^^^''•1 Faiiiui of tlic riiilka Lake ■ Crustacea Decapoda. 307 



Rambha to Nalbano and also occurred off Barnikuda, in Semanaddi, near Satpara 

 and m the vicmity of Barhampur I. It was invariably obtained on a bottom of 

 soft mud m water from 4 to 12 ft. in depth. Unlike A. crassimanus, it was never 

 seen under stones at the margin of the lake. Specimens were found at all times of 

 the year and the species is evidently able to tolerate changes in specific gravity 

 varying irom 1000 to 10265. Ovigerous females were found in November and 

 March. In the former of these months they occurred in water of very .slight salinity 

 whereas in the latter months they were obtained in water as salt as that of the sea 

 in the neighbourhood of the lake-mouth. 



The type specimens bear the nos. 9020-2/10 in the Museum register. 



Family ATYIDAE. 

 Genus CARIDINA, Milne-Edwards. 



1905. Caridina, Bouvier, Bull. sci. France Belgique, XXXIX, p. 67. 

 1913. Caridina, Bouvier, Trans. Linn. Soc, Zool. (2), XV, p. 447. 



Two species of this genus are commonly found in the Chilka I,ake among weeds. 

 Both occur abundantly in the Gangetic delta in brackish water. 



Caridina nilotica (Roux). 

 var. bengalensis, de Man. 



1908. Caridina nilotica, var. bengah-nsis, de Man, Rec. hid. Mn?., II, p. 265, pi. xx, figs. 6, 6a 6b. 



For the form of Caridina nilotica which occurs in the Chilka Lake I have 

 employed the varietal name given by de Man to the race found in the Gangetic delta. 



There are numerous series of Caridina nilotica in the Indian Museum obtained 

 at various points on the coasts of the Indian peninsula. \Aniere precise data are 

 available, it appears that these specimens were, with very few exceptions, obtained in 

 brackish water or in water that, though fresh at the time of their capture, is occa- 

 sionally subject to tidal influence. 



These series of individuals all agree in possessing the characters of the var. 

 bengalensis except that they show considerable variation in the dentition of the 

 rostrum. Kven in examples from the Gangetic delta the range of variation is much 

 greater than is apparent from de Man's account, the teeth forming the basal crest on 

 the upper margin varying in number from 15 to 30 and those on the lower margin 

 from II to 22. On examining long series from different places it is evident that 

 local distinctions exist in the number of rostral teeth ; but these distinctions are so 

 slight that it is only by taking the average of a large number of individuals that 

 they can be detected and they are, of course, far too trivial to justif)' nominal 

 recognition. 



It is, however, interesting to note that the Chilka Lake examples agree more 

 nearly with those from S. India than with those from the Gangetic delta, a fact 

 which is shown in the following table: — 



