244 The American Geologist. April, i897 
because of more or less transverse ice-movement. Under this 
liypolhesis, the individual ridges of a given section of the belt 
were not strictly contemporaneous, although nearly so. Evi- 
dence strongly supporting this proposition may be derived by 
an examination of the internal structure of the deposits. 
Where several ridges adjoin, the deposits of one will be found 
to overlap on those of a neighboring ridge, and to be in turn 
overlain b}'' those of another. 
In summarizing the evidence which is furnished by the 
phenomena connected with the stratified sand and gravel 
beds of this district, we find that it clearly proves (1) 
that they owe their existence to subglacial streams which 
flowed at so low a level as to derive the material very largely 
by direct erosion; (2) that these streams flowed westwardly 
into extra-glacial lakes; and (3) that the main period of ex- 
cessive melting occurred some time after the Kansan ice-sheet 
had begun to waste away or its front to "retreat." It less 
clearly shows (4) that the deposits as now seen are those 
which were laid down along the last few miles of the tunnels, 
and more particularly at or just beyond their mouths; and (5) 
that the " areas of special development" Indicate the position 
of a terminal line during a halt in the recession of the ice- 
front. The terminal lines here referred to apparently corres- 
pond to the terminal moraines of a later glacial epoch. It is 
evident that, if we are able to discriminate the several "special 
areas" in the different belts, and to correlate them, we can 
read the history of the moraineless drift of northwestern Illi- 
nois almost as readily and completely as that of the districts 
abounding in large and distinct moraines. However, to en- 
deavor to locate these terminal lines by means of the stratified, 
drift alone is rather a questionable proceeding; but if the 
lines sp indicated are supported by evidence of other classes, 
their positions may be considered as demonstrated. Under the 
next heading I shall endeavor to indicate the nature of this 
corroborative evidence as displayed in this district. 
Eskers are considered to have been formed by streams which 
flowed through or under the ice-sheet, generall}"- parallel to its 
direction of movement. To this class I would refer nine-tenths 
of the stratified drift ridges of this district. Kames are con- 
sidered to have been formed by streams which flowed down 
