250 Tne American Oeologist. April, i897 
which the}'- were formed. Obviously, the transporting power 
was supplied by the ff>rward ruoveraent of the ice-sheet. The 
advancing ice. in ascending a rock ridge, abraded the loose 
upper strata of Galena limestone, and forced them forward in 
front of it, continuing to do so as long as the front advanced. 
As soon as a retrograde motion of the ice-margin set in, the 
deposits of angular limestone gravel remained in ridged ac- 
cumulations along the terminal line of the just ended stage 
of advance. Since they have not subsequently been overrid- 
den by the ice, we may thus accurately determine the position 
occupied by the ice-margin at the culmination of each re-ad- 
vance. Where the deposits are heavily developed and a per- 
sistent line of knolls is found to correspond with the reputed 
trend of the ice margin in that portion of the district, we 
may safely conclude that this line represents the culminating 
limit of some stage of glacial re-advance. Furthermore, upon 
projecting our supposed terminal line across one of the more 
prominent esker belts, and finding it to intersect an " area of 
special development," we may conclude that the line thus es- 
tablished is not merely hypothetical. 
The combined phenomena of the angular gravel knolls and 
the "special areas" of the esker series constitute the Kansan 
representatives, in tlie Pecatonica basin, of the terminal mo- 
raines of the Wisconsin epoch. Along these morainic or ter- 
minal lines there is often a slight thickening of the ground 
moraine, with an increase in the amount of englacial drift. 
It is evident that a more plentiful supply of material and the 
proper length of time would have sulticed to form massive 
moraines. But in the eastern portion of Stephenson county, 
where the angular gravel deposits are most numerous, the 
amount of foreign drift is small. In fact, the general absence 
of any typical moraine material in the vicinity of the broken 
limestone deposits is one of the most remarkable characteris- 
tics of their occurrence. It indicates, I believe, that the knolls 
were formed with great rapidity. Between the main lines of 
knolls, which apparently represent terminal lines of a mo- 
raine-forming period, there are smaller deposits, each of which 
may possibly be the result of a single season's W'Ork. If this 
could be proved, it would furnish data for computing the rate 
of general recession of the ice-front across the district; but a 
