THE ATMOSPHERE. 
87 
calm belt passes it the rains cease, and the people in that latitude 
have no more wet weather till the fall, when the belt of calms re- 
crosses this parallel on its way to the south. By examining the 
*' Trade-wind Chart," it may be seen what the latitudes are that 
have two rainy seasons, and that Bogota is within the bi-rainy 
latitudes. 
133. The Rainless Regions. — The coast of Peru is within the 
region of perpetual southeast trade-w^inds. Though the Peruvian 
shores are on the verge of the great South Sea boiler, yet it never 
rains there. The reason is plain. 
The southeast trade-winds in the Atlantic Ocean first strike the 
water on the coast of Africa. Traveling to the northwest, they 
blow obliquely across the ocean until they reach the coast of 
Brazil. By this time they are heavily laden with vapor, which 
they continue to bear along across the continent, depositing it as 
they go, and supplying with it the sources of the Rio de la Plata 
and the southern tributaries of the Amazon. 
Finally they reach the snow-capped Andes, and here is wrung 
from them the last particle of moisture that that very low temper- 
ature can extract. 
Reaching the summit of that range, they now tumble down as 
cool and dry winds on the Pacific slopes beyond. Meeting with 
no evaporating surface, and with no temperature colder than that 
to which they were subjected on the mountain-tops, they reach 
the ocean before they become charged with fresh vapor, and be- 
fore, therefore, they have any which the Peruvian climate can ex- 
tract. Thus we see how the top of the Andes becomes the res- 
ervoir from which are supplied the rivers of Chili and Peru, 
134. The other rainless or almost rainless regions are the west- 
ern coasts of Mexico, the deserts of Africa, Asia, North America, 
and Australia. Now study the geographical features of the coun- 
try surrounding those regions ; see how the mountain ranges run ; 
then turn to Plate VIII. to see how the winds blow, and where the 
sources are 87) which supply them with vapors. This plate 
shows the prevailing direction of the wind only at sea ; but know- ^ 
ing it there, we may infer what it is on the land. Supposing it to 
prevail on the land as it generally does in corresponding latitudes 
at sea, then the Plate will suggest readily enough how the winds 
that blow over these deserts came to be robbed of their moisture, 
