352 EEPOKT— 1882. 



The Sea of Horkand extended to al-Mdmni, identified with Sumatra 

 not only by its position, but by its pi'oducts (snch as Fansiiri camphor, 

 elephants, brazil-wood, and cannibals), and by its great extent. The 

 compass of 800 parasangs ascribed to it corresponds roughly with the 

 estimates of Sumatra which we find in Marco Polo (' 2,000 miles or more '), 

 and in Barbosa (2,100 miles), and with the truth, which is about 2,300 

 miles. The navigators, crossing the Sea of Horkand with the western 

 monsoon, made land at the Islands of Lavja-, Lanlcha, or Lika-Bdlus, 

 where the naked inhabitants came off in their canoes, bringing ambergris 

 and coco-nuts for barter ; a description which, with the position, identifies 

 these islands with the Nicohars, Necuveram of Marco Polo, Ldka-vdram of 

 Rashid-uddin, and, I can hardly hesitate to say, the Harusae Islands of 

 Ptolemy. 



Beyond these, and not in the track usually followed, were the two 

 islands of the Anddmdn Sea, inhabited by dangerous and naked negro 

 cannibals. Still further out of the way in this dii-ection, and difBcult 

 of access, was a region of mountains containing mines of silver. The 

 landmark to reach these was a mountain called al-Khushndmi (' The 

 auspicious '). 



This 'Land of silver mines,' both position and this description iden- 

 tify with the Argijrc of Ptolemy. As no silver is known to exist in that 

 region (Arakan) it seems probable that the Arab indications to that effect 

 were adopted from the Ptolemaic charts. And this leads me to suggest 

 also that the Jibal Khush-ndml also was but a translation of the 'AynQov 

 caiixovor rijoroc, ov Islc of Good Fortune, in those maps, whilst I have 

 thought also that the name Anddmdn might have been adopted from a 

 transcript of the same name in Greek as ' 'Ay. ccu/jor. N.' ' 



At Koli, or Kfdis, I have said the Greek and Arab routes coincide. 

 For I take this K6U to be the Kalah of the Arabs, which was a month's 

 sail from Kaulam (Quilon) in Malabar, was a place dependant on the 

 Maharaja of Zabaj (i.e. Java, or the Great Islands), and near which were 

 the mountains producing tin. Ko-lo is also mentioned in the Chinese 

 History of the T'ang dynasty in terms which indicate its jDOsition some- 

 where in the region of Malacca.^ 



Kalah lay on the Sea of Shahthif (which we call the Straits of 

 Malacca), but was not very far from the entrance of tlie Sea of Kadranj, 

 a sea which embraced the Gulf of Siam ; therefore I presume that Kalah 

 was pretty far down the Malay Peninsula. It may, however, have been 

 Kadah, or Quedda, as we write it. For it was ten days' voyage from 

 Kalah to Tiydmah (written also Batnmali, Koi/)hnaJi, &c. — a variation 

 dependent on loose pointing chiefly), a place where they found supplies 

 of fresh water. And this I take it is Tiymnan (in charts corruptly 

 Timoan) on the eastern side of the Malay Peninsula. The island ' Thnon ' 

 is a point of note in Linschoten's ' Course from Malacca to Macau in 

 China ' (1597). ' Thereon,' he says, 'are two places where you find good 

 fresh water.' 



' BazaJwttr, and the ' Islc of Good Fortune,' may be taken as the Great and the Little 

 Andaman respective!}-. The Arab delation mentions, in an unconnected notice, an island 

 called Malhdn between Serendib and Kalah, i.e. between Cej'lon and the Malay Penin- 

 sula, ^^ hich was inhabited by black naked cannibals. This may bo another indication 

 ■of tlii Andaman group, and the name ma}' have been taken from I'tolemy's Maniolae, 

 which in his map occupy the position in question. 



- Sec Bretschneider on Chinese Botanical Works, Foocliow, 1870, p. 29. 



