796 SoLLAS — A Map to show the Distribution of Eskers in Ireland. 



of Hummel. Thus, while admitting that in some cases eskers may have been formed 

 by sub-glacial streams, he considers that, in general, the streams were super-glacial. 

 He thinks that the deposition of the eskers took place chiefly at the mouths of 

 these rivers, and extended along the valleys as fast as the ice-front retreated ; and 

 finally he no longer regards the large angular erratics, occasionally associated with 

 eskers, as having been left behind by a melting glacier, but as dropped from 

 floating ice. 



In the same year Professor Newberry* describes the kames of Ohio as differing 

 in some respects from typical eskers. " They occupy a topographical position 

 which makes it impossible that they should have been the beds of rivers ; for they 

 form a belt along the summit of the divide between the lake-basin and the Ohio 

 valley, all the way across the State" ; and he concludes that they "have been 

 formed by the action of water on the morainic material of the Erie clay " ; but 

 " how the necessary floods could be produced, there, is a difficult problem." 



A valuable contribution to the study of eskers was published by Professor 

 G. H. Stone in 1881. f Briefly summarised, Mr. Stone's results are as follows: — 

 The eskers of Maine are found at all elevations above the sea up to a height of 

 about 1600 feet. They freely cross transverse hills 100 feet in height, even when 

 a little deflection would give them a course through a valley. None surmount 

 hills of over 200 feet. No instance is known where a kame, after entering a 

 valley bordered by hills more than 200 feet high, has left it, however crooked the 

 valley may be. The only surface features which seem to invariably determine the 

 location of eskers are drainage basins and hills more than about 200 feet high, 

 measured on the north. In many instances eskers cross the beds of lakes, some- 

 times for many miles, and can be traced under water. 



The observations of the author on the transport of the material of the eskers 

 are so important that I give the following extract in full : — " With the exception 

 of one short outcrop of granite the course of system ix. lies wholly within a region 

 of slates, conglomerates, and Limestones, for nearly 100 miles. It then crosses the 

 granitic range of mountains extending north-eastward from Mount Desert. For 

 seven miles after entering the granite area in Aurora the kame is composed 

 almost wholl}^ of slate, although the country on both sides is covered by granitic 

 till, and shows multitudes of granite boulders and fragments. The granite then 

 begins to appear in the kame, and, a few miles farther south-eastward, the kame 

 plains are almost wholly composed of granite, although the underljang rock is 

 there a micaceous slate or schist, and the till shows that kind of rock freely. 

 These and other facts of similar import show that the kame materials have been 



* Rep. Geol. Surv. Ohio, toI. iii., p. 40, 1878. 



\ The Kames or Eskers of Maine, by George H. Stone. — Proc. Amer. Assoc. Adv. Sci., 1880 (1881), 

 p. 510 ; and On the Kames of Maine. — Proc. Boston See. Nat. Hist., 1880 (1881), p. 430. 



