APPARENTLY CORRELATED TO THE CONDITIONS OE LIFE. 
301 
miles Meregen Sai), but always without any marked rising of the ground ; wliere 
any elevation occurs the level at which the shells cease is always definite and striking. 
The absence of shells above a definite level seems to suggest that the sea has never 
in recent times extended over parts above that level. There is nothing to suggest 
that any Aral Sea deposits, higher than this line of demarcation, have been denuded. 
For, had denudation been the cause of the absence of Aral shells above this line, it 
would be expected that the shells would gradually disappear on a line travelling up 
from the sea, and that they would disappear at different levels in different j)laces, 
which is not the case. If, therefore, the Aral Sea did ever extend over a greater 
tract of country than that which would be covered by it if it rose about 15 feet 
above its present level, it can only be supposed that such a condition occurred in the 
remote past, and not that it has gradually diminished to its present size from a much 
greater extent, as has been often suggested. Moreover, if the Aral Sea had recently 
retired from a greatly extended area, it must have covered the Kara Kum entirely, 
extending to Lake Tschalkar, which is marked on the Russian maps as a lake about 
40 miles long and 25 miles broad, forming the termination of the great valleys of the 
Irghiz and Turgai streams. In the belief that such a connection might have formerly 
existed between Lake Tschalkar and the Sea of Aral, I travelled down the Irghiz 
river as far as the lake. I found it to be a vast sheet of salt mud, which 
becomes dry in summer in most places. The joint stream of the Irghiz and Turgai 
never reaches the main part of the lake, becoming lost in reedy morasses of nearly 
fresh water at the western end. The lake was so dry that my camels crossed the 
west end of it in the beginning of August. Its northern shore is bounded by a 
range of hills which rise about 600 to 800 feet from the lake. Their southern front, 
which faces towards the lake, is nearly vertical, and is cut in places by ravines. These 
hills are composed of horizontal beds containing Eocene fossils, similar to those which 
were found in the hills on the north-west of the Aral Sea. Above these beds was a 
deposit of horizontally stratified sand about 80 feet thick. 
In no case, eitlier in the ravines, or among the hills, or on the shores of the lake, 
or in the dehi'is thrown up at the mouths of the wells, were any shells found other 
than those of the fossiliferous beds. There was no trace of the previous j)resence ot 
the Aral Sea. 'The ground did not differ in any way from considerable low-lying 
tracts near the Aral Sea, which remain covered with Cockles ; and, had the sea recently 
been in Lake Tschalkar, these shells could not have failed to be found in quantity. 
Also in the Kara Kum, excepting the above-mentioned low-lying tracts, the ground is 
without Cockles, but on descending to these depressions the deposit of shells is 
suddenly reached. This is true in the case of the north end of the depression, 
Jaksi Klich, which, tiiough 15 miles from the Aral Sea, was formerly joined with it 
by a channel, and equally true of the steepest pai'ts of the bank, ns, for example, 
where the southern slopes of Togusken rise almost vertically from the water’s edge. 
I have also every reason to believe that those parts of the Kora Kum which I did not 
