239 
[Upham. 
placements. This seam of sand was doubtless quickly deposited 
by running subglacial water ; and it was thenceforward undisturbed 
while the deposition of the till continued over its whole extent. 
While the till both next below and next above, to a thickness va- 
rying from a few inches to two or three feet, but exceeding one 
foot only beneath the sand layer, was being deposited, sufficient per- 
colation of water took place to produce an imperfect stratification, 
the clay commonly enclosing nearly the ordinary proportion of 
gravel and stones in the till. Very generally, however, this clay is 
finely laminated and free from gravel for two or three inches next 
below the sand and for a half inch to one inch above it ; and in one 
place it was seen thus stratified and free from gravel for about a 
foot or a little more below the sand. Boulders and smaller rock 
fragments, occurring in contact with the sand layer, or in the asso- 
ciated more or less stratified part of the till, are enclosed by lam- 
inae which bend down beneath and arch upward over them. Such 
seams of sand or gravel enclosed in till are often sources of water 
in wells ; and a fine spring, situated only a few rods northwest from 
the Winter Hill station, probably issues from this porous layer. 
No other similar sand layer has been found during my examination 
of many extensive sections of the till forming drumlins in the vi- 
cinity of Boston ; but seams of modified drift have been observed 
rarely in lenticular hills and slopes of till in New Hampshire, and 
they are frequently encountered in digging wells in the undulating 
drift-sheet of Minnesota, where in many districts they have so great 
extent as to supply artesian water . 1 
Besides the evidence in the structural features of the drumlins, 
indicating that they were accumulated rapidly during the closing 
stage of the last glacial epoch, a strong argument toward the same 
conclusion is afforded by the prevailing east-southeast trend of the 
longer axes of these Iqlls about Boston, while the striation on the 
bed-rocks is mostly south-southeast, the difference between the two 
courses being forty-five degrees. Elsewhere in all the districts char- 
acterized by abundant drumlins in this country and the British Isles, 
their longer axes and the glacial striae are parallel ; and it seems 
sure that both were determined by the currents of the ice -sheet. 
Their difference in direction in the neighborhood of Boston, I be- 
lieve to be due to a deflection of the motion of the ice there during 
1 Geology of Minnesota, Eighth annual report, for 1879, pp. 113, 114; Final report, 
vols. i and ii. 
