292 THE GEOLOGY OF THE LABRADOR COAST. 



most subarctic and all arctic lands the soil is but a few 

 inches deep. 



In all temperate regions the superficial deposits have 

 been characterized by Prof. Desor* to be " a succession 

 of rocky hills and drift plateaus or valleys, which can be 

 traced to the highest elevation of the country, near the 

 dividing ridge, each following plateau or valley being 

 commonly at a higher level than the preceding." This 

 state of things obtains in Labrador, but there is an im- 

 mense disproportion between the rocky hills and the 

 drift deposits. We find no sandy plains or level tracts 

 of glacial drift, or marine clays, distributed at intervals 

 from the coast to the interior. They take the form of 

 occasional, isolated sand-banks and cliffs of clay, of slight 

 extent, overhanging rivers, and which by their secluded 

 and retired positions have escaped the general denuda- 

 tion by the Labrador current which must have passed 

 over the lower levels of the peninsula subsequent to the 

 glacial epoch. In travelling in the interior we find our- 

 selves walking, when it is possible to walk or climb at 

 all, over the rocky floor of this inhospitable region, 

 smoothed in spots, though rarely striated by glaciers, 

 but on the coast more generally mangled and torn by 

 the action of shore-ice and frosts, which have here shown 

 a vast power. 



The Leda clays are mostly confined to the head of re- 

 tired bays, or if in more exposed- situations, lie between 

 bold headlands. The vast sand barrens of Canada and 

 New England spreading into broad plains, are here rep- 

 resented by precipitous masses of sand hanging upon the 



* Foster and Whitney's Report on the Geology of Lake Superior. 



