234 THE PHYSICAL GEOGRAPHY OF THE SEA, 



into the Atlantic ; on tlie other hand, there are in Asia the Gan- 

 ges, and all the great rivers of China ; and in North America, in 

 the latitude of the Caspian Sea, is our great system of fresh-water 

 lakes ; all of these receive from the atmosphere immense volumes 

 of water, and pour it back into the sea in streams the most mag- 

 nificent. 



652. It is remarkable that none of these copiously-supplied 

 water-sheds have, to the southwest of them in the trade-wind re- 

 gions of the southern hemisphere, any considerable body of land ; 

 they are, all of them, under the lee of evaporating surfaces, of 

 ocean waters in the trade-wind regions of the south. Only those 

 countries in the extra tropical north which I have described as 

 lying under the lee of trade-wind South America and Africa are 

 scantily supplied with rains. Pray examine Plate VII. in this 

 connection. It tends to confirm the views taken in Chapter YI. 



653. The surface of the Caspian Sea is about equal to that of 

 our lakes; in it, evaporation is just equal to the precipitation. 

 Our lakes are between the same parallels, and about the same 

 distance from the western coast of America that the Caspian Sea 

 is from the western coast of Europe ; and yet the waters dis- 

 charged by the St. Lawrence give us an idea of how greatly the 

 precipitation upon it is in excess of the evaporation. To wind- 

 ward of the lakes, and in the trade-wind regions of the southern 

 hemisphere, is no land ; but to windward of the Caspian Sea, and 

 in the trade-wind region of the southern hemisphere, there is land. 

 Therefore, supposing the course of the vapor-distributing winds 

 to be such as I maintain it to be, ought they not to carry more 

 water from the ocean to the American lakes than it is possible for 

 them to carry from the land — from the interior of South Africa 

 and America — to the valley of the Caspian Sea ? 



654. In like manner (§ 393), extra-tropical iSTcw Holland and 

 South Africa have each land— not water— to the windward of 

 them in the trade-wind regions of the northern hemisphere, where, 

 according to this hypothesis, the vapor for their rains ought to be 

 taken up : they are both countries of little rain ; but extra-trop- 

 ical South America has, in the trade-wind region to windward of 

 it in the northern hemisphere, a great extent of ocean, and the 



