The route by sea is pointed out in a clear and satisfactory manner, by some of the ancient authors, particularly the author of the Periplus of the Erythrean Sea.  In enumerating the exports from Nelkundah, he particularly mentions silk stuffs, and adds, that they were brought to this place from countries further to the east.  Nelkundah was a town in Malabar, about twelve miles up a small river, at the mouth of which was the port of Barake; at this port, the vessels of the ancients rode till their lading was brought down from Nelkundah.  This place seems to have been the centrical mart between the countries that lie to the east and west of Cape Comorin, or the hither and further peninsula of India; fleets sailed from it to Khruse, which there is every reason to believe was part of the peninsula of Malacca; and we have the authority of Ptolemy, that there was a commercial communication between it and the northern provinces of China.  But at a later period than the age of the Periplus, silk was brought by sea from China to Ceylon, and thence conveyed to Africa and Europe.  Cosmos, who lived in the sixth century, informs us, that the Tzenistae or Chinese, brought to Ceylon, silks, aloes, cloves, and sandal wood.  That his Tzenistsae, are the Chinese, there can be no doubt; for he mentions them as inhabiting a country producing silk, beyond which there is no country, for the ocean encircles it oh the east.  From this it is evident that the Tzenistae of this author, and the Seres of the ancients, are the same; and in specifying the imports into Ceylon, he mentions silk thread, as coming from countries farther to the east, particularly from the Chinese.  We thus see by what sea route silk was brought from China to those places with which the western nations had a communication; it was imported either into the peninsula of Malacca by sea, and thence by sea to Nelkundah, whence it was brought by a third voyage to the Red Sea; or it was brought directly from China to Ceylon, from which place there was a regular sea communication also with the Red Sea.

The author of the Periplus informs us, that raw as well as manufactured silk were conveyed by land through Bactria, to Baraguza or Guzerat, and by the Ganges to Limurike; according to this first route, the silks of China must have come the whole length of Tartary, from the great wall, into Bactria; from Bactria, they passed the mountains to the sources of the Indus, and by that river they were brought down to Patala, or Barbarike, in Scindi, and thence to Guzerat:  the line must have been nearly the same when silk was brought to the sources of the Ganges; at the mouth of this river, it was embarked for Limurike in Canara.  All the silk, therefore, that went by land to Bactria, passed down the Indus to Guzerat; all that deviated more to the east, and came by Thibet, passed down the Ganges to Bengal.