TRANSACTIONS OF SECTION C. 777 



8. A Map of the Eslcer Systems of Ireland. 

 By Professor W. J. Sollas, D.Sc, F.R.S. 



[Communicated by permission of the Director-General of the Geological Survey."] 



Nowhere probably can the study of Karnes or Eskers be more profitably under- 

 taken than on the Central Plain of Ireland, over which they are strewn in count- 

 less numbers. Hitherto they have been investigated rather individually than 

 collectively, though, thanks to the careful mapping of the officers of the Geological 

 Survey, the material for establishing their connection as members of great groups 

 or systems lies ready to hand. 



Much, however, is to be learnt from the individual Esker. The current bedding 

 of the masses of well-rounded pebbles and sand of which it is composed is such as 

 to point to rapid accumulation in running water, while the numerous instances of 

 irregular disturbance and ' caving in ' can be most feasibly explained by the melt- 

 ing of enclosed masses of ice. One of the most striking peculiarities of form is the 

 steepness of the sides, which frequently approaches the angle of repose of the con- 

 stituent material, and forces upon one the idea of the existence during deposition 

 of a sustaining wall, by which the running water was prevented from distributing 

 its load far and wide over the surrounding plains. Such a wall might conceivably 

 have been furnished by a previous lateral extension of the Esker itself, since removed 

 by river action, but such a supposition is unsupported by evidence. A more 

 probable suggestion is that the support was furnished by ice, and that the Esker 

 may represent a ' cast,' as it were, of a glacier tunnel in gravel and sand. On this 

 hypothesis all the known characters of Eskers find an explanation, and many inci- 

 dental details, such as the long lakelet or shallow streams by which they are not 

 unfrequently flanked. 



All explanations of Eskers depending on marine action may be summarily dis- 

 missed, for not only do they fail to afford a single parallel instance to the point, 

 but they are directly negatived by the universal absence of marine shells; of the 

 thousands of existing Irish Eskers, not one has aflbrded a fragment of a contempo- 

 raneous marine fossil, in spite of most persistent and careful search. Either, then, 

 we must admit, on the hypothesis of an Esker sea, that marine shells were absent 

 from its floor over the whole breadth of Ireland, and through a bathymetrical 

 range of 300 feet, or that having once existed they have since entirely disappeared. 

 One alternative is not more improbable than the other, as is shown by the frequent 

 occurrence of fragmentary marine shells in the sands and gravels of the Middle 

 Glacial Drift, as on Ballyedmonduff" and elsewhere at elevations of over 1,000 feet. 



The fluviatile origin of Eskers, so ably advocated by the American geologists, 

 Chamberlin, Lewis, and Wright, finds its strongest support in their relations to 

 one another as parts of a system. In the map exhibited Eskers may be traced 

 pursuing their winding, serpentine path for miles together, but at the same time 

 with a convergence which ends frequently in their joining one by one together, like 

 the tributaries of a river, to form a main stream. As with tributary rivers so 

 here, the apices of the angles at the places of junction point in one general direc- 

 tion, that of the general convergence. From individual ridges also small spurs are 

 frequently given off, usually including an acute angle, which points in the same 

 direction as those made by the main bi'anches. When, as sometimes happens, the 

 direction is reversed, signs are not wanting that this is the result of a ' loop,' such 

 as is so common in the course of imdulating streams, and of which the Shannon, as 

 it winds among the Eskers, affords instructive examples for comparison. 



Accepting the fluviatile origin of Eskers, one may deduce from their present 

 distribution that of the ancient drainage systems of the Irish glaciers. From the 

 map two systems are clearly discernible, a smaller, corresponding to the glacier of 

 .Sligo and Roscommon, and the other vastly larger, embracing the whole Central 

 Plain, with a general flow from west to east and a discharge probably by the basin 

 of the Liffey. 



