276 THE EFFECT OF MOISTURE 



sisted in (1) a diminution in the thickness of the shells, 

 this being first apparent in the shells of the third ter- 

 race. So marked was this change that the shells of the 

 seventh or lowest terrace were almost horny and semi- 

 transparent, and their weight was not a third that of 

 shells from the first two terraces; (2) a diminution in 

 the size of the beak; (3) a high colouration in the 

 shells. This change occurred almost uniformly, the 

 shells of each terrace being very nearly alike in texture, 

 thickness, and degree of colouration; (4) grooves be- 

 tween the ribs appearing on the inside of the shell as 

 ridges with rectangular faces; (5) a great diminution 

 in the absolute size of the shells on the lowest terrace; 

 (6) an increase in the length (greatest antero-posterior 

 dimension) of the shells in proportion to their breadth, 

 this ranging from the average ratio of 1 : .80 in the shells 

 from the first terrace to 1:.725 in shells from the 

 seventh. In Jaksi Klich lake the shells from the 

 lowest and most saline deposit were even more elon- 

 gated, the ratio of length to breadth being as 1: .68 for 

 samples of smaller shells, and 1:.66 for samples of 

 larger ones. Those from the lowest deposit of Jaman 

 Klich showed about the same degree of elongation as 

 those from the lowest terrace of Shumish Kul. 



It was very noticeable that the shells of each sample, 

 whether from a separate lake or only from a particular 

 terrace, resembled each other more closely than they 

 did shells from one of the other lakes, or those from 

 another terrace in the same lake, as at Shumish Kul. 



In each of the three lakes mentioned (and also in an 

 entirely distinct locality, the lagoon of Abu Kir near to 

 Alexandria), it was thus found that shells which had 



