SoLLAS — A Map to show the Distribution of EsJcers in Ireland. 795 



obviously and naturally unknown to him. He was also unaware of the work of 

 Hummel and Hoist, this being then very recent, and moreover published in the 

 Swedish language. Great credit further is due to Upham, owing to the fact that 

 the phenomena are not so simple in New Hampshire as elsewhere, the eskers in 

 that State being buried to a great extent beneath plains of alluvial deposit. I 

 give this author's theory in his own words : — " Through the whole glacial period 

 rivers probably existed, to some extent, beneath the ice ; but their volume and 

 strength . . . greatly increased during the final melting of the ice-sheet. In many 

 instances these probably filled deep channels along lines of depression upon the 

 surface of the glacier, which, at the last part of the melting, would coincide nearly 

 with our present valleys ; or of tener, probably, they formed for themselves great 

 tunnels beneath the ice, seeking, of course, the lowest land for their route. By 

 these glacial rivers, which flowed beneath, or on, the ice, discharging the water 

 supplied by its melting, there were deposited, from the low water of winter, layers 

 of sand, and from the strong currents of summer layers of gi'avel, often very coarse, 

 which would be very irregularly bedded, here sand and there gravel accumulating, 

 and, without much order, interstratlfied with each other. At the melting of the 

 ice-walls on each side of these glacial rivers, the materials, which had collected to 

 a great depth in their channels, were left in long ridges, a section of which would 

 show an irregular anticlinal stratification. . . . Mounds or irregular short ridges 

 . . . have probably resulted from the rapid deposition of these rivers as the last of 

 the glacier was melting away, ridges and irregular masses of ice having existed 

 where there are now hollows or ponds. These irregular accumulations of sand and 

 gravel were many years ago attributed to similar causes by Dr. Edward Hitchcock, 

 who accordingly named them moraine terraces. The occurrence of occasional 

 angular boulders enclosed within, or lying on, the surface of kames is readily 

 explained, as the course of the rivers was probably in most cases beneath the 

 glacier, and they would be dropped from the melting of the ice overhead. From 

 explorations in the Alps and Greenland we know that streams in summer are 

 found flowing on the surface of glaciers, and that falling through crevasses these 

 gather to form considerable rivers beneath the ice." 



In the year 1877, the second edition of " The Great Ice Age," by Professor 

 James Geikie, was published. It contains a notice of Hoist's views, and an 

 account of Hummel's theory, which the author, abandoning the marine theory, 

 accepts. 



Warren Upham* published a full and complete account of the eskers of New 

 Hampshire in the following year. In this he considerably modifies the views he 

 had at first expressed, so that they now more closely resemble those of Hoist than 



* The Geology of New Hampshire, 1878. 



6U2 



