^{page 495 note 1} According to de Morgan, (Les Premières Civilisations, p. 314)Google Scholar it ended some time between 1500 and 1200 b.c.

^{page 496 note 1} The attempts to find any trace of them in the Vendidad or Ṝig Veda are now, I think, generally abandoned. For the Vendidad see Darmesteter, S.B.E. The Zend Avesta, i, pp. 1–2Google Scholar; and for the Veda, Ṝig, Hopkins, , Religions of India, pp. 15, 30.Google Scholar

^{page 496 note 2} De Morgan, , op. cit., p. 266.Google Scholar

^{page 497 note 1} Armenians, Kurds, Cassites, Elamifces with Sumerians in Babylonia, and Arabs on the eastern shores of the Persian Gulf. The Sumerians may have been of Turki or of Ugro-Altaic stock, but they were strongly Semitized. The Elamites are often regarded as of the Turki race, but we have little to guide us except the language.

^{page 497 note 2} Haddon, , Wanderings of Peoples, p. 27Google Scholar; Races of Man, p. 60.Google Scholar

^{page 497 note 3} Of the Medes Herodotus says: καλέοντο δ πάλαι πρς πάντων Ἄριοι (Herod, , vii, 62)Google Scholar; and the Armenian historians continue to use the name. Cf. Rawlinson, , Herod, , i, p. 388.Google Scholar

^{page 497 note 4} It is convenient to have a generic name for the whole group of the Indo-European languages; and usage has sanctioned the term Aryan. But the Āryas were only one of the numerous Scythic tribes who used an Aryan speech; and had nothing to do with its origin. There are older forms than either Sanskrit or Avestic. The steppe dwellers carried their speech all over Europe, as well as Western Asia and India; but of how it arose, and where, we know nothing.

^{page 498 note 1} JRAS, 1909, pp. 1106 ff., and 1911, pp. 42–6.Google Scholar

^{page 498 note 2} Sergi assigns the skeletons found by Pumpelly at Anau to the Homo Mediterraneus, Pumpelly (Prehistoric Civilisations of Anau, ii).

^{page 498 note 3} Ripley, , The Races of Europe, pp. 448–51.Google Scholar “There can be no doubt of their racial affiliation with our Berbers, Greeks, Italians, and Spaniards. They are all members of the same race, at once the widest in its geographical extension, the most populous, and the most primitive of our three European types.”

^{page 499 note 1} It is sometimes diificult to distinguish in a photograph a British officer dressed in a Sikh uniform and turban from his native comrades.

^{page 499 note 2} JRAS. 1909, p. 1113.Google Scholar

^{page 499 note 3} They were fair, but not the fairest of the fair. In the Ṝig Veda, i, 100, 18, they are aided by “white-hued friends” against the Dasyu; i.e. they were brunets (Macdonell and Keith, , Vedic Glossary, i, 356, n. 6).Google Scholar

^{page 499 note 4} Their hair was not woolly, but straight; they were physically more nrobust, and their complexions not so dark (Strabo, , II, iii, 7, p. 103; xv, 24, p. 696).Google Scholar

^{page 499 note 5} τν δ’ νθρώπων οἱ μν μεσημβρινο τος Αἰθίοψίν ὅμοιοι κατ τν Χροιάν, κατ δ τν ψιν κα τν τρίχωσιν τος ἄλλοις (οὐδ γρ οὐλοτριχοσι δι τν ὑγρότητα το έρος), οἱ δ βόρειοι τος (Strabo, , xv, 13, p. 690).Google Scholar

^{page 499 note 6} Ctesias, Indica, par. 9 (Fragmenta Ctesiœ Cnidii, ed. Müller). He says: ὅτι οἱ Ἰνδο οὐχ ὑπ το λίου εἰσ μέλανες, λλ φύσει, Εναι γάρ φησιν ν αὐτος κα ἄνδρας κα γυνακας λευκοτάςτους πάντων, εἰ κα π’ ἔλαττον.

^{page 500 note 1} For the Indian accounts v. Macdonell, and Keith, , Vedic Index, vol. i, pp. 347 and 356 ff.Google Scholar, s.v. Dasyu and dāsa. For their beards and figure, Arriani, Indica, c. 16 and 17.Google Scholar “The Indians are in person slender and tall, and of much lighter weight than other men” (MacCrindle's trans., c. 17).

^{page 500 note 2} Arriani, Anabasis, v, 19, 1; Diod. Sic., xvii, 91.Google Scholar

^{page 500 note 3} I have already remarked that one-third of the Brahmans in the United Provinces are over 5 ft. 7 in.

^{page 500 note 4} Curtius, Q., de rebus gestis Alex. Mag., vii, 4, 6.Google Scholar “Quorum neminem adeo humilem esse ut huineri ejus non possent Macedonis militis verticem æquare.”

^{page 501 note 1} The Negrito element, dark-skinned with curly but not woolly hair, can be traced along the sea-coast of Southern Asia, and in regions not far from it, for a great distance. Negritos are found in the neighbourhood of Aden and along the southern coast of Arabia; they were probably the “Black-heads” of Babylonia; Houssay discovered traces of them in ancient Susiana; and they are found in the South of India and in the Malay Peninsula. The Veddas of Ceylon and the Toala of the Celebes belong to the same stock. To these many add the Melanesians and Australian blacks (Haddon, , Races of Man, pp. 12–13).Google Scholar Cf. Thurston, , Castes and Tribes of Southern India, i, Introduction, pp. xx–xxiii.Google Scholar

^{page 502 note 1} Thurston has treated of the Dravidians very fully in his Introduction to The Castes and Tribes of Southern India, vol. i.Google Scholar

^{page 503 note 1} For these Mundā-speaking tribes v. Census of India, 1911, pp. 322–7Google Scholar; SirGrierson, G., JRAS., 1907, pp. 187 ff. and 743Google Scholar; Haddon, , Races of Man, p. 64.Google Scholar Cf. Thurston, , op. cit., vol. i, p. xxiii.Google Scholar For the ease with which small tribes change their language v. Gait in Census of India, 1911, p. 328.Google Scholar

^{page 504 note 1} Strabo, , xv, c. 37, p. 702.Google Scholar For the physical geography of the Panjab v. Ibbetson, 's Outlines of Panjāb EthnographyGoogle Scholar, pars. 6–24.

^{page 504 note 2} This interpretation is modern; the word was originally Sarhind, but the modern version admirably connotes the fact.

^{page 505 note 1} An almost imperceptible watershed is generally quite sufficient to separate one herd of black buck from another herd. They seldom cross it except when they are driven.

^{page 505 note 2} Iyengar, , Age of the Mantras, p. 15Google Scholar, says that when the Āryas reached India “the country was not a barren waste, but its rich valleys were filled with a teeming population, speaking dialects of the Dravidian and Mundā, languages. They were not primitive tribes, but tilled the ground, and raised crops of various kinds … they worked in metals … they traded with foreign countries”. But Iyengar confounds times early and late. Macdonell, and Keith, , Vedic Index, i, p. 358Google Scholar, say more truly: “The wealth of the DĀsas was no doubt considerable, but in civilization there is no reason to suppose that they were ever equal to the invaders.”

^{page 507 note 1} My chief authorities for this chapter are Macdonell, and Keith, 's Vedic IndexGoogle Scholar; Iyengar, Srinivas, Life in Ancient IndiaGoogle Scholar; and Chanda, R., The Indo-Aryan Race.Google Scholar

^{page 507 note 2} Hopkins, , Religions of India, p. 30.Google Scholar

^{page 507 note 3} Ibid., c. 2; and Iyengar, P. T. Srinivas, Life in Ancient India.Google Scholar

^{page 508 note 1} There is no tin in India, and the swords and other implements found in the Ganges Valley are of copper. V. Smith, I.A. xxxiv, 1905, pp. 229 ff. Dr. Smith says: “This essay will be primarily devoted to proving that in the greater part of Northern India a copper age intervened between the neolithic and the iron age; and secondarily, to proving that India had no bronze age.” Cf. Macdonell, and Keith, , Vedic Index, s.v. loha; ii, p. 234.Google Scholar

^{page 508 note 2} Varuna is the only one of the Vedic gods who bears a really ethical character; he flees from sin, and punishes sinners. But this side of his character is a comparatively late development (v. Hopkins, , op. cit., pp. 61 ff).Google Scholar

^{page 509 note 1} The Kathaians, Malli, Sibi, Oxydrakæ were all “kingless”. For Kōsala-Vidēha, v. Davids, Rhys, Buddhist India, pp. 17 ff.Google Scholar

^{page 510 note 1} Macdonell and Keith's Vedic Index gives the names of nearly seventy tribes which are mentioned in Vedic literature. Iyengar, , op. cit., p. 8Google Scholar, says the “Mantras” mention about forty.

^{page 510 note 2} Thus the Bharatas had their own Agni, and a special Τύχη in a goddess Bhāratī. Agni bore different names among the Eastern and the Western Āryas. Cf. Iyengar, , op. cit., p. 127Google Scholar; Macdonell, and Keith, , op. cit., ii, 97.Google Scholar

^{page 511 note 1} Macdonell, and Keith, , Vedic Index, i, pp. 466 ff.Google Scholar, s.v. Pancajanāh (Hopkins, , op. cit., p. 26).Google Scholar The five peoples are sometimes said to be the Anus, Druhyus, Yadus, Turvasas, and Pūrus.

^{page 511 note 2} But the meaning is obscure. The Turvasas appear later to have been merged in the Pañchālas (ibid., i, p. 116).

^{page 512 note 1} For the civilization of the aborigines v. Iyengar, , op. cit., p. 13Google Scholar, and Vedic Index, i, 356Google Scholar, s.v. Dāsa.

^{page 513 note 1} For the Tṛitsus v. Macdonell, and Keith, , Vedic Index, i, pp. 320–3.Google Scholar

^{page 513 note 2} For the Bharatas, Turvasas, Pūrus, etc., v. Macdonell, and Keith, , Vedic Index, i, pp. 313 ff.Google Scholar; ii, pp. 11–13, 94–7. The exploits of Sudās and his father Divōdasa are narrated in the iii, vi, and vii Maṇḍalas of the Ṝig Veda. For the Kurus, v. op. cit., i, pp. 165–9.Google Scholar

^{page 514 note 1} Macdonell, and Keith, , Vedic Index, i, p. 468Google Scholar, s.v. Pañcāla.

^{page 514 note 2} Ibid., i, pp. 165–9; ii, pp. 125–6.

^{page 515 note 1} For Kōsala-Vidēha v. Davids, Rhys, Buddhist India, pp. 25 ff.Google Scholar, and Macdonell, and Keith, , Vedic Index, i, pp. 153–5; ii, pp. 298 ff.Google Scholar

^{page 517 note 1} Strabo, , xv, c. 28, p. 698.Google Scholar

^{page 518 note 1} Macdonell, and Keith, , Vedic Index, ii, 126Google Scholar; Laws of Manu (SBE. xxv), x, 44.Google Scholar Manu, however, has chiefly foreigners in view—Yavanas, Pahlavas, Śakas, etc. The Mahābhārata calls the tribes on the Indus Mlechchas or barbarians.

^{page 518 note 2} Herod, , in, 98–101.Google Scholar

^{page 519 note 1} Strabo, , xv, 54, p. 710.Google Scholar

^{page 519 note 2} Herod, , iii, 102.Google Scholar

^{page 520 note 1} Macdonell, and Keith, , Vedic Index, ii, 125–6.Google Scholar The country of the Kuru-Pañchālas is the country “where the Brāhmanas and the later Saṃhitas were produced”. Cf. ibid., i, 154, 468.

^{page 520 note 2} Davids, Rhys, Buddhist India, p. 3.Google Scholar

^{page 520 note 3} Macdonell, and Keith, , Vedic Index, i, 168.Google Scholar

^{page 521 note 1} Ibid., i, 153–5, s.v. Kās¯.

^{page 521 note 2} Ibid., i, 165–9; ii, 94–7, 273.

^{page 521 note 3} Pargiter, , JRAS., 1908, 334.Google Scholar

^{page 521 note 4} Macdonell, and Keith, , Vedic Index, ii, 126.Google Scholar

^{page 522 note 1} Ibid., i, 153–5; ii, 116–8, 298; Davids, Rhys, Buddhist India, pp. 60 ff.Google Scholar

^{page 522 note 2} Macdonell, and Keith, , Vedic Index, i, 154–5.Google Scholar

^{page 523 note 1} This appears to me the most probable explanation of the historical germ of the Rāmāyaṇa; but everyone is at liberty to frame his own theory. The identification of Lankā with Ceylon is late, and possibly due to the Buddhists. In the original poem, says Jacobi, Lankā was not an island but a town, and the Indian astronomers placed it on the equator, where it was intersected by the meridian of Ujjayini.

^{page 526 note 1} Strabo, , xv, 41, p. 702Google Scholar; Pliny, , HN., vi, 19, 66.Google Scholar

^{page 527 note 1} Sir G. A. Grierson tells me that he lays even more stress upon the phonetic differences than upon the grammatical.

^{page 528 note 1} Vedic Index, ii, p. 126Google Scholar, n. 14. The theory in question is set forth in the chapters on Ethnology and Language in the Census of India Report, 1901Google Scholar, and summarized in the Imperial Gazetteer of India, the Indian Empire, vol. i, pp. 303, 358Google Scholar; v. also SirRisley, H., The People of India, p. 55.Google Scholar The theory had its origin in the linguistic crux, and was for a long time confined to the philologists, who posited a separate section of the Āryas speaking a Vedio speech different in grammar and phonetics from the rest. Hillebrandt, in his Vedische Mythologie, vol. i, pp. 94–116Google Scholar, gave these speculations definite shape. The deeds of Sudās and his father Divodāsa are celebrated in the 6th and 7th Mandalas of the R. V. Hillebrandt holds that the Divodāsa of the 6th Mandala is a wholly different person from the Divodāsa of the 7th. The Divodāsa of tlie 6th Mandala combats the Panis, Dāsas, Pārāvatas, and others; and Hillebrandt equates these with the Scythic nomads, the Parni or Assarni, and the Dahæ; the Pārāvatas are Ptolemy's Παρουται, and the Sarasvatī of this Mandala is the Hara-quaiti, the Arachotvis or Helmund. This earlier Divodāsa lived on the Helmund in Arachosia; the Pūrus were his neighbours; and at a much later period his descendants led the Bharatas to India, whither the Pūrus also migrated. He concludes that there was a double invasion, an earlier and a later, and from two different quarters. Sir H. Risley improved upon this. He placed the main body of the Āryas in Arachosia and Seistan, while a second body of adventurers, travelling without wives, entered India by way of Uilgit and Chitral. I am not aware that any ethnologist has followed Sir H. Risley in this matter. Any invaders of India coming from Arachosia or Seistan would come by way of Kandahar and the Bolan, and would naturally spread in the first instance over the Lower Indus, whereas the Āryas of whom we have knowledge were on the Kabul River and in the North-West Panjab. All these speculations appear to me baseless. I am not a philologist, and I can offer no solution of the linguistic puzzle; but surely that does not require such violent hypotheses. As Sir G. Grierson has said with regard to it: “It is immaterial whether we are to look upon the state of affairs as two invasions, or as the earlier and later invasions of a series extending over a long period of time.” The Bharatas were late-comers; they regarded the speech of the Pūrus as barbarous; and philologists may have some right in claiming them as the missing element, provided that there is no other solution of the question. For a fuller discussion of the subject v. SirGrierson, G. A.'s and ProfessorKeith, Berriedale's papers in the J.R.A.S. for 1908 and 1917.Google Scholar