170 PROCEEDINGS OF THE ACADEMY OF [1885. 



junction of two streams (ancient or modern). These deposits do 

 not show an anticlinal structure, and instead of occurring in the 

 centre of a valley, are seen at points where an eddy in the drift- 

 laden waters, or more commonly the shelter of a projecting rock, 

 has allowed them to be formed. They may be termed eddy-ridges 

 or terrace-deltas, and are clearly of entirely different origin from 

 kames. The writer has seen a number of such ridges in northern 

 Pennsylvania, and there is little difficulty in distinguishing them 

 from kames. 



That stratified deposits ma}' exist beneath a glacier, undisturbed 

 by the weight or motion of the ice, has been shown by observa- 

 tion at the base of the Swiss glaciers. Among the most interesting 

 of these observations, because so well correlated with the phe- 

 nomena of the American ice sheet, are those of Prof. Chamberlin, 

 made at the base of the Rhone glacier. He remarks : J "At other 

 points, near the centre of the valley, the ice may be seen resting 

 directly upon well-assorted stratified sand and gravel. Level 

 sheets of fine detrital matter extend without disturbance of con- 

 tinuit} r or surface beneath the edge of the glacier. The assorting 

 and stratification of this material was apparently accomplished 

 by subglacial streams, which seem afterwards to have found 

 other avenues, when the ice occupied their place either by 

 settling down from above, or advancing from behind. The 

 singular fact is that the stratified sands should not have been 

 disturbed." 



The similarity between the contours of kames and of moraines 

 is accounted for on the supposition that both were moulded 

 beneath the ice sheet. While the moraine shows the edge of the 

 glacier, the kames indicate the direction of its drainage. Flat 

 terrace plains, such as that at Berwick, on the Susquehanna, 

 are made by floods issuing from the front of the glacier, but when 

 the drainage was subglacial, whether forward or backward, kames 

 are the result. 



Siil>glacial Drainage. — The most important conclusion arrived 

 at by a study of the backward-draining marginal kames is that 

 of a great subglacial drainage. The same conclusion is drawn 

 from observations on the terminal moraine itself, which in many 

 places shows evidence of having been drained northward, not 

 southward. Many sucli instances might be described. 



1 Wis. Geol. Survey Report for 1878, p. 17. 



