near 



Mount Washington. 



9 



upon by streams, will eventually be changed to gravel and sand, 

 and be washed down the mountain, leaving fresh surfaces for the 

 renewed winter freezing. In this way, little by little, the exca- 

 vation goes on, the deep, ragged ravines notching the mountains 

 where the formation happens to be slaty and permeated with 

 numerous joints." * 



Fig. 3. 



Fig. 3. King Ravine and Mt. Adams from Randolph. Photograph by 

 Shorey. In the floor of the cirque may be seen the great pile of blocks 

 stretching forward from the foot of the headwail. 



It is true that the foliation planes and well-developed joints 

 of the mica schists of the Presidential Kange permit frost 

 and gravity to act very effectively on the steep ravine walls. 

 A study of the joints in these ravines fails, however, to dis- 

 close in them any satisfactory reason why the ravine heads 

 should have grown semicircular. The joints do not run in a 

 wide arc like that made by the cresceritic rim of the cirque ; 

 they run in the usual way, in systems which are straight for 

 long distances and which intersect obliquely, yielding polyhe- 

 dral blocks. It is hard to see why the side walls of the ravines 

 should have been eaten back by frost so fast, almost keeping 

 pace with the recession of the headwalls, for the declivity of 

 the side of the mountains is very considerable, and headwarcl 



* Geology of New Hampshire, vol. i, p. 623. 



