Sec. i] MARCO POLO'S LOP AND HSÜAN-TSANG'S NA-FU-PO 319
At all three, earlier settlements are attested by ruins. But closer examination shows that Desert route
neither Vâsh-shahri nor Mirân can lay claim to represent the Venetian's ' town of Lop '. His btôzn of
account states that travellers proposing to cross the desert used to halt here for a week ' to refresh Lop'.
themselves and their cattle '. ' On quitting this city they enter the Desert.' Now from what has
been said above of the relative importance of the settlements actually extant at Vâsh-shahri and
Charkhlik, and of the respective size of the rivers upon which their existence depends, it follows
that Charkhlik, with its far more ample supply of water and much greater area of cultivation, could
not possibly have been an abandoned waste at a period when Vash-shahri was still an oasis,
as Marco's reference implies. The route from Charchan to Sha-chou or Tun-huang must always
have passed through Charkhlik, and if the ' town of Lop ' were to be placed at Vash-shahri, it
would be impossible to explain why Marco Polo should have described it as the last halting-
place of the caravans starting for the month's desert journey to Sha-chou.
At first sight his description would seem to fit Mirân far better ; for it lies two marches beyond 14lirân
Charkhlik on the direct route towards Tun-huang and on the edge of absolutely bare gravel desert.? before one d
But here again identification with Marco's ' town of Lop ' is precluded by the result of my Marco Polo.
excavations.8 These have brought to light conclusive archaeological evidence showing that the
ruins of Mirân were abandoned centuries before the Mongol period. Even if we assume that some
modest cultivation survived by the side of the Jahan-sai River, as it now exists at a distance from
the ancient site, this could not have sufficed for a town, nor furnished the resources which caravans
preparing there for a long desert journey would have regularly needed.
Thus we are forced to conclude that the place meant by Marco Polo's ' town of Lop ' must be Charkhlik
located at the Charkhlik oasis. With this conclusion the character and present appearance of the Marco
ruined circumvallation is in agreements How long after Ser Marco's passage the ' town ' and oasis passage.
were abandoned we do not know. It is very unlikely that they retained their importance after the middle of the fourteenth century, when, with the downfall of the Mongol dynasty, China reverted to the old policy of seclusion, and free intercourse and trade with Central Asia ceased. At the beginning of the fifteenth century, as we know from a statement of Shah Rukh's embassy, the southern route through the desert was unfrequented.'°
About the same period Lop is mentioned among the waste places where Vais Khan, a Moghul chief, was believed to have hunted wild camels.t' Mirza l;laidar, who records this about the middle of the sixteenth century, knew Lob merely as the name of a ruined town situated somewhere in the south-eastern part of the Tarim Basin. ' To the east and south of Kashghar and Khotan are deserts, which consist of nothing but heaps of shifting sands, impenetrable jungles, waste lands, and
7 See below, p. 346.
° Sir Henry Yule was guided by a just topographical sense when he conjectured (in a note contributed to Prejevalsky, From Ku ja to Lob-nor, p. 77, note 4) that Marco Polo's city of Lop might be located at what Prejevalsky briefly mentions as the 'traces of a third very large city near Lob-nor, at a place called merely Kunia-Shari, i.e. old town'. Prejevalsky's map shows that the Mirân ruins are meant, and on the strength of Sir Henry Yule's conjecture our maps have since continued to show them as ' Ruins of old Lob'. It was reserved for the spade to prove that the main ruins of Miran must have been deserted for many centuries before Marco Polo's caravan passed them.
° That Marco Polo's town of Lop corresponds to the
present Charkhlik was first recognized by M. Grenard, who, though he did not visit the ground himself, has correctly emphasized the importance of Charkhlik as the only possible site for a larger settlement at the crossways of several great routes ; cf. Mission Dutreu:l de Rhins, iii. pp. r 49 sq. There is much justice, too, in the remark that he adds : ' Il me paraît absolument certain que toutes ces anciennes villes entre Khotan et le Lob nor étaient de pauvres bourgades qui n'ont acquis quelque célébrité que parce qu'elles étaient pour les voyageurs des étapes obligées, clairsemées dans l'étendue déserte.'
10 Cf. Yule, Cathay, i. p. ccxi ; also cxxxiv ; and below, chap. xrv, sec. iii.
" Cf. Elias and Ross, Tärikh-i-Rashidi, p. 67.