FAU^A OF THE INLAND SEA OF SINGGORA. 97 



stake-nets and dip-nets. Tlie stake-net season terminated in 191G about 

 the beginning of Febniaiy, but in tlie week before it did so I was able 

 to make a large collection. Most of the species are marine and probably 

 swim in ami out of the mouth of the lake, but the estuarine forms noted 

 at Ijumpam also occur with others of like habits. Among the Rays, I 

 collected examples of Ithi/nchol>atis (iiicluding R. thouini, which is a 

 scarce form in the Bay of Bengal), Triji/on, Hijpolophus, Pteroplatea 

 and Aeiohatis. Several of the best of the Indian food fishes, for exam[)le 

 the Bekhti ( Lates cakarifer ), and the Hilsa, (Clupea ilisha), were 

 abundant. 



From a z )ological point of vidw, however, the most interesting 

 forms were certain minute (iobies dredged from the bottom of the lake 

 and, in particular, a peculiar little transparent fish which seems to 

 belong to the family Salangidae. This family, which is believed to 

 consist of degenerate relatives ot the Salmon idae, has not been found 

 hitherto west of China. Its members, which are more or less anadro- 

 mons, are remarkable (m- the transparency of their tissues, for their 

 elongate form and peculiar flattened narrow triangular heads All 

 the specimens found in the Tale Sap were unfortunately immature, but 

 I have little doubt that they represent post-larval stages in the life- 

 historj' of a species of ;S((?((7ia; or Some allied genus that occurs in the 

 Gulf of Siani. 



Molluscs. A remarkabl ! feature in the fauna of the outer lake 

 was the large number of dead bivalve shells obtained from the bottom 

 ill mv nets. A g'reat fljod, in which enormous volumes of fresh water 

 had been carried through the outer laice and out of its mouth into 

 the sea b}" a strong and steady wind, had occurred shortly before my 

 visit, and it is possible that this flood had killed some ot the molluscs. 

 We noted in tiie Chilka Lake' that in some species a larg-^ proiiortion 

 of the individuals were killed by the monsoon floods. Some of the beds 

 of dead shell;- in the Tale Sap are, however, probably of older and less 

 incidental origin. This is indicated by the fact that they include 

 hirge numbers of acorn-barnacle shells which must have been attached 

 to solid bodies of some kind, and that these shells were lying perfectly 



1- Annandale and Keuip, Men. Ind. Afui.. vol. V, p. 3;>7 (1916). 

 VOL. II, DEC. 1S16. 



