EASTING OF THE TRADE-WINDS, ETC. 161 



for each of the four quarters in every band was ascertained. 

 Considering difference of temperature between these various 

 bands to be one of the chief causes of movement in the atmo- 

 sphere, — that the extremes on one hand are near the equator, 

 and on the other about the poles ; — considering that the tendency 

 of every wind (§ 234) is to blow along the arc of a great circle, 

 and that consequently every wand that was observed in any one 

 of these bands must have moved in a path crossing these bands 

 2nore or less obliquely, and that therefore the general movements 

 in the atmosphere might be classed accordingly, as winds either 

 with northing or with southing in them ; — we have so classed 

 them ; and we have so classed them that we might study to more 

 advantage the general movements of the great atmospherical 

 machinery. See Plate XV. 



353. The medial hands. — Thus, when, after so classing them, 

 we come to examine those movements in the band between 5° 

 and 10° south, and to contrast them with the movements in the 

 band between 55° and 60° south, for example, we find the general 

 movements to be exactly in opposite directions. Observations 

 show that during the year the winds in the former blow toivards 

 the equator 283, and/rora it 73 days ;, and in the latter they blow 

 toward the pole for 224, and from it 132 days. These facts show 

 that there must be a place of rarefaction — of low barometer, an 

 indraught tow^ards the poles as well as the equator ; — and that 

 consequently, also, there must be a medial line or band some- 

 where between the parallels of 10° and 55° south, on one side of 

 w^hich the prevailing direction of the wind is towards the equator, 

 -on the other towards the 'pole. So, in the northern hemisphere, 

 the same series of observations point this medial band out to us. 

 They show that one is near the calm belt of Capricorn, the other 

 near the calm belt of Cancer, and that they both probably lie 

 between the parallels of 35° and 40°, where the winds north and 

 south are equal, as per table, page 162. 



The wind curves (Plate XV. and the table) aJBford a very 

 striking view of these medial bands, as the parallels in either 

 hemisphere between w^hicli the winds with northing and the 

 winds with southing are on the yearly average exactly equal. 

 In the northern hemisphere the debatable ground appears by the 

 table to extend pretty nearly from 25° to 50° N. By the plate 

 the two winds first become equal between 25° and 30° ; the two 

 curves then recede and approach very closely again, but without 



M 



