§ 285-287. RAINS AND RIVERS. 107 



cording to the latitude in which they reach the shore, and con- 

 sequently with an average dew-point not higher than the mean 

 temperature. Classifying the winds of this part of the ocean ac- 

 cording to the halves of the horizon as east and west, the mean 

 of 44,999 observations in the log-books of the Observatory shows 

 that, on the average, the west winds blow annually 230 and the 

 east winds 122 days. 



285. Taking all these facts and circumstances into considera- 

 Tiie vapor-springs for tion, and without prctcuding to determine how 



all these rivers not in ^ n ,^ , ^ ' ^ .^ • ok • ^ 



tiie Atlantic Ocean, mucu oi the watcp wmch the rivers 01 America and 

 Europe carry into this part of the ocean comes from it again, we 

 may with confidence assume that the winds do not get vapor 

 enough from this part of the ocean to give rain to Europe, to the 

 Mississippi Yalley, to our Atlantic slopes, and the western half 

 of Asiatic Eussia. We are authorized in this conclusion, just as 

 we have authority to say that the evaporation from the Mediter- 

 ranean is greater in amount than the volume of water discharged 

 into it again by the rivers and the rains ; only in this case the re- 

 verse takes place, for the rivers empty more water into the At- 

 lantic than the winds carry from it. This fact also is confirmed by 

 the hydrometer, for it shows that the water of the North Atlantic 

 is, parallel for parallel, lighter than water in the Southern Ocean. 



286. The inferenc;, then, from all this is, that the place in the 

 The places in the sea sca (§ 276) whcucc comc thc' watcrs of the Missis- 

 rrs^oTthenorthtSl- sippi and othcr great rivers of the northern hemi- 

 c?Snt^tThTcaim sphcrc is to bc found in these southern oceans, and 

 ^^^^^' the channels by which they come are to be search- 

 ed out aloft, in the upper currents of the air. Thus we bring evi- 

 dence and facts which seem to call for a crossing of air at the calm 

 belts, as represented by the diagram of the winds, Plate I. It 

 remains for those who deny that there is any such crossing — who 

 also deny that extra-tropical rivers of the northern are fed by 

 rains condensed from vapors taken up in the southern hemisphere 

 — to show whence come the hundreds of cubic miles of water 

 which these rivers annually pour into the Atlantic and the Arc- 

 tic Oceans. In finding the "place" of all this water, it is incum- 

 bent upon them to show us the winds which bring it also, and 

 point out its channels. 



287. "In the greater number of physical investigations some 



