The atmospheric circulation, then, may be readily regarded as the movement, over the terrestrial surface, of gaseous masses which absorb and return energy in regions of high and low velocity exactly in the fashion explained above for solid material. In their movement from polar towards equatorial regions these masses, by the action of the aqueous vapour (§ 38), absorb energy (axial energy) and expand upwards against gravity. Here we have an energy operation identical in nature with that embodied in the movements of a pendulum mass simultaneously over a spherical surface and against radial springs as in the system of rotating pendulums, or identical with the equatorial and radial movement of the auxiliary masses in the hypothetical system. The return movement of the aerial masses over the terrestrial surface in the opposite direction from equatorial to polar regions provides also exactly the same phenomena of energy return as the return movement of the masses in our illustrative systems. These systems, in fact, portray the general operation of mechanical energy precisely as it occurs in the terrestrial atmospheric machine. But obviously they cannot illustrate the natural conditions in their entirety. The passage or flow of the atmospheric air masses over the earth's surface is a movement of an exceedingly complex nature, impossible to illustrate by experimental apparatus. And indeed, such illustration is quite unnecessary. As already pointed out (§ 38), no matter what may be the precise path of an aerial mass in its movement towards the planetary surface the final energy return is the same. Sooner or later its energy of position is restored in the original axial form.
The terrestrial atmospheric machine will be thus readily recognised as essentially a material mechanical machine corresponding in general nature to the illustrative examples described above. The combination of its various energy processes is embodied in a complete cyclical and reversible operation. Its energy capacity, as in the simpler cases, is strictly limited by the total mass of the operating material. The active or working energy is well within the limit for reversibility (§ 23), and the machine is therefore essentially stable in nature. The continuous abstraction of axial energy by the aqueous vapour is balanced by an equally continuous return from the air masses, and the system, so far as its energy properties are concerned, is absolutely conservative. Energy transmission from or to any external source is neither admissible nor necessary for its working.