﻿466 SMITHSONIAN MISCELLANEOUS COLLECTIONS [VOL. 47 



at the center. For a short distance the bands may be traced, upon 

 the Lefroy side, into crevasses which are swinging- from a transverse 

 into a longitudinal position. Some 23 of these bands may be made 

 out (pi. lxhi), becoming more crowded together and more sharply 

 bent before they are obliterated by surface melting. The explanation 

 of this feature as given by Tyndall seems to be the correct one. As 

 the ice flowing down a relatively gentle incline suddenly changes 

 to one sufficiently steep, there is developed a system of transverse 

 crevasses, because of the inability of the ice to yield suddenly with- 

 out rupture. During the summer when the motion is most rapid 

 there is produced either a single such crevasse or a series of parallel 

 crevasses, more or less closely approximated. The sun rounds off 

 the edges of the single crevasse and cuts rapidly into the exposed 

 walls which separate neighboring ones (pi. lxiv, b). At the bottom 

 of the slope, or upon the slope itself, when the crevasses are healed, 

 there still remains a more or less well defined depression, into which 

 atmospheric dust and the debris from the surface of the glacier may 

 collect. During the winter there will be a fewer number of crevasses 

 formed and either from the weaker solar action, or the protection 

 afforded by the snow, the edges are not appreciably affected and the 

 crevasses later are perfectly healed. The result of this action is 

 that there is formed each year a depression and a ridge, each of 

 which becomes more and more curved because of the more rapid 

 central motion of the ice. Where the incline is steeper the motion 

 is greater and the successive bands are farther apart, but as the ice 

 becomes more sluggish upon the gentler slope the bands become 

 more crowded. Each season, as the result of melting and rains. 

 more and more material is washed into the depressions, so that they 

 become better defined up to a certain point when, as a result of 

 melting the surface acquires a uniform slope, the dirt becomes more 

 uniformly distributed and the bands are obliterated. The method 

 of formation here described is beautifully shown in the Deville 

 glacier (fig. 72, a), situated in the Selkirks and here reproduced 

 through the courtesy of Air. Arthur ( ). Wheeler and the Canadian 

 Topographic Survey. 



Especial interest and importance is to be attached to this type of 

 " dirt band," since it gives a clue to the actual, as well as relative, 

 forward motion of the ice and a certain insight into the nature of 

 this motion. Standing upon the Victoria it is possible to recognize 

 about three-fourths of the bands which may be counted from a 

 distance, and to determine approximately the successive intervals. 

 Starting with the one at the base of the ice slope and measuring 



