MOVEMENTS OF THE POLAR ICE. 39 



therefore the geographer draws upon his map an ideal 

 line, marking the equatorial margin of the Polar ice, 

 it is well understood to signify that such is merely its 

 approximate mean limit during ordinary seasons, and 

 that it by no means presumes to fix an invariable line 

 from which the great ice movements never deviate 

 one way or the other. 



The boundaries which we have ventured to fix to 

 each of our Climatic Zones are therefore to be under- 

 stood to be subject to similar conditions. In each case 

 we have been careful to set forth the reasons which 

 have seemed to warrant our decision, and also to ex- 

 plain, as far as that has been possible, the causes which 

 produce exceptional climatic conditions in particular 

 regions ; as for instance in South America : Why 

 should large portions of the western seaboard be 

 practically a rainless and treeless desert, while upon 

 the same parallel of latitude, upon its eastern and 

 central portions, there is a tremendous rainfall, and 

 the land is covered by impenetrable forests of the 

 most luxuriant growth? The reasons for this will be 

 apparent when we come to consider this question. 

 Fortunately, in this case, the causes of these wonderful 

 differences of climate are so manifest as to be ac- 

 cepted by geographers in general as furnishing a 

 complete explanation of this wonderful phenomenon. 

 The matter is merely referred to at present, however, 

 with a view to show that exceptional causes are apt 

 to produce exceptional effects in every part of the 

 world, which of course still further complicates the 

 difficulty of saying what are the limits of, say, the 

 Desert Zone, with any certainty in any part of the 

 world. 



