270
tudinal extension of Asia; but the evidence of direct trade between Rome and China is remarkable.
The first part of the route was minutely described before our author’s time, in the Mansiones Parthica: of Isidorus of Charax Spasini.
This route of Maes the Macedonian followed very nearly the same direction as the Chinese Nan-lu, after leaving Bactra, crossing the Pamirs diagonally to Kashgar, on th ePci-lu, but then turning south- ward through Yarkand to Khotan, and in passing “Thagura” took a more southerly, and also a more direct route than the Nan-lu itself, which it joined half-way between Lop Nor and the Bulunzir (the “river of the Hiong-nu”); east of which all three routes were iden- tical as far as Singanfu.
(See map to face p. 500, Vol. I, of Richthofen’s China-,— Stieler’s Hand-Alias, maps 61-2; — Stanford, Atlas of the Chinese Empire, plates 12, 13, 19, 21 Lansdell, Chinese Central Asia, Vol. II; — Stein, op. cit., chap. v. and map.)
At Bactra this overland trade-route branched again, following westward through the Parthian highlands to the Euphrates, or southward to Bamian, the Cabul valley, the Khyber Pass and the Indus. From Taxilathe highway of the Maurya dynasty led through the Panjab to the capital at Palibothra, with a branch from Mathura southward to Ozene and the Deccan. The route down the Indus to its mouth was less important owing to the character of the tribes living on the lower reaches. This is indicated by the text, which says far more of the products carried by the overland route to Barygaza than of those coming to Barbaricum.
Yet a part of the Chinese trade was, apparently, localized at the mouth of the Indus. While the valuable silk cloth went to Barygaza, the yarn, or thread, went to Barbaricum, where it was exchanged for a product always more highly valued in China than in India — namely, frankincense; the white incense, or shehri luban, which Marco Polo still found in extensive use in China under the name of “milk per- fume.” This is not listed in the Periplus among the imports at other Indian ports, and evidently found its way up the Indus to Peucelaotis and Bactra, and thence to China. The silk yarn, in return, went to Arabia, where it was used in making the embroidered and silk-shot fabrics for which Arabia and Syria w T ere so famous in the Roman market.
Concerning the frankincense of the Deir-el-Bahri reliefs Mr. R. E. Drake-Brockman writes again from Bulhar, Sept. 18, 1910, that the cattle shown in those reliefs are not the humped cattle peculiar to
