358 2he Ainei'ican Geologist. D.-iomber, iso^ 
On some other ureas, and perliaps nioi'e eonnnonly. di-unilins 
may have been formed from the enulaeuil stratum of drift dni'uig 
a time of re-udvanoe of tlie ice-sheet, carrying the diift forward 
so tliat it would be accumulated on a land sui'face. This appears 
to have l)een the case in central New Vork, whei'e Prof. W. M. 
Davis finds that the sections of drumlins fre(iuently show strati- 
lied gravel autl sand underlying the till, and that often the rela- 
tionship of these formations is such as to pi'ovethat the stratified 
beds were somewhat eroded before the deposition of the till. But 
the absence of such sections in Massachusetts and generally in 
New England makes it probable, as stati'd, that here the recession 
of the ice was continued, though with a much <liminished rate, 
while the drumlins were being amassed. 
Rcrii'ir iif ()/ij(c/ii,iis t(i f/iis Hxiihinnfiini. — At first sight, this 
explanation of the accumulation of the drumlins api)ears to Ite 
opposed by two conspicuous objections, which must be answered. 
The first is the local derivation of much of their material. Where 
the peculiarities and restricted limits of the adjacent rock forma- 
tions ou the north i)ermit an approximate determination of the 
distances of transportation of the drift forming the drumlins, it is 
found that a large part, sometimes more than half, has been car- 
ried only a few miles. It seems surprising that local material 
should constitute so important an element of the drift contained 
within the ice at consideral)le hights. until we consider how fast 
it would be uplifted l)y even a very slight upward inclination of the 
basal current of the ice. If the drift eroded from any place was 
carried up with an av(U'age ascent of only one degree, it would 
rise, within one mile, to an altitude of !)2 feet aV)ove the ground, 
and within two to three miles would l)e as high as the tops of the 
most prominent drumlins. Currents ascejiding at this rate, or 
even two or three degrees or more, may very probably have ex- 
isted in the lowest part of the ice-sheet, on account of the accel- 
eration of its upper I'urrents, within distances from 20 to 5(1 miles 
or more back from its boundary. Hy these currents much drift 
eroded from the land surface would be gradually incorporated in 
the conij)aratively sluggish lower part of the ice, reaching alti- 
tudes 100 to 1,000 feet al)ove the ground within a few miles from 
its sources. 
When the boundary receded, the upper currents of the outer 
belt of the ice, ujion a width of probably ten miles, would pour 
