SoLLAs — A Map to show the Distribution of Eskers in Ireland. 793 



esker -whicli winds along through the same great area of gravel. Since the water- 

 worn material diminishes from S.W. to N.E., it is clear that the water flowed in 

 this direction, while the glacial striae show that the movement of the ice was from 

 N.E. to S.W. The existing river now flows in the same direction as that which 

 carried the gravel. 



Hummel's famous Paper was followed by one equally well-known bv 

 Dr. Hoist.* A short account (by Dr. Lindahl) of this Paper was published 

 in the American Naturalist, vol. xxii, 1888, pp. 589 and 705. The author com- 

 mences with a powerful criticism on the marine and mud-plain theories, which 

 he has little difficulty in demolishing. He rightly points out that the mud-plain 

 theory had, at least, the merit of perceiving that the esker required sustaining 

 walls for its formation, and that these have since disappeared, and that it 

 recognised the resemblance of an esker to a river deposit. [One may remark 

 that the same material, viz. mud, was at one time invoked to do other duty, 

 which has since been assigned to ice, such as transpoi'ting erratics and polishing 

 and striating rock surfaces. J Hoist fully recognises the merit of Hummel's theory 

 so far as it depends on the jDresence of the inland ice, but he thinks it less fortunate 

 in its details, particularly in assigning the origin of the eskers to tunnels below the 

 ice. The formation of the subglacial tunnels, as described by Hummel, affords a 

 fair opportunity for a damaging attack ; but since the observations of Russel on 

 the Malaspina Glacier, the whole treatment of this part of the subject, both by 

 Hoist and Hummel, appears speculative and artificial. 



Dr. Hoist then proceeds to explain his own views. On the melting of the 

 inland ice, streams form upon its surface, which work back from the margin of the 

 ice until they form ice-canons ; at the same time englacial drift (which is assumed 

 to exist in considerable quantity) is set free, worn by running water into rounded 

 gravel, and sorted and arranged in layers on the bed of the stream. Thus the 

 englacial fragments, which lay near the upper surface of the ice, form tlie lower 

 layers of the esker. 



On the final melting of the ice the deposits in the river-bed remain behind as 

 an esker. Nordenskiold's description of a broad and deep river flowing on the 

 surface of the Greenland ice is appealed to ; and it is pointed out that the size of 

 this river points to its having flowed for some distance before arriving at the place 

 where it was observed. This river, a little lower down, certainly precipitates 

 itself through the ice and disappears ; but since crevasses depend on the move- 

 ment of the ice, when the ice was melting away and becoming stationary, they 

 would diminish and might almost disappear. Or if crevasses occur in the ice, 

 they must be in union with one another and with the free margin of the ice ; 



* Om de glaciala rulhtensdsa/rne, by N. 0. Hoist, Geol. Forens. i Stockholm Forhandlmgar, Bd. III. 

 p. 97, 1876 and 1877. 



lEANS, KOT. DUB. SOC, N.S. TOL. V., PAST XIH, 5 J^ 



