the Finnish or Ugrian family, with special affinities to the Ostiak. This connection seems to be radical, though remote, and established by particulars of primary importance.

Mr. Caldwell has arrived at his conclusions by a comparison of the Dravidian dialects-of which he has a thorough and accurate scientific and practical knowledge-with the grammars and vocabularies of the group in which he classes them. He acknowledges that a great diversity exists among the members of this group; so great, indeed, that, while the Indo-European idioms form only one family or genus, of which the ten families classified under that term are but species, in the Scythian family five or six authenticated genera have been enumerated, each of which includes as many species as are contained in the solitary Indo-European genus, besides twenty or thirty isolated languages, which have up to this time resisted every effort to classify them.

Notwithstanding this diversity, however, the generic characteristics of the Scythian group are very strongly marked, and incapable of being mistaken. The Ugrian and Turkish families, for instance, can be proved by their grammatical structure and vital spirit to be cognate, with as much certainty as the Gothic and the Sanskrit, or the Zend and the Greek.

I. The history of the Dravidian people is not unfavorable to the hypothesis of the Scythian relationship of their languages. There is sufficient evidence that the Drâvidas lived in the Indian peninsula long prior to the commencement of history, and before the Sanskrit-speaking race had made their way over the snow-capped mountains which separated their ancestral home from the plains of the Ganges, Nerbudda, and Cavery. The Drâvidas were doubtless the earliest inhabitants of India; or, at least, the first to enter from the northwest and cross the Indus. There is no evidence from Sanskrit authors-and they are our only authority on this point-that the Dravidians ever had any relations with the primitive Aryans but those of a peaceable and friendly character; and this could not have been true, had they followed that race into India. There is evidence that the Brahmans crossed the Vindhya mountains and entered the Dekhan and Southern India, not as conquerors, but as colonists; as priests and instructors, not as soldiers. The kings of the Pandiyas, Cholas, Calingas, and other Dravidians, appear to have been simply Dravidian chieftains, dignified by the new Brahman priests with Aryan titles. At the time when these events were taking place-some 500 years, perhaps, before the Christian erathe Dravidians were destitute of a written language, and unacquainted with the higher arts of life; but, from an examination of their language, it appears that they had acquired at least the elements of civilization. By a reference to the vocabulary of the early Tamilians, for instance, we gather, by our author's aid, the following items of information:

·

"They had 'kings,' who dwelt in 'fortified houses,' and ruled over small 'districts of country;' they were without books, but they had 'minstrels' who recited 'songs' at 'festivals;' they were without hereditary priests and idols, and appear to have had no idea of heaven or hell, of the soul or sin; but they acknowledged the existence of God, whom they styled ko or king,' a realis tic title which is unknown to orthodox Hinduism; they erected to his honor a temple, which they called ko-il, 'God's house.' They were acquainted with all the ordinary metals,' with the exception of tin and zinc; with the 'planets' which were ordinarily known to the ancients, excepting Mercury and Saturn. They had numerals up to a 'hundred,' some of them to a 'thousand;' but were ignorant of the higher denominations, a lakh and a crore; they had 'medicines,' but no medical science, and no doctors; 'hamlets' and 'towns,' but no cities; 'canoes,' 'boats,' and even 'ships'-i. e. small 'decked' coasting vesselsbut no foreign commerce; and no word expressive of the geographical idea of island, or continent. They were well acquainted with 'agriculture,' and delighted in 'war.' They understood cotton-weaving' and 'dyeing.' They had no acquaintance with painting, sculpture, architecture, astronomy, astrology, philosophy, or grammar. Their only words for the mind were 'diaphragm, the inner parts,' or 'interior;' they had a word for thought,' but no word distinct from this for memory, judgment, conscience, or will; to express the will, they would have been obliged to describe it as 'that which in the inner parts says, I am going to do so and so." But although there existed among them these elements of civilization previous to the arrival of the Brahmans, in intellectual, social, and political standing they were centuries behind this priestly race. They soon, however, rose in the social scale, and formed communities and states in the Dekhan rivalling those of the Aryans in the north.

เ

II. The absence of physiological evidence to the contrary. It is acknowledged that, while in some instances physiology has contributed much to the discovery of the affiliations of races, in the effort to prove the Scythian relationship of the Dravidians it renders no aid; but seems, so far as the study has been pursued, to be utterly at fault. The Dravidians might, on the ground of physical characteristics only, as well be classed with the Caucasians, or would readily admit of being affiliated with the Indo-Europeans; for no essential difference is observed between the heads and features of the Dravidians and those of the Brahmans; and, in fact, the Dravidian type of head will even bear to be directly compared with the European, with more definite marks of suppleness and subtlety in the former, and of straight-forward moral and mental energy in the latter.

It is not safe, however, in the presence of the strong lingual

evidences to be adduced, to draw any conclusion on this ground adverse to their Mongolian or Scythian origin; for a similar change has passed upon the features of the Mohammedans of India, who are all, without doubt, of Tatar-Mongolian extraction: with the exception of a somewhat greater breadth of face and head, and a more olive complexion, they do not differ physiolog ically from the Hindus, properly so called. A change appears to have passed over them, similar to that which is observed in the Osmanli Turks since they settled in Europe, which has transformed them from Tatars into Europeans.

It may farther be suggested in this connection, that possibly the distinctive Mongolian type, the absence of which is acknowledged in the Dravidians, has been developed in the course of time, since the period when the plains of India were first colon ized by the progenitors of their race.

III. Evidence derived from religious usages. In proving the origin and relationship of any people, the evidence gathered from their religious usages is always more satisfactory and reliable than that which is founded on physiological comparisons. The religions of the ancient Indo-European nations and those of the old Scythians of Upper Asia present many essential points of difference. In Shamanism-so is termed the superstition which prevails among the Ugrians of Siberia and elsewhere, and which was the religion of the whole Tatar race before Buddhism and Mohammedanism were disseminated among them-there was nothing which resembled the three prominent characteristics of the religion of the Indo-European family: viz., the doctrine of metempsychosis; the worship of the elements of nature, or of a pantheon of heroes and heroines; and the maintenance of a distinct and generally hereditary order of priests.

Shamanism acknowledges the existence of a Supreme Being, but no worship is rendered to him; nor are the objects of worship an inferior order of gods or heroes, but wicked and cruel spirits or demons. Any one who pleases may at any time officiate as priest, though ordinarily the father of the family, or the head-man of the hamlet or community, fills that office. Bloody sacrifices are offered with wild dances; the officiating priest or magician meanwhile exciting himself to frenzy, professes to have ascertained the mind of the propitiated demon, and, when the ceremonies are over, communicates it to those who consult him. Such is Shamanism, and the demonolatry practiced in India by the more primitive Dravidian tribes is not only similar to this, but the very same. The Brahmans by whom the Aryan civilization and superstition was grafted on the ruder Dravidian stock labored assiduously to extirpate their religion, and in this they were generally successful; yet is it still possible to discriminate between the doctrines and practices introduced by them and the

older religion of the people. Many vestiges of the primitive superstitions still remain, and in some districts they prevail extensively, especially among the Shânârs, and other rude and less Aryanized tribes, inhabiting the provinces in the extreme south of the peninsula. So far as yet appears, every religious usage of the Dravidians which is not of Brahmanical origin is either identical with Shamanism, or closely allied to it.

IV. Evidence furnished by the Behistun tablets. Before proceeding to the proofs derived from direct linguistic analysis, we notice an incidental evidence of the Scythian relationship of the Dravidian tongues. The famous inscriptions on the tablets at Behistun, in Beluchistan, which record the political autobiography of Darius Hystaspes, in the old Persian, Babylonian, Scythian, and Medo-Persian languages, have recently been translated. The translation of the Scythic portion enables us to compare the Dravidian idioms with a fully developed copious language of the Scythian family, as spoken in the fifth century B. C. The principal points of resemblance between the Dravidian dialects and the language of the tablets are: 1. The use of the cerebral class of consonants, t, d, n, which are indigenous to the Dravidian languages. 2. The use of the same consonant as a surd when initial and when doubled, and as a sonant when single and medial. 3. The employment in both of similar suffixes for the genitive and the dative cases of nouns, and the accusative of pronouns. 4. The use of a similar word for the numeral 'one' (the only numeral which occurs in letters in the tablets), and the uniform employment in both of the same suffix to express the ordinal numbers. 5. The pronoun of the second person singular is exactly the same in the tablets as in the Dravidian languages. The plural, unfortunately, does not occur. 6. The use of a relative participle. Perhaps this is the most remarkable characteristic of every unaltered dialect of the Scythian family. 7. The analogous etymons in the tablets are: nan, 'to say,' corresponding to the Dravidian an or en; uri, 'make known,' Dravidian urai; pori, 'to go,' Dravidian pô; ko, 'a king,' Dravidian ko. From the discovery of these analogies, Mr. Caldwell concludes that "the Dravidian race, though resident in India from a period long prior to the commencement of history, originated in the central tracts of Asia, the seed-plot of nations (and languages); and that from thence, after parting company with the rest of the Ugro-Turanian horde, and leaving a colony in Beluchistan, they entered India by way of the Indus."

V. Evidence from grammatical analysis.

1. The laws of sound. The phonetic laws which govern the Dravidian languages contribute to determine the question of their affiliation.

a. Vowels. The only point of resemblance noticed under this head is what is termed "the harmonic sequence of vowels," which appears in all the languages of the Scythian group, and in the phonetic systems of at least two of the Dravidian languages. The law of harmonic sequence is that a given vowel occurring in one syllable requires a vowel of the same class in the following syllables of the same word, and the vowels of such syllables are altered accordingly. In Telugu, the range of this law, although restricted to the two vowels i and u, appears to be identical with that of the Scythian law; u being changed into i, and i into u, according to the nature of the accompanying vowel. In some cases, the vowels of the appended particles are changed through the attraction of the roots to which they are suffixed; in other instances, the vowel of one of the suffixed particles draws that of the root and that of its other appendages also into harmony with itself: e. g. kalugu, 'to be able,' from which is formed with perfect regularity the aorist first pers. sing. kalugudu-nu; but the preterit first person is kaligi-ti-ni, where the change of the two final vowels of the root kalugu to kaligi, and of the personal termination nu to ni, is effected by the particle ti, which is the characteristic of the tense; for in the inflexion of Telugu words the most influential particles are those which indicate the time.

b. Consonants. One distinctive peculiarity of the Dravidian consonants is the convertibility of surds and sonants. There are four surd letters which are thus convertible; they are k, t, t, p: k is convertible into its related sonant g; tinto d; tinto d; and p into b. They are said to be convertible, because they are pronounced as surds at the beginning of words, and whenever they are doubled; and they are always pronounced as sonants when single and mediate. A sonant cannot commence a word, neither is a surd admissible in the middle except when doubled. In Tamil, and partly in Malayalam, one set of consonants serves for both purposes, and the change is made in the pronunciation alone. This peculiarity is not found in any of the Indo-European languages; but the resemblances which are found to exist between it and the laws of sound which prevail in some of the languages of the Scythian family amounts to identity. In the Finnish and Lappish there is a clearly marked distinction between surds and sonants: a sonant never commences a word in either tongue. The same remark has been already made of the Scythic version of the Behistun tablets.

The Tamil differs from the other Dravidian dialects in refusing to combine the surd lingual with the lingual nasal n, changing it in such a combination into its corresponding sonant d. This is in accordance with a general law of sound in that language, which is, that nasals will not combine with surds, but with sonants only.