284 ZOOIvOGY OF THE FAR EAST. 



moderatel}^ excavate anteriorly ; I have not seen any individual with this segment as 

 slender as in de Man's examples from the Nargi River in Flores (de Man, I.e., 1892, 

 p. 403, pi. XXV, fig. 3i(^). The fingers are usually a little longer than the palm. 



In the second peraeopods the carpus is from i| to i| times as long as the chela 

 and is from 4 to 4J times as long as its greatest breadth. 



The usual spines are present on the ischium, merus and carpus of the last three 

 peraeopods. The dactylus of the third pair generally bears from 8 to 10 spines, but 

 in specimens from the Patani River the number is higher, from 12 to 15. In the fifth 

 peraeopods the propodus is from 3I (Tinnevelly) to 4 J times (Patani R.) the length of 

 the dactylus. The latter segment usually bears from 32 to 39 spinules; but, as in 

 the case of the third pair, the number is higher in specimens from the Patani River, 

 varying from 45 to 55. 



The outer uropod is provided with from 8 to 11 movable spines. 



Large specimens reach a total length of about 38 mm. 



The size of the eggs is somewhat variable. In specimens from Calcutta, Portu- 

 guese India and the Patani River they are from 0-35 to 043 mm. in length and from 

 0-23 to 028 mr^ in breadth. In those from Travancore and Tinnevelly they are 

 slightly, but noticeably larger, from 0-50 to 0-52 mm. in length and from 032 to 0-33 

 mm. in breadth. The lowest of these determinations agrees with de Man's description, 

 in which the length is stated to be \ mm. Even between the extremes the variation 

 is small, but it is noteworthy that the specimens from Travancore and Tinnevelly 

 that possess the largest eggs were found in fresh water, whereas all the rest, including 

 those from which de Man drew up his description, were obtained in places within 

 the reach of tidal influence. 



Summarizing the foregoing observations it may be stated that material from five 

 distinct regions (four situated in the Indian Peninsula and one in Siam) shows 

 little signs of local variation. Three points only call for emphasis, — (i) in specimens 

 from the Tinnevelly district in S. India the average number of upper rostral teeth is 

 below normal, (ii) in specimens from Lower Siam the number of spines on the 

 dactyh of the last three legs is above normal, and (iii) specimens from water that is 

 brackish or subject to tidal influence have smaller eggs than those obtained in 

 fresh water. 



The colouration of Hving specimens is distinctive. The animal as a whole is 

 translucent with the rostrum, the lower surface of the last abdominal somite, the 

 distal two-thirds of the telson and frequently the tips of the uropods deeply pigmented. 

 The carapace is without markings, but there is a short transverse row of chromato- 

 phores on the third abdominal somite and a longitudinal line of similar chromato- 

 phores near the inferior margin of the first five somites. The depth of pigmentation 

 is variable. In extreme cases the whole of the rostrum, the antennules, the inner edge 

 of the antennal scale and the tail-fan are deeply pigmented and there is a broad lateral 

 longitudinal band on either side of the abdomen. 



In my experience C.gracilirostris is a scarce form, much less abundant than other 

 species of the same genus with which it is found associated. 



