''ROCK GLACIERS'' OR CHRYSTOCRENES 553 



its contained fragments of ro:k, not only moves outward from the 

 hills or mountains and down the valleys, but that when the upper 

 parts of the ice melt under the influence of the warm weather in the 

 spring, the rocks which were contained in it slide downward and form 

 the concentric ridges which are so well shown in one of Mr. Capps's 

 photographs. 



The clear ice which constitutes the matrix for the mass of ice and 

 stones in the talus is certainly not formed by the freezing of water 

 during the warm weather of summer, when surface water is abundant, 

 when streams are flowing in all the valleys, and when the temperature 

 of the air often rises to 70° or even 80° F. It is equally certain that 

 this icy matrix is formed during the winter, when, in the Klondike 

 district, as well as in the interior of Alaska, the surface drainage is 

 completely arrested, and the small streams are frozen to the bottom, 

 and when the only water which reaches the surface is derived from 

 springs with moderately deep-seated sources, though this spring 

 water is quickly frozen into ice when it comes within the influence of 

 the extremely low temperature of the overlying atmosphere. 



The positions of these springs may readily be determined in winter 

 time by the presence of chrystocrenes, which form more or less con- 

 spicuous mounds of ice elevated above the general snowy surface, 

 but in the summer they are often very difficult to detect, for it may be 

 impossible to distinguish between water derived from springs and 

 that from surface drainage. 



Therefore, the presence of the springs, and their possible signifi- 

 cance in supplying ice to the talus heaps in order to enable these to 

 move downward and outward more easily, would probably escape 

 notice if investigations had been conducted exclusively in summer. 

 Investigations conducted during the winter would probably prove 

 the existence of springs behind these "rock glaciers," in which case 

 they might be considered as chrystocrenes filled with broken fragments 

 of rock. 



