1921.] The Eighth Indian Science Congress. XCIX 
as that at the surface. The molecules down here have a 
- mean free path of only about 10-5 cm., while at the upper 
height considered they have 1000 kilometres (600 miles) to 
travel on the average before they make a single molecular 
with the dynamical process of cooling by expansion. Conduc 
is a slow process, and only appreciable when the overlying air 
is dry and therefore non productive; and intermixture, though it 
certainly may produce condensation, is seldom extensive enough 
in nature to show more than an insignificant effect. That some 
rain formation may occur follows from the fact that the capacity 
s 
temperature of the air itself, and that consequently if two 
masses of air, at or near saturation but at different tempera- 
tures, are forced to mingle, the resultant mixture may pass the 
saturation point corresponding with the ne 
the whole. But that the effect must always be small can be 
seen by a numerical example ; for ii we start 
extreme ease of two masses of saturated air, each mass 1000 
metres high, and originally at temperatures 75° and 95 res- 
pectively, we shall extract from the hotter mass by its admix- 
ture with the cooler a fall of only 43 cents of rain, out of its 
total water capacity of 34 times that amount. : 
- 19. Now instead of mixing our columns of air, assume 
that the column at 95° F is merely raised mechanically above 
the other column and extends from the 1000 m. level to some- 
thing beyond 2000 m. It will then be found that the upper 
column will drop some 90%, of all the vapour it holds, giving 
