MAGNETISM AND CIRCULATION OF THE ATMOSPHERE. 107 
regions of the atmosphere from the equator, do not, after arriving 
at the calms of Cancer, and descending, return to the equator on 
the surface, but that they, continue on the surface toward the pole- 
But why should they ? What agent in nature is there that can 
compel these, rather than any other winds, to take such a circuit ? 
183. The following are some of the facts and circumstances 
which give strength to the supposition that these winds do contin- 
ue from the calm belt of Cancer toward the pole as the prevailing 
southwesterly winds of the extra-tropical north : 
We have seen (Plate 1.) that, on the north side of this calm 
zone of Cancer, the prevailing winds on the surface are from this 
zone toward the pole, and that these winds return as A through 
the upper regions from the pole ; that, arriving at the calms of 
Cancer, this upper current A meets another upper current G from 
the equator, where they neutralize each other, produce a calm, 
descend, and come out as surface winds, viz., A as B, or the trade- 
winds ; and G as H, or the variable winds. 
184. Now observations have shown that the winds represented 
by H are rain winds ; those represented by B, dry winds ; and it 
is evident that A could not bring any vapors to these calms to 
serve for H to make rains of ; for the winds represented by A have 
already performed the circuit of surface winds as far as the pole, 
during which journey they parted with all their moisture, and, re- 
turning through the upper regions of the air to the calm belt of 
Cancer, they arrived there as dry winds. The winds represented 
by B are dry winds ; therefore it was supposed that these are but 
a continuation of the winds A. 
185. On the other hand, if the winds A, after descending, do 
turn about and become the surface winds H, they would first have 
to remain a long time in contact with the sea, in order to be sup- 
plied with vapor enough to feed the great rivers, and supply the 
rains for the whole earth between us and the north pole. 
In this case, we should have an evaporating region on the north 
as well as on the south side of this zone of Cancer; but investi- 
gation shows no such region ; I speak exclusively of the ocean. 
186. Hence it was inferred that A and G do come out on the 
surface as represented by Plate I. But what is the agent that 
should lead them out by such opposite paths ? 
187. According to this mode of reasoning, the vapors which 
