1890.] 
117 
[Crosby. 
lowest part of the area ; while the coarser material is usually de- 
posited much nearer its source and hence somewhat uniformly along 
all the drainage lines of the district. This general fact that the 
coarse silt is local while, in this region, the fine silt is gathered from 
a large and indefinite area and deposited mainly in the sea and al- 
most wholly below sea-level, makes a direct quantitative compari- 
son quite out of the question. Another point that should be 
considered in this connection is that some clay must have been re- 
moved from the general mass of the till, and from much till which 
hasneverbeen thoroughly assorted or modified, by subglacial waters, 
before the final melting of the ice-sheet ; for the constant grinding 
and kneading of the ground moraine in the presence of water would 
necessarily result in the washing out of considerable clay wherever 
there was any subglacial drainage. Although it thus appears im- 
possible to obtain a strictly unmodified till, the product of ice action 
alone, and equally impossible to determine even approximately 
the extent of the clay beds corresponding to the stratified sand and 
gravel of any given area ; it has yet seemed worth while to test the 
validity of the general impression that clay is a principal constitu- 
ent of the till as we now know it. 
The primary object of the examinations of the till detailed in the 
following pages was thus to determine whether the observed facts 
concerning the relative abundance of the coarse and fine modified 
drift are in harmony with the constitution of the till from which 
both were derived ; but it soon became apparent that they might 
throw important light upon the problem of glacial erosion, possibly 
enabling us to settle the vexed question as to whether the till itself 
is the product chiefly of the mechanical action of the ice-sheet and 
its accompanying torrents, or of quiet chemical erosion in pregla- 
cial times. As the investigation advanced there also appeared to 
be some ground for the hope that we might find here criteria that 
would aid in the distinction of glacial and non-glacial silts, even 
in the earlier geological formations. It is needless to add that these 
are now regarded as by fai the most important and interesting as- 
pects o/ this study. 
SOURCE OF THE MATERIAL ANALYZED. 
All of the till for the following analyses was collected with con- 
siderable care from the drumlins of the Boston Basin, and always 
at points where deep artificial or natural excavations had laid bare 
