THE ATMOSPHERE. 83 



be the source of at least two systems of currents in the air, viz., 

 an upper and a lower. The two agents comhined, viz., that which 

 changes level or volume, and that which changes specific gravity, 

 give us the general currents under consideration. Hence we say 

 that the j>rimum mohile of the air is derived from change of spe- 

 cific gravity induced by the freezing temperature of the polar re- 

 gions, as well as from change of specific gravity due the expand- 

 ing force of the sun's rays within the tropics. 



154. Therefore, fairly to appreciate the extent of the influence 

 due the heat of the sun in causing the winds, it should he recol- 

 lected that we may with as much reason ascribe to the inter-trop- 

 ical heat of the sun the northwest winds, which are the prevailing 

 winds of the extra-tropical regions of the southern hemisphere, or 

 the southwest winds, which are the prevaiHng winds of the extra- 

 tropical regions of the northern hemisphere, as we may the trade- 

 winds, which blow in the opposite directions. Paradoxical, there- 

 fore, as it seems for us to say that the heat of the sun causes the 

 winds between the parallels of 25° or 30° north and south to 

 blow toward the equator, and that it also causes the prevaiHng 

 winds on the polar sides of these same parallels to blow toward 

 the poles, yet the paradox ceases when we come to recollect that 

 by the process of equatorial heating and polar cooling which is 

 going on in the atmosphere, the specific gravity of the air is 

 changed as weU as its level. Nevertheless, as HaUey said, in his 

 paper read before the Royal Society in London in 1686, and as 

 we also have said (§ 133), "it is likewise very hard to conceive 

 why the limits of the trade-wind should be fixed about the paral- 

 lel of latitude 30° all around the globe, and that they should so 

 seldom exceed or fall short of those bounds." 



155. Operated upon by the equilibrating tendency of the at- 

 mosphere and by diurnal rotation, the wind approaches the north 

 pole, for example, by a series of spu'als from the southwest. If 

 we draw a circle about this pole on a common terrestrial globe, 

 and intersect it by spirals to represent the direction of the wind, 

 we shall see that the wind enters aU parts of this circle from the 

 southwest, and that, consequently, there should be about the poles 

 a disc or circular space of calms, in which the air ceases to move 



