——eo- 
A.—MATHEMATICS AND PHYSICS. 21 
The anticyclone, on the other hand, was considered to be a region of cold 
descending air. Thus cyclone and anticyclone were regions of ascent of 
warm and descent of cold air respectively. 
But I have already shown that the thermal stratification of the atmo- 
sphere, except in the Doldrums and occasionally in other regions, is 
prohibitive of such ascending and descending currents. Further, 
observations have shown that there is no direct flow offair from the 
equator to the poles in the upper atmosphere, and measurements of 
temperature in cyclones and anticyclones have shown that the former are 
not warm and the latter are not cold. 
Although the old ideas were wrong in detail, they were, of course, 
right in principle, for the potential energy inherent in masses of air at 
different temperatures must be the origin of the kinetic energy of the 
winds, the difference in temperature between the equator and the poles 
being responsible for the general circulation of the atmosphere, and the 
difference in temperature between neighbouring masses of air for the 
energy of cyclones and anticyclones. The only question is, how does the 
transfer from potential to kinetic energy take place ? 
The solution of the problem was given by Margules in a series of 
papers, commencing in 1903, in which he investigated the energy developed 
in storms. Unfortunately, these papers were very obscure, and, as few 
meteorologists had then realised the insufficiency of the old theories, they 
received little attention outside Vienna. 
Margules’ papers were difficult because he dealt with the problem in a 
general way and treated the actual case of a compressible atmosphere, 
the density of each part of which varies with temperature and pressure. 
The physical principles involved are, however, extremely simple and are 
similar to those met with in problems dealing with the equilibrium of 
fluids of different densities. Suppose that we have a vessel in the form 
of a tank with a vertical division in the middle separating it into two parts, 
and let one of these parts be filled with oil and the other with water, both 
to the same depth (Fig. 2). If the partition is now withdrawn, the water 
settles down and flows under the oil, while the oil rises and flows into a 
horizontal layer over the water. A simple diagram of the position of the 
centre of gravity of the whole mass of liquid before and after the change 
shows that the centre of gravity has fallen. This change in the position 
of the centre of gravity has released the energy which set the liquid in 
‘motion. In this experiment the oil has been lifted, but it did not rise 
through the water, it was pushed up along the surface of discontinuity, 
which continued sharply marked during the whole period of change. In 
his complete investigation Margules showed that if two bodies of air 
having different temperatures are brought side by side, they react towards 
one another like the oil and water in the experiment just described, and 
he showed that the energy released by the lowering of the centre of gravity 
of the mass of the air as a whole is sufficient to account for the most 
violent storms met with in the atmosphere. 
This work leads to an entirely new idea as to the method in which solar 
energy is converted into the kinetic energy of atmospheric motion. 
Instead of warm air rising vertically like the warm gases in a chimney, 
drawing air in at the bottom and delivering it at the top, we see two 
bodies of air, one warm and the other cold, brought side by side, then the 
