4 JOURNAL R. A. S. CEYLON. - [EXTRA ffO. 



All the inhabitants of these islands are Miisalmans, pious 

 and honest people. They are divided into regions or zones, each 

 of which is ruled by a governor called Cordoiiiy. Among these 

 zones the following are distinguished: (1) the zone of Pali- 

 pour; (2) Cannalous; (3) Mahal, the province after which all the 

 islands are called, and at which their sovereigns reside ; (4) Teld- 

 dib ; (5) Caraidou; (6) Te'im ; (7) Teledommety ; (8) Heledommety „ 

 the name differing from the preceding only by having its 

 first letter an h: (9) Bere'idou ; (10) Candacal ; (11) Molouc; (12) 

 Souwe'id. The last is the most distant of all.* All the Mai- 

 dive islands are destitute of grain, except that in the province 



* The French editors identify these names with those given by Pyrard as 

 follows: — " Pdlipour — ' Padypolo' : Mahal — 'Male': Caraidou — 1 Caridou': Tele- 

 dommety — 1 Tilla dou niatis' : Heledommety — ' Milla done mad cue' : Bere'idou 

 — ' Poulisdous' : Molouc — i Molucque' : Souweid — ' Souadou.' The Cannalous 

 of Ibn Batuta, they acid,- is perhaps the ' Collomadous' of Pyrard, or, as 

 Horsburgh writes it, 4 Colomandous. 



["The majority of the above 4 are easily recognisable,' but five (viz., 

 Cannalous, Telddib, Te'im, Heledommety, and Candacal) present more diffi- 

 culty,— -some, indeed, being shaped, more Arabico, out of all but unsatisfactory 

 conjectural recognition. 



"As regards Heledommety, the French translators would appear to be at 

 fault. It may perhaps with more reason be taken as equivalent to Hela- 

 du(m)-mati (S. liela, 1 white,' or even sela, = Jav. sela, 1 rock' -L diiva, ' island'-f- 

 matu-plta, 'above') passing by contraction into Had-dummatL Addu may 

 similarly be a contracted form of Hela-du or Hulu-du — the name of one of the 

 islands in that Atol. 



" If it be open to question the orthography of the MS. and, supposing an 

 error of the copyist, read Nelddib for Telddib — t and n are not unlike in 

 Arabic—it can be at once fitted to Nilandu Atol. Even accepting the received 

 form as correct, the identification may possibly be not considered too far- 

 fetched. Cf. Tojaree = ? Nausdri ; AccaneeanA Kaluftee islands' (Lakkadive 

 group) = Aucutta, Kalpeni. Adm. Chart. (Col. Yule in Indian Antiquary 



Vol. III., pp. 212-4 on " Names in the Tohfat-al Majahidin.") ' 



" To attempt to twist the rest into probable coincidence with the modern 

 names of the remaining Atols seems hopeless. But the identification of Carai- 

 dou with Caridou (Kaharidu) — if accepted — suggests a likely clue to their 

 origin. Admitting that this island — now-a-days of comparative insignificance 



