Woodworth.J 204 [Keb. 7, 
The crest-line of some eskers agrees closely in elevation with 
the height of neighboring extraglacial deiDOsits, n)oraine terraces, 
sand plaini!', and kame-fields. The exceptions often indicate a 
failure to attain the elevation reached by frontal and marginal 
deposits. Examples of this relation are to be seen at Auburndale, 
Mass., Pawtucket, and Providence, R. I. In a channel open to 
the sky, the limiting condition to the accumulation of detritus 
would accord with that of ordinary streams. When tlie ice 
melted away, the gravitative distribution of material on the sides 
might affect narrow and high segments of the deposit. Where 
this prmciple is not involved, the limit of construction must lie in 
the ice-sheet itself, as in the presence of an arch. 
The second class of variations in crest-line corresponding with 
the relief of the terrane traversed by an esker is commonly 
observed and constitutes one of the most puzzling features. 
Eskers in southern New England tend to lie in nearly north- 
south valleys ; but they occasionally are found crossing east-west 
depressions, often undergoing a change of elevation of nearly 100 
feet. According to the view which maintains the superglacial 
origin of eskers, these last examples, like the first, originally filled 
channels in the ice at an elevation equal to the summits of the 
adjacent divides, the present position of the ridge being due to 
settling through the ice sheet, something like the passage of a 
wire loaded with weights through a block of ice. The objections 
to this view arise mainly from the frequent uniform preservation 
of the esker type throughout a segment having this distribution.! 
Such a change of position may be compared to the stretching of 
a linear mass from a chord to its ai-c. The eskers of this descrip- 
tion, so far as I have been able to see their structure, do not 
exhibit evidences of this extension. The Cape Ann example is 
less steei»-sided and well developed than the typical eskers of 
north-south valleys. I am not sufficiently acquainted with other 
examples to attempt to show their identity with those eskers 
which I believe to be of truly subglacial origin. 
The Direction of Eskers. 
Eskers tend to meander. Their sinuous and often tortuous 
course led Professor Shaler to name them "serpent kames." 
» For other objections, see Salisbury, Geol. surv. of N. J., Ann. rept. for 1892, p. 42. 
