AN ISLAND IN THE CHILKA LAKE. 271 



''The substance of shells from west coast estuaries is distinctly thinner than that 

 of those from the east coast, due possibly to a smaller lime content in the west coast 

 rivers. Here I should mention that much of the variation found in the form and 

 proportions of the umbones and hinge region of M. casta is due to the effect upon the 

 size and relative relationship of these parts, caused by the exceedingly variable rates 

 at which the shell increases in thickness under different conditions. Exceptionally 

 rapid growth in thickness tends to produce a humped and corbicular shape, while 

 rapid growth in length and breadth with slow deposits of lime salts give a compressed 

 form with flat umbonal region. Under certain conditions not now prevailing, but 

 which existed at the very recent geological period when the shell-pits of swamps 

 round the margins of the larger backwaters of the eastern coast were being formed, 

 the deposit of lime salts must have progressed at a greater rate than the most rapid 

 now existing, for in these subfossil deposits we got an immensely massive form, of 

 M. casta. This shell at first sight appears so different in hinge form and in general 

 shape from the type of M. casta that one does not hesitate to treat it as a different 

 species. Hence we find Preston describing it as a Cyrenid under the name of Corhi- 

 cula {Velorita) satparaensis. He saw some cause to modify this opinion later and in 

 Rec. Ind. Mus., XI, p. 300, he rightly assigned it to M. casta. In this I agree with 

 him, after a comparison of a large series of young individuals of the massive form as 

 opposed to those of the type. This gave conclusive evidence of identity, as the 

 extreme comparative solidity of the variety is rapidly lost as we descend in the 

 series till at last among small individuals of the type and of the variety, of f inch in 

 length, we attain practical identity and it becomes impossible to differentiate the one 

 from the other." 



Specimens from the island do not belong to the extreme phase of the race 

 but rather to a transitionary one. They are comparable to Hornell's fig. 21 on his 

 plate V and fig. 29 on his plate vi, but as a rule are a little more nearly symmetrical 

 bilaterally. A large valve is only 6-5 mm. thick in the middle of its disk. The 

 following table of measurement shows in a genaral way the difference in size 

 and proportions between these shells and a comparable series from the beds at the 

 head of Rambha Bay noticed by the late Dr. W. T. Blanford.' I will discuss these 

 differences further in comparing the deposits. 



Measurements of Shells o/Meretrix casta satparaensis. — 



Barkuda Rambha beds 



Height 257 30 31-5 33 34-5 38 5 33 39 40 41 43 445 



Breadth 267 32 32 35-5 37-5 41 32-5 367 43*5 42*5 42-5 44"5 



An unusuall}'- large shell, with the valves still adhering, picked out of the 

 soil at the village of Samulia on the mainland between Rambha and Barkuda, is 57 

 mm. high, 575 mm. broad and 43'5 mm. in diameter with the two valves together. 

 In this shell and in a large one from Rambha the thickness of the valve in the 

 centre is slightly over 8 mm. 



I Mem. Geol. Surv, Ind., I, p. 275 (i 



