292 '^HE PHYSICAL GEOGRAPHY OF THE SEA. 



produce geological phenomena in another, I return to the Dead 

 Sea and the great inland basins of Asia, and ask, How far is it 

 possible for the elevation of the South American continent, and 

 the upheaval of its mountains, to havQ had any effect upon the 

 water-level of those seas ? There are indications (§ 535) that the j 

 all once had a higher water-level than they now have, and that 

 formerly the amount of precipitation was greater than it now is ; 

 then what has become of the sources of vapor? What has di- 

 minished its supply ? Its supply would be diminished (§ 588) 

 either by the substitution of dry land for water-surface in those 

 parts of the ocean which used to supply that vapor ; or the quan- 

 tity of vapor deposited in the hydrographical basins of those seas 

 would have been lessened if a snow-capped range of mountains 

 (§ 536) had been elevated across the path of these winds, between 

 the places where they were supplied with vapor and these basins. 

 A chain of evidence which it would be difficult to set aside is 

 contained in the chapters beginning severally at pp. 71, 127, and 

 148, going to show that the vapor which supplies the extra-trop- 

 ical regions of the north with rains comes, in all probability, from 

 the trade-wind regions of the southern hemisphere. 



541. Now if it be true that the trade-winds from that part of 

 The path of the S.E. ^hc world take up there the water which is to be 

 trthtnorthem hem- rained in the extra-tropical north, the path ascribed 

 isphere. ^q ^\^q southeast tradcs of Africa and America, after 



they descend and become the prevailing southwest winds of the 

 northern hemisphere, should pass over a region of less precipita- 

 tion generally than they would do if, while performing the office 

 of southeast trades, they had blown over water instead of land. 

 The southeast trade-winds, with their load of vapor, whether great 

 or small, take, after ascending in the equatorial calms, a north- 

 easterly direction ; they continue to flow in the upper regions of 

 the air in that direction until they cross the tropic of Cancer. 

 The places of least rain, then, between this tropic and the pole, 

 should be precisely those places which depend for their rains upon 

 the vapor which the winds that blow over southeast trade-wind 

 Africa and America convey. Now, if we could trace the path of 

 the winds through the extra- tropical regions of the northern hemi- 

 sphere, we should be able to identify the track of these Andean 

 winds by the droppings of the clouds ; for the path of the winds 



