Wooclworth.l 202 [F<^h. 7, 
Thk Crest-line of Eskers. 
The crest-line of eskers varies in elevation not only with refer- 
ence to the slope of the terrane on which the deposit rests, but 
also with reference to the base of the esker itself. Gilbert^ notes 
of eskers "that the ridge tends to equality of height rather than 
to horizontality." 
The crest-line is seldom horizontal for more than a few rods ; 
but long, gentle grades of even crest-line occur, having lengths 
varying from a few rods to nearly half a mile. Such grades are 
seldom straight lines but are commonly curved. An esker may 
increase in thickness in this manner from almost nothing to 100 
or more feet, gaining correspondingly in breadth of base. The 
Auburndale, Mass., esker illustrates this class of changes. 
Contrasted with the gently sloping crest-line are the segments 
of an esker having short, steep grades, with materials at the 
angle of repose. Where these slopes recur at short intervals, a 
karae-like surface is assumed, and not infrequently the esker 
widens out into a small kame-field. Examples of this class obtain 
in segments of most long eskers. 
According to one view, the hummocky outline of an esker 
reflects the elevations of depressions in the arch of the subglacial 
stream in which it is supposed the esker was deposited. By 
another conception, the same outline is supposed to be due to 
irregularities induced by the gradual lowering of the bed of a 
superglacial stream to the base of the ice. This last view per- 
mits several variations, and like the preceding ascribes the 
variations of the crest-line to vertical changes in the configura- 
tion, or rate of melting, of the ice-sheet. The ice-tunnel and 
ice-canon have been supposed to operate to the same end. Still 
another view explains the rise and fall of the esker crest as the 
effect of violent currents in sweeping away detritus in constricted 
passages and laying it down where the velocity of transportation 
was diminished by the enlargement of the water way. 
There is yet another way in which the elevation of the crest- 
line of an esker may be changed, and a simple experiment with 
blocks of ice so placed as to bound a channel of varying width 
may readily be conducted to demonstrate the principle I shall 
I Op. cit.,p. 87. 
