38 SHIFTING BOUNDARIES OF DESERTS. 



Prima facie, it would seem that it ought to be very 

 easy to see where a desert begins, or where it comes 

 to an end; and yet it is not so: as a matter of actual 

 fact, it is one of the most difficult things possible to 

 lay down a limit to the desert that will stand the test 

 of criticism ; and for this reason, that the almost rainless 

 zone, which forms the desert, is periodically visited by rain 

 storms of extreme violence, as well as by more moderate 

 showers which produce a season of verdant growth, 

 however short it may be, so that the limits of the 

 desert are constantly shifting backwards and forwards. 

 What is at one period of the year an expanse of 

 absolutely sterile sand or clay, at another time is 

 covered with verdure, and cattle may be pastured 

 upon it. 



In fixing the limits of the desert, or any other 

 zone, therefore, all that we aspire to do is to show 

 the approximate mean limits to which, as we humbly 

 conceive, the desert may be taken to extend. 



But we cannot too emphatically point out that there 

 exists no hard and fast line anywhere in Nature, ex- 

 actly circumscribing the geographical limits of any of 

 her productions: all her transformations are gradual, 

 and pass one into the other so imperceptibly that no 

 absolute line of demarcation is visible. If we take 

 any of the phenomena of Nature, and carefully examine 

 its conditions, we shall find this the invariable rule. 

 Take a well-known case, such as that of the limits of 

 the Polar ice: Arctic navigators are well aware that 

 these limits are continually shifting; some years the 

 sea is open much farther towards the pole; while at 

 other times the ice extends a long way further out 

 into the ocean than it did the previous year. When 



