APPARENTLY CORRELATED TO THE CONDPnONS OF ETFF, 
almost all due to a change in the proportion wljich the greatest length bears to 
the greatest breadth at right angles to it. The measurements, owing to the 
irregularities of the shell, were made accurate to half a millimetre, and I believe that 
any difference due to variation in the selection of the exactly comparable morphological 
point on the ventral edge of the shell would be found to be within this limit of error. 
1 am also aware that concholoo'ists use the term leimth for the shorter of these 
two measurements ; but, as this appears confusing to the general reader, it seems better 
in a paper of more general biological interest to use the terms in their ordinary sense. 
In comparing these shells of the upper terraces with those of the lower, it will 
be found that the greatest length is^xeater in proportion to the greatest width than it 
is in those shells wdiich have been exposed to salter water. 
I have made tables which are intended to show this change in the proportions of the 
shells in a tabular form. The Tables bring out three points :—■ 
{a.) That the change in proportions does not occur in all the shells, nor to an equal 
degree in tliose in whicli it is found. Thereby it differs from the changes which occur 
in the texture and colouring. A few shells may be found in any terrace at Shumish 
Kul which do not differ materially in shape from normal shells. In the case of Jaksi 
Klich, however, almost all the shells are affected. 
(6.) The second point noticeable in the occurrence of this variation is that it is far 
more marked in shells of greater absolute size (that is, presumably, of greater 
age) than in smaller and younger ones. This fact is brought out in the second column 
of the Tables. 
(c.) The third fact which appears on comparing the averages is that the lengthening 
of the shells occurred slightly in the shells of the second terrace ; increasingly in those 
of the third and fourth ; reaching a point in the fifth terrace which is practically not 
afterwards exceeded in shells found as much as 30 feet lower, though the changes in 
texture, &c., had greatly progressed in these latter. Evidence will be given, moreover, 
which tends to show that this lengthening of the shells is more probably due to some 
other consequence of the diminished size of the lake than to the increase in saltness ; 
for example, to its increasing shallowness and consequent high average temperature in 
summer. Examples will be given of Cockles from lagoons both of the Aral Sea and 
in Egypt, which, while differing entirely from those of these salt lakes in general 
appeau’ance, are yet like them in the proportion of length to breadth. The whole 
question will be fully considered after the specimens have been described. The shells 
on the several terraces may now be described in detail. 
First Terrace .—The shells on the first terrace were, no doubt, living at the time 
when the Aral Sea was connected with this series of lakes and, perhaps, also for 
a short period after its separation from them. They lie at the foot of the hills coming 
down to the lake, and, though mostly covered with earth, it was possible to get plenty 
of them, especially among the debris thrown out by burrowing animals. Ther^ are, for 
the most part, smallish shells, being chiefiy 19 mm. to 24 mm. in length. They are 
MDCCCLXXXIX-B. 2 R 
