160 THE rilYSlCAL GEOGRAPHY OF THE SEA. 



the air? Heat and cold, the early and the latter rain, clouds and 

 sunshine, are not, we may rely upon it, distributed over the earth 

 by chance ; they are distributed in obedience to laws that are as 

 certain and as sure in their operations as the seasons in their 

 rounds. If it depended upon chance whether the dry air should 

 come out on this side or on that of this calm belt, or whether the 

 moist air should return or not whence it came — if such were the 

 case in nature, we perceive that, so far from any regularity as to 

 seasons, we should have, or might have, years of drought the most 

 excessive, and then again seasons of rains the most destructive ; 

 but, so far from this, we find for each place a mean annual pro- 

 portion of both, and that so regulated withal, that year after year 

 the quantity is preserved with remarkable regularity. Having 

 thus shown that there is no reason for supposing that the upper 

 currents of air, when they meet over the calms of Cancer and Cap- 

 ricorn, are turned back to the equator, but having shown that 

 there is reason for supposing that the air of each current, after de- 

 scending, continues on in the direction toward which it was trav- 

 eling before it descended, we may go farther, and, by a similar 

 train of circumstantial evidence, afforded by these researches and 

 other sources of information, show that the air, kept in motion on 

 the surface by the two systems of trade- winds, when it arrives at 

 the belt of equatorial calms and ascends, continues on thence, each 

 current toward the pole which it was approaching while on the 

 surface. In a problem like this, demonstration in the positive 

 way is difficult, if not impossible. We must rely for our proof 

 upon philosophical deduction, guided by the lights of reason ; and 

 in all cases, in which positive proof can not be adduced, it is per- 

 mitted to bring in circumstantial evidence ; and the circumstan- 

 tial evidence afforded by my investigations goes to show that the 

 winds represented by O, Q, § 215, do become those represented by 

 R, S, T, U, V, A, and A, B, C, D, B, F respectively. In the first 

 place, O, Q represents the southeast trade-winds — i. e., all the winds 

 of the southern hemisphere as they approach the equator ; and is 

 there any reason for supposing that the atmosphere does not pass 

 freely from one hemisphere to another ? On the contrary, many 

 reasons present themselves for supposing that it does. If it did 

 not, the proportion of land and water, and consequently of plants 

 and warm-blooded animals, being so different in the two hemi- 



