entirely antecedent to what follows, it should not be inferred that 
such was its relation to the study of sand plateaus and their feeding 
eskers in the field. The two divisions of the study advanced to- 
gether, and one part was continually employed to check the other. 
In this way it is thought that a fuller understanding of both 
was secured than if field work had been pursued, as some would 
advise, before a general knowledge of the studies of others was 
gained. The reasons for such advice appear to be based on a fear 
that if the conclusions and theories of others are present to the 
mind while field study is going on, the observer may warp his 
facts into line with preconceived opinions, and fail to discover the 
mistakes of his predecessors as well as to appreciate the real 
meaning of the new facts of his own observation. I believe that 
a sufficient counter to this objection is to be found in careful 
training. If the danger that truly exists is kept in mind and 
clearly realized, and the mind is required to take a serious and 
judicial attitude, avoiding hasty conclusions and testing every 
hypothesis, then the advantage of the double method of investiga- 
tion is certainly greater than its disadvantage. Indeed, those who 
are too poorly trained to be trusted with it had best not under- 
take independent study at all. 
5. THE PROBLEM OF THIS ESSAY. 
The special subject indicated by the title of this essay is one of 
those smaller problems of glacial geology to which a considerable 
attention has been directed by observers of the minuter forms 
of drift deposits for some years past. Although a relatively tri- 
fling matter in itself, it is one of those details whose right under- 
standing contributes a valued share to the correct interpretation 
of the processes of the glacial period as a whole. 
Some have advocated that the long ridges of gravel, variously 
called kames, serpentine kames, Indian ridges, osars and eskers — 
the latter being the name now adopted by Chamberlin and others 
— were formed in channels underneath the ice during the latter 
stages of the disappearance of the ice-sheet ; others believe them 
to have been deposited in channels on the surface of the ice, from 
which they have settled down to the ground as the ice afterwards 
melted away. There are difficulties attending either view. If 
formed in subglacial channels it must be admitted that the gravels 
