resemblances between these languages, a Hindu might feel disposed to draw the conclusion that Sanskrit is the source of all the other kindred tongues, instead of being derived together with them from an older language, the common parent of them all,-to obviate this erroneous inference, it is next shown that the whole grammatical character of Greek and Latin is that of independent languages; that in this respect they differ entirely from the Indian Prakrits (which have evidently resulted from the decomposition of Sanskrit), and that they even contain various forms which are older than those of the Sanskrit; while the greater part of their vocabulary is different. The same considerations apply, though not so strongly, to Zend. In Section iii. (pp. 267-278) the inference is drawn that affinity in language implies affinity in race;" and that, therefore, the ancestors of the Hindus must at one time have lived in the same country, as a part of one and the same community, with the forefathers of the Persians, Greeks, and Romans. In such a case as is here supposed, those branches of the original nation which separated earliest from the others, would in aftertimes exhibit the fewest points of resemblance in language and institutions to the rest, while those which remained. longest together would show in all respects the closest mutual affinities. In Section iv. (pp. 279286) it is argued that there is no objection arising from physiological considerations, i.e. from colour or

6

[In this second edition, this proposition is modified. I only affirm now that affinity in language affords some presumption of affinity in race.] [This assertion is in the 2nd edition changed into an inquiry whether there is any objection.]

bodily structure, to classing the Hindus among the Indo-European races. Section v. (pp. 287-300) exhibits the grounds which exist for supposing that the ancestors of the Indians and Iranians (or Persians) continued to form one community after the other kindred tribes had separated from them, and departed to distant regions. These grounds are, first, the closer affinity which subsists between Zend, the language of the ancient Persians, and Sanskrit (of which some illustrations are furnished); secondly, the fact that both nations in former times applied to themselves the appellation of Arya; and, thirdly, the nearer and more numerous coincidences which are discoverable between the early mythologies of the two peoples, of which some details are adduced. From this more intimate affinity between the Indians and Persians, independent as both are of each other in their origin and development (see also pp. 312-317), a strong confirmation is derived to the general conclusion (deduced mainly from language) of the common origin of all the nations called IndoEuropean. In Section vi. (pp. 301-306) the theory of Mr. Curzon, that India was the original country of the Indo-European races, from which they issued to conquer, occupy, and civilize the countries lying to the north-west, is stated, together with some of the arguments by which he supports it. The remarks of Mr. Elphinstone, who leaves it undecided whether the Hindus were autochthonous or immigrant, are also quoted. In Section vii. (pp. 306-322) I cite the opinions of Schlegel, Lassen, Benfey, Müller, Weber,

Spiegel, Renan, and Pictet, who concur in the conclusion that the cradle of the Indo-European race must be sought, not in India, but, as Schlegel, Lassen, and Pictet argue, in some central tract, from which the dif ferent branches of this great family could most easily have diffused themselves towards the widely-separated countries which they eventually occupied; a condition which would not be fulfilled by supposing a remote and southerly region, such as Hindustan, to be the point of departure. Some of these writers draw the same inference from the relation in which the Indo-Arians stood to the aboriginal tribes whom they encountered in India. In opposition to Mr. Curzon, who represents the language and religion of India as the sources from which those of all the other kindred races issued, Professor Spiegel maintains that the Iranian language and mythology, though owning a common origin with, are in their development perfectly independent of those of, the Indians. In the same section it is further urged that as neither the languages nor the mythology of the Greeks and Romans are derived from those of the IndoArians, there is no ground for supposing that the former nations emigrated from India at any period whatever. Section viii. (pp. 322-329) contains the few passages

I have been able to discover in the Indian authors which may be supposed to embody any reference (in no case, it must be confessed, other than a very obscure one) to the trans-Himalayan origin of their ancestors. The chief of these are the interesting paragraph of the Śatapatha-brahmana, which contains the legend of the

8

Deluge in the oldest form in which it occurs in any Sanskrit work,' and some texts relating to the northerly region of Uttara Kuru, the Ottorocorras of Ptolemy. In Section ix. (pp. 329-334) I have quoted, according to the versions of Spiegel and Haug, the first chapter of the Vendidad, which contains the oldest tradition of the Persians relative to Airyana-vaejo, the supposed primeval abode of their forefathers. Section x. (pp. 335341) discusses the route by which the Aryas immigrated into India. Schlegel and Lassen are of opinion that they must have penetrated into India from the west by the route of Kabul and across the Indus. Roth and Weber also regard the Panjab as the earliest seat of the Indo-Arians in Hindustan. In Section xi. (pp. 341– 357) I have endeavoured to show by quotations from the Vedas, that at the period when the hymns were composed, the Indians, though not unacquainted with the central provinces of northern India, were most familiar with the countries bordering on, or beyond, the Indus, and the north-western parts of Hindustan generally. From this fact, and from the testimony of later writers to their intercourse with tribes, apparently Arian in descent and language, residing in the Panjab and on the other side of the Indus, I derive a confirmation of the view that the Hindus entered India from the northwest.

In the Third Chapter (pp. 358-444) I have sought

7 [This passage has been omitted in the present edition for the reasons stated in note 96, p. 323.]

[A sentence referring to an opinion of Professor Benfey subsequently altered is here omitted.]

to adduce further arguments in support of the same conclusion, (1) from the distinction drawn by the authors of the Vedic hymns between their own kinsmen, the Aryas, and the tribes differing from them in complexion, customs, and religion, whom they designate as Dasyus; (2) from the accounts occurring in the Brahmanas and post-Vedic writings, of the gradual advance of the Aryas from the north-west of India to the east and south; and (3) from the well-established fact that the south-Indian languages are fundamentally different from the Sanskrit, and imply a non-Arian origin in the people by whom they were originally spoken. Section i. (pp. 359-369) contains a selection of passages from the Rigveda, in which the Aryas and the Dasyus are distinguished from one another, and reference is made to the enmity existing between the two. In most (or at least some) of these passages, it appears, human enemies and not demons must be intended under the appellation of Dasyus, as I infer both from the tenor of the texts themselves, and because in later writings, the Aitareyabrāhmaṇa, the Institutes of Manu, etc., this word is always applied to barbarous tribes. Section ii. (pp. 369-396) supplies a further collection of Vedic texts bearing upon the relations of the Āryas and Dasyus, and the characteristics of the latter as degraded, darkcomplexioned, irreligious, neglecters of sacrifice, etc, There are, indeed, other texts in which these Dasyus are regarded as demons, and this creates a difficulty. An attempt is made at the close of the section to explain, (1) from the original position of the Āryas,