SEA ROUTES, CAL3I BELTS, AND VARIABLE WINDS. 355 



lent. Now, in each case, the prevailing winds blow (§ 657) from 

 the high to the low barometer (Plate I.). 



G60. The barometric ridges. — The fact of two barometric ridges 

 encircling the earth, as the high barometer of the tropical calm 

 belts do, and as they may be called (Plate I.), suggests a place of 

 low barometer on the polar side as naturally as the ascent of a 

 hill on one side suggests to the traveller a descent on the other ; 

 and, had not actual observations revealed the fact, theory should 

 have taught us (§ 654) the existence of a low barometer towards 

 the polar regions as well as towards the equatoriah 



661. TJiey make a depression in the atmosphere. — Let us contem- 

 plate for a moment this accumidation of air in the tropical belt 

 about the earth in each hemisphere. Because it is an accumula- 

 tion of atmospheric air about the calms ; — because the barometer 

 stands higher under the calm belt of Capricorn, for instance, than 

 it does on any other parallel between that calm belt and the ]Dole 

 on one side, or the equator on the other, it is not to be inferred 

 that therefore there is a piling — a ridging up — of the atmosphere 

 there. On the contrary, ^vere the upper surface of our atmo- 

 sphere visible, and could we take a view of it from above, we 

 should discover rather a valley than a ridge over this belt of 

 greatest pressure ; and over the belt of least pressure, as the 

 equatorial calm belt, we should discover (§ 520), not a valley, 

 but a ridge, and for these reasons : In the belts of low^ barometer, 

 that is, in both the equatorial and polar calms, the air is ex- 

 panded, made light, and caused to ascend chie% by the latent 

 heat that is liberated by the heavy precipitation which takes 

 place there. This causes the air which ascends there to rise up 

 and swell out far above the mean level of the great aerial ocean. 

 This intumescence at the equatorial calm belt has been estimated 

 to be several miles above the general level of the atmosphere. 

 This calm belt air, therefore, as it boils up and flow^s off through 

 the upper regions, north and south, to the tropical calm belts, 

 does not so flow by reason of any diiference of barometric pres- 

 sure, like that which causes the surface winds to blow, but it so 

 flows by reason of difference as to level. 



662. TJie upper surface of the atmosphere. — The troi^ical calm 

 belts (§ 278) are places where the mean amount of precipitation 

 is small. The air there is comparatively dry air. So far from 

 being expanded by heat, or swelled out by vapour, this air is 

 contracted by cold, for the chief source of its supply is through 



2 A 2 



