20 SECTIONAL ADDRESSES. 
of condensation which supplies heat to an ascending current. Yet air 
cannot descend through the stratification without the necessary heat being 
extracted. On the other hand, we do know that air descends, for the air 
which goes up in the ascending currents, or rather, an equivalent amount, 
must come down somewhere. The solution of the problem is that air 
practically never descends through its environment, but comes down by 
the gradual subsidence of a whole column. This is generally brought 
about by the air at the bottom of the column spreading under the sur- 
rounding air and so lowering the air above in a way to be described in 
greater detail later. 
If now we consider the undisturbed atmosphere in different parts of 
the world, we find that each has its own stratification, which is mainly 
determined by the local radiation. At the equator the stratification is 
not so close as at the poles, and equivalent strata are higher in the atmo- 
sphere the further we move from the equator. If a large mass of air is 
transported as a whole without gain or loss of heat, no change in entropy 
occurs, and therefore it retains its original stratification. It is therefore 
clear that if masses of polar and tropical air are brought together the 
strata will not fit. The process is something like removing two geological 
specimens from different parts of a stratified rock and then placing them 
side by side. We can recognise the surface where the two masses meet 
by the discontinuity in the strata; in geology such a surface of discon- 
tinuity is called a fault. We shall consider later the consequence of 
bringing together masses of air of different origin in this way, and it will 
be shown that they interact like separate fluids, but throughout the 
resulting motion they retain their stratification, although this stratification 
becomes modified and distorted. 
This idea of the stratification of the atmosphere which has caused us 
to recognise that ascending and descending currents are relatively rare 
occurrences raises new problems as to how the solar energy is converted 
into the kinetic energy of winds. This leads me to the second subject of 
this address. 
The Mechanism of the Atmospheric Heat Engine. 
Brunt has calculated from considerations of wind and atmospheric 
friction that 25 x 1011 kilowatts of energy are required to maintain the 
motion of the atmosphere. It is generally agreed that this energy is 
derived from the solar radiation which falls on the earth, the atmosphere 
itself acting as a gigantic heat engine to convert the solar energy into the 
kinetic energy of the winds. How the atmospheric heat engine works is 
the problem which we are now to discuss. 
Until quite recently this problem seemed to present no difficulty. 
All atmospheric motion was referred in one form or another to the ascent 
of warm air through cold air and the descent of cold air through warm 
air. The so-called general circulation of the atmosphere was considered 
to be the direct consequence of the ascent of warm air at the equator and 
the descent of cold air at the poles, there being a permanent circulation 
from the equator to the poles in the upper atmosphere, with a return flow 
in the surface or middle layers. Similarly, cyclones were considered to 
form in regions where the air is warmer than the surrounding air, with a 
consequent upward motion of the warm air through its colder environment. 
Oe 
