Sec. iv] ENDERE AND THE ' OLD TU-HUO-LO COUNTRY' 287
brought into subjection in Bactria .2 It is not likely that they will be definitely settled until further researches, perhaps on Bactrian soil itself, have furnished additional materials. Still less do I feel called upon here to investigate the location at different periods of those Tukhdra who have figured in Sanskrit literature since the Mahabharata, and whose name is certainly identical with that of the classical Tochari in Strabo and Ptolemy and the Tu-huo-lo of the Chinese records. But in the course of the discussions on this subject Hsiian-tsang's mention of the ruins of the ' old Tu-huo-lo country' has been repeatedly treated as if it afforded conclusive proof that the early seats of the Tochari were in this particular locality before their conquest of Bactria.3 It is this assumption which must be examined here in the light of the newly acquired archaeological evidence.
This shows quite clearly that the town walls, etc., which Hsiian-tsang saw in ruins about A.D. 645 at the deserted settlement must have been occupied down to the third or fourth century of our era. It follows that the abandonment of the site could have nothing to do with the supposed emigration of the Tochari from this tract to Bactria some eight centuries before the time of Hsiian-tsang's visit. In view of this it must appear very doubtful whether it is necessary or even justifiable to put upon Hsiian-tsang's brief reference to the ' old Tu-huo-lo country' that interpretation which has been generally presumed in the discussions reviewed in the preceding note, viz. that this was the country
2 These questions have been discussed more recently and with critical care by Prof. Marquart, Érân ahr, pp. 200 sqq. ; Franke, Tiirkvölker, pp. 24 sqq. For a review of earlier opinions and for some observations which still deserve attention, reference is useful to the paper of M. VasconcellosAbreu, De l'origine probable des Toukhares et de leurs migrations à travers l'Asie, in Le Museron (1883), ü. pp. 165-88.
3 As far as I can trace the matter in the literature at present accessible to me, Sir Henry Yule appears to have been the first to notice the possible bearing of this passage in Hsiian-tsang's record upon the question of the ethnic origin of the Tochari or Tokhara. In his masterly Notes on Hwen Thsang's Account of the Principalities of Tokhâristdn (1892), J.R.A.S., N. S., vi. p. 95, he cautiously refers to Hwen Thsang, on his journey homewards, using the same name of Tukhdra in connexion (as it would seem) with the original seats of the Yuéchi beside the Gobi Desert'.
This early location of the Tukhâra between Niya and Charchan previous to their appearance in the Oxus region became an important argument for Baron von Richthofen; in an elaborate note of his great work, China( I 877), i. pp. 439 sqq., he endeavoured to demonstrate that the designations of Tukhâra and Yüeh-chih applied to one and the same people which had their original seats east of Khotan. M. VasconcellosAbreu, to whom credit is due for having called attention to certain serious difficulties about this theory of the great geographer, appears to have had also critical misgivings as to the probative force of the argument derived from Hsüantsang's brief mention of the ruined settlement and as to the exact meaning of the designation, ' old Tu-huo-lo country', which he gives to it ; cf. his paper quoted above, Illusion, 1883, pp. 167 sqq., especially pp. 175, 185.
Prof. Marquart by a series of ingenious observations endeavours to prove that the Tochari (Tukhâra, Tu-huo-lo) who took Bactria from the last Greek rulers on the Oxus were identical with the Ta-hsia whose territory, as we see from the Former Han Annals, must subsequently have been
occupied by the Yüeh-chih some time after 126 B. C. ; cf. L'rânkahr, pp. zoo sqq. He assumes that these TochariTa-hsia had emigrated into Bactria in the latter half of the second century B. c. from the tract where Hsiian-tsang eight centuries later still found the ruins of the old Tu-huo-lo kingdom ' abandoned to the desert. Prof. Marquart lays special stress on Hsiian-tsang describing the deserted settlements he saw as ' towns '. He takes this as a proof of the high civilization which he assumes to have been already attained by the Tochari before their movement westwards to the Oxus ; cf. Érânsahr, p. 2o7.
Prof. Franke, who has last discussed these questions with much thoroughness and the special qualifications of a Sinologist, does not accept the identification of the Tochari and Ta-hsia, but assumes that the Tochari, originally seated in Hsiian-tsang's old Tu-huo-lo country ', attached themselves to the Ytieh-chih when the latter fleeing before the Hsiung-nu or Huns about 170 B. c. passed from their old seats in the extreme north-west of Kan-su to the T'ien-shan region through the Taklamakân; cf. Tiirkvölker, pp. 28 sqq. He believes the Tochari to have been nomadic at the time.
It seems to me impossible to reconcile such a nomadic existence of a large tribe on the desert edge with the physical conditions which during historical times must always have prevailed there more or less as at present. In the same way geographical considerations completely preclude the idea that the migration towards the Tien-shan of a large nomadic tribe, such as Chang-ch'ien's record shows the great Yüehchih to have been before their defeat by the Huns, could have taken place through the Taklamakan or along the southern edge of it. That the Tarim Basin with its barren wastes of sand or gravel, broken only by a narrow fringe of cultivated oases, was throughout historical times a region utterly unsuited to nomadic migrations is a geographical fact which deserves to be reckoned with in historical speculations more than hitherto has been the case.