182 Mr W. Bateson, On Variations of Cardium edule. [Feb.13, 
In the case of Shumish Kul the sides of the lake have been 
terraced probably by the oscillation of the water under the wind. 
On these terraces the cockles are found, generally as paired 
valves with their ligaments more or less preserved, sitting up 
in their natural attitude. As the lake dried up, the water must 
have become salter; hence the cockles on any one terrace must 
have lived in salter water than that which was inhabited by 
the cockles of the terrace above it. Passing from above down- 
wards a progressive series is thus obtained. On comparing samples 
from each terrace it appears that under the influence of increasing 
salinity the shells (1) became thinner (ten shells from the lowest 
bed weigh 1:54 gr. while ten shells of the same size from the 
highest weigh 5:18 gr.), (2) their length in an antero-posterior 
direction increased relatively to their dorso-ventral width, (3) they 
became much more highly coloured, (4) only the lowest shells 
shew diminution in absolute size, (5) deformed shells are very 
common on the lower terraces. 
The chief interest of these facts lies in the almost perfect 
uniformity of each sample as regards texture and colour. The 
uniformity in shape is not so great. Even amongst the longest 
shells individual rounder ones occur, but if measurements of a 
number be taken together, the increase in average length is con- 
siderable. 
The same changes are visible in the case of the cockles in Jaksi 
Klich and in Jaman Klich, though these lake-beds shew hardly 
any terracing. As all these three lakes were isolated from each 
other, it may be concluded that the variations observed occurred 
as the specific consequence of the changed conditions. 
(2) The character of the geological formation a factor in Zoo- 
geographical Distribution, illustrated by observations in Portugal 
and Spain. By H. Gavow, M.A., King’s College. 
The observations concern the distribution of all the species of 
Amphibia and Reptiles in the Peninsula, based upon a great 
number of data. 
The method adopted is the followmg. The various species 
and localities are arranged in six classes, which represent the 
most typical geological formations, so far as the purpose of the 
investigation is concerned. The proportionate areas of the various 
formations are given in per cent. of the whole country. 
On Granitic ground, equal to 29 per cent. of the total area of 
Portugal, 65 localities for Amphibia were recorded. On 42 per 
cent. there should be 94, but there are only 15 on record, these 
42 per cent. being Palaeozoic. Consequently Granitic ground is 
about six times more favourable to Amphibian life than is Palaeo- 
