408 PROCEEDINGS : BOSTON SOCIETY NATURAL HISTORY. 
Stone’s monograph is the most complete contribution j^et made 
to the natural history of eskers, and no one has discussed the theory 
of eskers more fully and impartially. In fact, his work is particu¬ 
larly notable for the judicial and fair-minded attitude toward the 
rival hypotheses which it reveals. Although holding, with Chamber¬ 
lin, Davis, and others, that the subglacial hypothesis affords the best 
explanation of a certain ideal type, he is disposed, as we have seen, 
to refer the wide eskers, the branching eskers, the reticulated eskers, 
the unstratified eskers, and perhaps others, to the agency of super¬ 
glacial streams. These concessions are certainly sufficient to give 
the superglacial hypothesis a good standing ; and the chief objection 
urged against a still broader application of this hypothesis is the 
supposed prevalence of crevasses in the marginal zone of the ice 
sheet, a supposition which is, in my opinion, essentially groundless. 
Source of the material of eskers and their terminal plains .— The 
all-important question here is as to whether the material was derived 
chiefly from the englacial or the subglacial drift, and if from the 
former, whether it was supplied to superglacial streams through 
superficial ablation or to subglacial streams through the erosive 
action of the stream itself on the roof and sides of its tunnel. 
Assuming, in view of the preceding discussions, that the englacial 
drift was sufficiently abundant and extended to a sufficient height 
in the ice to meet the requirements of esker and plain formation 
through the agency of superglacial streams, and recalling that all 
the drift set free by superficial ablation is virtually within easy reach 
of the superglacial streams and available for the aggrading of their 
beds and deltas, we may now give our attention particularly to the 
subglacial streams. 
The subglacial drainage could derive but little detritus from the 
englacial drift without such an enlargement of the tunnel as would 
cause its collapse, and it is, therefore, practically limited to the sub¬ 
glacial drift or ground moraine, for crevasses are probably wanting; 
and even if they were not, to depend upon detritus washed into 
them from the surface of the ice sheet would be, as we have already 
noted, to grant a greater volume and height of englacial drift than 
subglacialists have heretofore been willing to allow. What oppor¬ 
tunity has the subglacial stream to erode the ground moraine ? On 
the first obstruction of its mouth, by standing water or terminal 
deposit, it must begin to aggrade its bed, and erosion here almost 
