136 THE PHYSICAL GEOGRAPHY OF THE SEA. 



345. After impinging upon the cold hill-tops of the Patagonian 

 coast, and passing the snow-clad summits of the Andes, this same 

 wind tumbles down upon the eastern slopes of the range as a dry 

 wind ; as such, it traverses the almost rainless and barren regions 

 of cis-Andean Patagonia and South Buenos Ajres. 



346. These conditions, the direction of the prevailing w^nds, 

 and the amount of precipitation, may be regarded as evidence af- 

 forded by nature, if not in favor of, certainly not against, the con- 

 jecture that such may have been the voyage of this vapor through 

 the air. At any rate, here is proof of the immense quantity of 

 vapor which these winds of the extra-tropical regions carry along 

 with them toward the poles ; and I can imagine no other place 

 than that suggested, whence these winds could get so much va- 

 por. 



I am not unaware ol the theory, or of the weight attached to it, 

 which requires precipitation to take place in the upper regions of 

 the atmosphere on account of the cold there, irrespective of prox- 

 imity to mountain tops and snow-clad hills. 



347. But the facts and conditions developed by this system of 

 research upon the high seas are in many respects irreconcilable 

 with that theory. With a new system of facts before me, I have, 

 independent of all preconceived notions and opinions, set about to 

 seek among them for explanations and reconciliations. 



348. These may not in all cases be satisfactory to every one ; 

 indeed, notwithstanding the amount of circumstantial evidence 

 that -has already been brought to show that the air which the 

 northeast and the southeast trade-winds discharge into the belts 

 of equatorial calms, does, in ascending, cross — that from the south- 

 ern passing over into the northern, and that from the northern 

 passing over into the southern hemisphere (see F and G, B and 

 C, Plate I.) — yet some have implied doubt by asking the ques- 

 tion, "How are two such currents of air to pass each other?" 

 And, for the w^ant of light upon this point, the correctness of rea- 

 soning, facts, inferences, and deductions have been questioned. 



349. In the first place, it may be said in reply, the belt of equa- 

 torial calms is often several lumdred miles across, seldom less than 

 sixty ; whereas the depth of the volume of air that the trade- winds 



