88 The American Geologist. February, i904. 
ern slope of this part of the moraine has become so much cut 
up is probably to be found in the greater thickness of the ice- 
sheet at this point, caused by the stowing of the ice, which, 
maintaining a slow but strong southern course, suddenly met 
with the barrier. During successive melting phases it dismissed 
mighty torrents upon the southern highlands, thus giving ori- 
gin to the now widened valleys of the Brule creeks. 
Besides the outlets there is another characteristic feature of 
the moraine : the basins and ponds, which usually occupy the 
inner slope, but also to some extent the back of the northern 
member of the moraine. The basins, which were once filled by 
the waters of the retreating ice sheet and w^iich may have been 
at that time deeper than they are now, are as a rule small and 
shallow, covering rarely more than a square-mile of surface, 
and being partially occupied by ponds, which dry out during a 
time of lasting drought or hold more or less water throughout 
the year. They are less developed on the bottom of the trough 
and near "ts axis, and the circumstance that they are not at all 
or at least very rarely met with on the surface and the outer 
and inner slopes of the "Canton ridge," is readily explained by 
the great declivity of the inner and the torn-up condition of the 
outer slope ; while the flat back of the northern member as well 
as the less pronounced inclination of its inner slope furnished 
more favorable conditions for the formation of these basins. 
Another feature, which is not directly connected wnth the 
moraine, as it rather indicates a phase of the retreat of the 
glacier after the moraine had become accumulated, but which 
might be taken up in this connection, is the druinlins met with 
in the area under discussion. They are low hills, not exceeding 
lo to 15 feet in hight, of more or less oval shape and composed 
of gravels, sands and the "bowlder-clay" of the drift, their long 
axis lying m the direction of the drift. They are in most cases 
the accumulation-points of erratics. These are of all possible 
kinds, although the composition of this material varies. Two 
of them have been observed by the writer, not very distant from 
each other, one within and the other without the northern 
boundary of Lincoln county, within T.ioiN., R.51W., an imag- 
inary line connecting them and coinciding with their axes, run- 
ning N.W.-S.E. Another one lies north of the road that leads 
one mile south of Tea to Harrisburg, within a fourth of a mile 
