142 THE PHYSICAL GEOGRAPHY OF THE SEA. 



part, tliej bring their moisture with them from the trade-wind re- 

 gions of the opposite hemisphere. 



374. I have marked on Plate YII. the supposed track of the 

 "Passat-Staub," showing where it was taken up in South Amer- 

 ica, as at P, P, and where it was found, as at S, S ; the part of 

 the line in dots denoting where it was in the upper current, and 

 the unbroken line where it was wafted by a surface current ; also 

 on the same plate is designated the part of the South Pacific in 

 which the vapor-springs for the Mississippi rains are supposed to 

 be. The hands (|^^) point out the direction of the wind. Where 

 the shading is light, the vapor is supposed to be carried by an up- 

 per current. 



375. Such is the character of the circumstantial evidence which 

 induced me to suspect that some agent, whose office in the grand 

 system of atmospherical circulation is neither understood nor rec- 

 ognized, was at work in these calm belts. 



'376. Dr. Faraday has shown that, as the temperature of oxygen 

 is raised, its paramagnetic force diminishes, being resumed as the 

 temperature falls again. 



" These properties it carries into the atmosphere, so that the 

 latter is, in reality, a magnetic medium, ever varying, from the 

 influence of natural circumstances, in its magnetic power. If a 

 mass of air be cooled, it becomes more paramagnetic ; if heated, 

 it becomes less paramagnetic (or diamagnetic), as compared with 

 the air in a mean or normal condition."* 



377. 'Now, is it not more than probable that here we have, in 

 the magnetism of the atmosphere, that agent which guides the 

 air from the south (§ 373) through the calms of Capricorn, of the 

 equator, and of Cancer, and conducts it into the north ; that agent 

 which causes the atmosphere, with its vapors and infusoria, to flow 

 above the clouds from one hemisphere into the other, and whose 

 footprints had become so palpable ? 



378. Taking up the theory of Ampere with regard to the mag- 

 netic polarity induced by an electrical current, according as it 

 passes through wire coiled 7mt/i or coiled against the sun, and ex- 



* Philosophical Magazine and Journal of Science, 4th Scries, No. I., January, 1851, 

 page 73. 



