THE ANTARCTIC REGIONS AND THEIR CLIMATOLOGY. 459 



sufficient to keep the air away from the poles and produce almost 

 a vacuum there. In this state of things, the same air would 

 continue to revolve about the poles were not some other agent, 

 such as heat, brought in to expand and drive it away. Being 

 expanded and pufied out above the general atmospherical level, 

 but retaining its velocity — for the supposition is that it moves 

 without friction — and returning through the upper regions, it 

 would flow back as it went, viz., as a westerly wind, and arrive 

 at its null belt in the direction of the meridian. But the wind 

 has friction, and is resisted in every movement ; the atmosphere 

 partakes of the spheroidal form, which has been impressed upon 

 the earth itself by its axial rotation. That form is to it the form 

 of stability. The water at the pole is about 13 miles nearer to 

 the centre of the earth than the water at the equator ; but there 

 is not on that account any tendency in the sea to flow back from 

 the equator towards the poles ; neither is there any tendency to 

 motion one way or the other in the atmospherical ocean by 

 reason of the oblateness of its surface. To produce the polar and 

 equatorial movements of the air, there must be an agent both at 

 the equator and the poles to prevent such stability by constantly 

 disturbing equilibrium there, and that agent is heat ; therefore, 

 whatever be the degree of depression due the polar barometer in 

 consequence of axial rotation, such depression could, of itself, 

 produce neither trade nor counter-trade wind ; it could no more 

 produce currents in the air than in the sea, nor could axial rota- 

 tion produce a high barometer at one pole, a low barometer at 

 the other ; consequently, the difference in the pressure of the 

 atmosphere about the two poles, as shown by the diagram (Plate 

 XVI.), cannot be ascribed to the influence of axial rotation. It 

 is doubtless due to the excess in antarctic regions of aqueous 

 vapour and its latent heat. 



862. PsycJirometry of polar winds. — The arctic circle lies chiefly 

 on the land, the antarctic on the water. As the winds enter one, 

 they are loaded with vapour ; but on their way to the other they 

 are desiccated (§826). Northern mountains and the hills wring 

 from them water for the great rivers of Siberia and Arctic 

 America. These winds, then, sweep comparatively dry air across 

 the arctic circle; and when they arrive at the calm disc— the 

 place of ascent there — the vapour which is condensed in the act 

 of ascending does not liberate heat enough to produce a rarefac- 

 tion sufficient to call forth a decided indrau2;ht from a crreater 



