404 PROCEEDINGS : BOSTON SOCIETY NATURAL HISTORY. 
r 
more that the topographic relations of eskers are, practically without 
exception, more accordant with the snperglacial than the suhglacial 
hypothesis.' The very usual haphazard relation of an esker to the 
contours of the bed rock and till is seen to be entirely normal for a 
superglacial deposit, while from the subglacial point of view, it is a 
perpetual enigma. Besides the general indifference of their trends 
to the modern drainage, which has perhaj^s been sufficiently consid¬ 
ered, we have the fact that they are not, as a rule, distinctly con¬ 
trasted in size or in the coarseness of the gravel on northern and 
southern slopes. The subglacial stream was comparable with the 
flow of water in a pipe, and the velocity for any given head was 
inversely proportional to the diameter of the tunnel, and independ¬ 
ent of the local gradient, the velocity being the same on northern or 
ascending as on southern or descending slopes. But, obviously, the 
tendency to aggrading of the stream bed is far greater on northern 
than on southern slopes, in spite of the uniform velocity; and there¬ 
fore the esker of subglacial origin should be stronger, more perfectly 
aggraded, and composed of much coarser material on the up than on 
the down slopes. This contrast may, perhaps, be noted occasionally, 
but it is by no means so marked or general as the hypothesis requires. 
Eskers belong chiefly to the moderately dissected peneplain tracts 
and are not specially characteristic of deep mountain valleys where, 
unquestionably, the conditions were most favorable to crevassing 
and a concentration of the glacial drainage beneath the ice. Again, 
the usual lateral position of valley eskers is entirely normal for the 
superglacial and seemingly inexplicable for the subglacial hypothe¬ 
sis. To cite a single instance, in the village of Bridgewater, Nova 
Scotia, on the southwest side of the La Have river, the valley of 
which is here some three hundred feet in depth, a well-formed 
esker trends approximately parallel with the valley and at an eleva¬ 
tion of about one hundred feet above the river. That a subglacial 
stream of water could have hung on this steep slope is well-nigh 
inconceivable and contrary, as previously noted, to all our observa¬ 
tions on existing glacial streams. The fact, that, in general, eskers 
trend towards cols or depressions in water partings, calls for no 
special comment, since it is entirely consistent with both hypotheses. 
Relations of eskers to the ground moraine. —Eskers, as well as 
their terminal plains, normally overlie the ground moraine or till, 
the only important exception being when, locally, the till is wanting 
