CiiAr. II.] INDIAN, ARABLVN, PEESIAN AUTIIOEITIES. 



591 



Ceylon, and in tlie course of the narrative, Garsharsp 

 and his fleet reach their destination at Kalah, and there 

 achieve a victory over the " Sliah of Serendib." ^ 



It must be observed, that one form of tlie Arabic 

 letter k is sounded like G, so that Kalah would be pro- 

 nounced Gala} The identity, however, is estabhshed 

 not merely by similarity of sound, but by the concurrent 

 testimony of Cosmas and the Arabian geographers ^, as 

 to the nature and extent of the intercourse between 

 China and Persia, statements which are intelligible if 

 referred to that particular point, but inapplicable to any 

 other. 



Coupled with these considerations, however, the iden- 

 tity of name is not without its significance. It was 

 the habit of the Singhalese to apply to a district the 

 name of the principal place within it ; thus Lanka, 

 which m the epic of the Hindus was originally the ca- 

 pital and castle of Eavana, was afterwards apphed to 

 the island in general ; and according to the Mahawanso^ 

 Tambapani. the point of the coast where Wijayo landed, 

 came to designate first the Avooded country that sur- 

 rounded it, and eventually the whole area of Ceylon.'^ 

 In the same manner Galla served to describe not only 

 the harbour of that name, but the district north and 

 east of it to the extent of 600 square miles, and De 

 Barros, De Couto, and Eibeyro, the chroniclers of the 

 Portuguese in Ceylon, record it as a tradition of the 

 island, that the inhabitants of that region had acquu^ed 



^ Ottseley's Travels, vol. i. p. 48. 



2 Kalah may possibly be identical 

 with the Singhalese word f/ala, which 

 means an " enclosnre," and the deeply 

 bayed harbour of Galle woidd serve 

 to jvistify the name. Galla signifies 

 a rock, and this derivation would be 

 equally sustained by the natural fea- 

 tures of the place, and dangennis coral 

 reefs which obstruct the entrance to 

 the poi-t. 



3 DuLATTRiEE, in the Journal 



Asiatkine for Sept. 184G, vol. xlix. 

 p. 209, has brought together the 

 authorities of Aboulfeda, Kazwini, 

 and others, to show that Kalah nuist 

 be situated in Ceylon, and he has 

 combated the conjecture of M. Alfred 

 Maury that it may be identical with 

 Kedah in tlie Malay Peninsula. — 

 Reinaud, Relation, cS'r. Disc, pp. 

 xli. — Ixxxiv., Introcl. Aboulfeda, p. 

 ccxviii. 



■* MaJwuHinso, ch. vii. p. 50. 



