Islam and Farmer-General of the Customs of the Persian Gulf, who resided at Kish. The contract price of the horses was fixed at 220 ducats a head, whilst the cost of' those lost at sea was borne by the contractor in Persia. In the time of Abubekr, the Salghur Atabeg of Hormuz, when that kingdom was in its glory, 10,000 horses yearly used to be shipped to India, bringing to the sellers a revenue of 3,500,000 pieces of gold !
" Two of the native chiefs of Maabar who contended for the throne were SINDARBANDI and PIREBANDI, the former the legitimate, the latter the illegitimate son of Gilishdiur Raja of Maabar" (probably KulesaDewar ; Von Hammer does not seem to see that this diur is the title Dewar which has just been specified), "a prince who had reigned prosperously for forty years without ever having been laid up by illness or attacked by a foe. He had named Pirebandi his successor, which so enraged Sindarbandi that he slew his father, and took forcible possession of Shahrmandi, where his enormous treasures were laid up. Pirebandi gathered an army to avenge his father's murder, and a battle took place beside a lake which the people of India called Talâji" (Tulà, a Tank? and perhaps the saine as the Celai (for Telai) of Odoric, p. 65.) " Eventually Pirebandi, aided by his cousin Bermal (Perumal?) was successful; whereon Sindarbandi fled to the court of Dehli, and sought . help from Alauddin against his brother." This led to the invasion of Kaffir.
This historian also speaks of Jamaluddin Abdarrahman Et-Thaibi as the Farmer-General and Keeper of the Marches of Maabar, apparently the same whom Rashid states to have succeeded to the Dewar in 1292. His son Surajjuddin, it is also stated, was plundered of all his wealth by the army of Kafur, upon which his son Nizamuddin betook himself to Dehli to make complaint, and obtained, with some partial restoration of property, the administration of the finances in Maabar, which had been held by his father and grandfather. (See Hammer Fur g-stall, Gesch. der Ilchane, ii, 51 seq., and 197 seq.) There are evident discrepancies between the accounts of Rashid and Wassaf, which it would be vain to attempt to reconcile without further knowledge. Nor do either their notices nor anything that I can gather from the works of Wilson and Taylor suffice to show to what dynasty belonged these princes of Maabar of' whom Polo and the Persian historians speak. The names of the chiefs, Sindarbandi (Sundara-Pandi), Parebandi (Vira-Pandi) Gilish (Kalesa), are all indeed such as occur repeatedly among the half-mythical lists of the Pandyan dynasty of Madura, but there seems some reason to believe that the chiefs in question may have been rather princes of Marawa, or of some family of adventurers.. The title Dewar, though not peculiar to the Setupatis has been specially affected by all the Marawas down to our own time, and Professor Wilson finds reason to believe that these were for a long time paramount over Madura, and for three reigns held the whole of that kingdom in their hands. (Catal. of hllackenzie Coll., i, 195; J. R. A. S., vol. iii, 165 and 223; Madras Journ., 1836, p. 35, seq.) The time indeed of this is left undetermined, except that it was before the rise of Vijayanagar in the fourteenth century.
The nearest approach in the Tamul Annals to an indication of the