204 DISCUSSION OF THE DESICCATION QUESTION. 



What has been said in the present section in regard to precipitation, and 

 the general nature of the causes regulating its amount and distribution, will 

 find its proper application farther on. It seemed impossible to proceed in this 

 discussion without giving the reader a sub-stratum of fiicts on which to base 

 the arguments which are next to follow. The main points sought to be 

 made, in reference to the amount of precipitation as influenced by and de- 

 pendent on temperature and area of surface exposed to evaporation, will 

 be admitted by all without hesitation. In reference to the Glacial epoch, 

 difficulties may arise in seeking for the causes of a fall of rain and snow so 

 irregular in its distribution as that demanded by the conditions then pre- 

 vailing. But it will be better to state these difficulties frankly, even if it 

 be found impossible to meet them satisfactorily, rather than to pass them 

 over entirely, or to assume, in pointing them out, that they need no expla- 

 nation because quite analogous to what is now taking place on the eai th's 

 surface, while in reality all the facts are entirely opposed to any such 

 assumption. 



A few words may be added in reference to the conditions favoring pre- 

 cipitation in the form of snow ; but not much need be said on that subject 

 at present, since it will necessarily come up again in various forms in the 

 succeeding chapter. The reader must bear in mind, however, that while a 

 low temperature is necessary to the formation of snow, the moisture from 

 which that snow is formed must have come from a comparatively warm 

 region. Thus the grand glaciers on the summits of the Himalayan ranges 

 are made up of frozen Avater which has been brought — in large part, at least 

 — from the surface of the Indian Ocean. Portions of the earth where the 

 mean temperature is verv low remain almost bare of snow throughout the 

 year, while other regions, in close proximity, and where the temperature is 

 higher, are covered with immense glaciers, as will be more fully set forth in 

 the next chapter. 



Since decrease of temperature is the residt of increased elevation, we are 

 able to have any amount of cold in any latitude. Even at the equator 

 eternal snow may and does exist ; and if conditions were favorable, such 

 snow accumulations might occur on the most gigantic scale. In fact, most 

 of that part of the earth in high latitudes which has a very low mean tem- 

 perature is but thinly covered by snow, and that not remaining through the 

 year ; while the cold, because elevated, regions within the tropics, or in their 

 vicinity, are always the abode of snow and ice. 



