerly included in the list of aboriginal tribes, were transferred to the Hindu section of the population. In Bengal, the Census officers explain that the non-registration of the aboriginal element is in some cases due to 'radical differences in the system upon which the castes, and especially the sub-divisions of castes, were classified in 1872 and in 1881.' In the NorthWestern Provinces and Oudh, the special officer states that his system of classification is not compatible with the modern doctrine which divides the population of India into Aryan and aboriginal.' No common data for 1872 and 1881. Hinduiz ing tendencies. Under these circumstances it would be misleading to attempt a comparison between the returns of the aboriginal or nonAryan population in 1872 and in 1881. On the one hand, there can be no doubt that the aboriginal castes and tribes are, in many parts of the country, tending towards Hinduism; and that many of them, as they rise in the scale of civilisation, lose their identity in the Hindu community. On the other hand, it is evident that the decreased returns of the aboriginal tribes and castes in 1881 are not entirely, or indeed chiefly, due to this process. It would be erroneous, therefore, to infer that the balance of 12 millions between the 17 millions of aborigines returned for British India in 1872 and the 4 millions nominally returned in 1881, had become Hindus. A Hinduizing process is going on both among the aboriginal low castes in Hindu Provinces, and among the aboriginal tribes who border on such Provinces. But the apparent disappearance of nearly 13 millions of aborigines between 1872 and 1881 is due, not so much to this Hinduizing process, as to differences in the system of classification and registration adopted by the Census officers. That the disappearance of the Indian aborigines is apparent and not real, can be proved. The birth-rate among some of the aboriginal races is unusually high; and, with exceptions, the aboriginal tribes and castes are numerically increasing, although they are partially merging their separate identity in the Hindu community. In Bengal and Assam, the aboriginal races are divided into |