CKOSBY: ORIGIN OF ESKERS. 
397 
drift due to and in turn determining unequal ablation, as well as 
the natural tendency of streams to meander with diminishing 
gradient. In brief, the trend of the superglacial esker is entirely 
consistent with the known facts and the reasonable probabilities of 
the case; and it matters not, so far as this feature is concerned, 
/ whether the ice was absolutely stagnant at the time the esker was 
formed or still retained a part or the whole of its velocity when in 
its prime. 
On the other hand, the only factors which may be considered as 
directly influencing the direction of subglacial streams and eskers are 
the ground topography, which is often contrary or indifferent; cre¬ 
vasses, which, in so far as they exist at all, must be chiefly transverse 
and therefore indifferent; and the movement of the ice, or the differ¬ 
ential pressure of the ice upon the ground resulting from its move¬ 
ment. That the basal movement of a sheet of ice thick enough to 
flow over a deeply and irregularly corrugated surface of very low 
average gradient (the gradient being often opposite or transverse to 
the movement) would tend to develop and maintain furrows or tun¬ 
nels in the bottom of the ice is extremely doubtful. Tunnels meet¬ 
ing the requirements of esker formation would not usually lie in the 
lee of prominent topographic reliefs — ledges or hills, which might, 
conceivably, have grooved the bottom of the ice as it flowed past 
them. Nor is it easy to see how tunnels due in any direct way to 
ice flow and pressure could have been so indifferent to the topography 
as eskers are often observed to be, forsaking broad valleys hundreds 
of feet in depth to continue, perhaps in a very oblique or closely 
parallel course, across an uneven upland. As previously noted, 
there must be, as long as the ice is in motion, a tendency to equalize 
the basal pressures and prevent the development of open spaces by 
flow of the ice and also by flow, in obedience to the ice pressure, of 
the underlying drift or ground moraine ; and, furthermore, the effects 
of unequal pressure would be, at least to some degree, neutralized by 
the differential or localized melting induced by the pressure through 
the lowering of the melting point. Again, in their longer reaches 
as well as in their minor meanders, eskers are often obliquely or 
directly transverse to the last recorded ice movement, and to that 
extent the movement of the ice must have tended strongly to oblit¬ 
erate or sweep away the deposits of the subglacial streams as fast 
as they were formed. Stone says (’99, p. 426), “The longer mean- 
