308 
MR. W. BATESOR ON SOME VARIATIONS OF CAEDIUM EDULE 
ridges with flat sides'. The beaks project very little from the general curve of the 
shell. 
The average ratio of leno’th to breadth in 30 shells the lengths of which were 
between IG and 21 mm. is 1: 0'725 ; that is to say, the average breadth of a shell 
20 mm. long is 14‘50 mm., as compared with 15‘98 mm. in the case of the shells of 
the highest level. 
Dreisseva 'polymorpha .—At one side of the lake on the level of the third terrace 
were found many shells of this form, which did not differ from those of the Aral Sea. 
The same is true of Hydrohia idvcB, winch was found in fair quantities on most of the 
terraces. 
Jaksi Klich. 
This is the largest, superficially, of the three dry lakes containing Cockles. Ifs 
length is about 10 miles, and its breadth 3 miles. It differs from Shumish Kul 
in being comparatively shallow. While the former must have been nearly 60 feet 
deep at the time of the separation from the Aral Sea, the basin of Jaksi Klich 
cannot have been more than 15 to 20 feet deep. There is not in it a distinct 
series of terraces, as at Shumish Kul, but the shells occur in two chief deposits, 
the one marking the original high level of the water, and the other forming a 
band round the salt which now Alls the bottom of the lake. Moreover, owing to 
the fact that the shells of the outermost deposit are almost all single valves, and not 
paired shells in situ, as at Shumish Kul, a good deal of mixing has become possible 
amongst them, which was, no doubt, facilitated by the shallowness of the lake ; as the 
banks ai-e so flat that at the time when the lake was low it may have happened that 
under a strong wind the water was driven upon the shore even as high as its original 
level. Hence it results that the upper deposit of shells at Jaksi Klich is more mixed 
in character than the deposits hitlierto described. I wall first desci’ibe the condition 
of the shells found at the bottom of the lake. They occur there in enormous numbers, 
being for the most part washed up into banks. A certain number of paired shells 
occurs between the ridges. Their texture is uniformly thin and papeiy, and they are 
very highly coloured, thus resembling the shells of the lower terraces of Shumish Kul, 
especially those of the sixth terrace. Their length is very great, and this feature is 
found in almost every individual shell. While they thus resemble in many respects 
the shells from the salter levels of Shumish Kul, they yet have several features peculiar 
to themselves, especially the enormously greater degree to wdrich they are elongated; 
also, though their colour resembles the Shumish Kul shells in being much brighter 
than that of ordinary Aral Sea Cockles, it has a character of its own, wdiich wmuld 
make it impossible to mistake a shell from either locality. 
As will be seen in the Tables, the average ratio of lenofth to breadth in 30 shells 
varying in length betwmen 30 mm. and 25‘5 mm. is 1 ; O’GGO ; and in 30 shells 
varying in length from 25'5 mm. to 19 mm. is 1 :0‘G82. It wall he seen, therefore, 
that the increased proportional length is greater in these shells than in any others that 
wmre ol^tained. 
