548 GEOLOGY OF OLD HAMPSHIRE COUNTY, MASS. 



three later glacial deposits with as many intervening sands, which could be 

 followed for 3,350 feet in the open cutting. 



Several of these hills of most regvilar shape are so built up on the 

 steeply sloping rocky valley side (the valley runs here north and south) 

 that looking vxj) from below one seems to have before one a di-umlin of the 

 largest size, while looking down from higher up the hillside one sees only 

 a small ridge interrupting the eastward slope. These hills are directed 

 southerly, as are the neighboring striae, but they lie near the western 

 boundary, between the areas of southeast and of southern motion of the 

 ice, like a line of bars between two currents meeting from different direc- 

 tions; and in Northampton, where the ice was deflected in the vaUey 

 southwestward, the drumlins have the same direction and swing in a great 

 curve across the reentrant angle in the rocky border. 



In the town of Amherst nearly every hill is a drumlin, and in several 

 cases they are laid side by side in pairs and coalesce laterally. It is fur- 

 ther interesting that this group of drumlins in Amherst runs right up to the 

 steep northern base of the Holyoke range, which here traverses both the 

 valley and the direction of the ice, and whose crest of trap is finely covered 

 with north-south striae. 



I have mentioned above that tlie drumlin exposed in the Camp 

 Meeting cutting, a little higher in the valley, is covered by three sepa- 

 rate glacial beds, representing, doubtless, as many oscillations in the ice 

 at the time of its retreat, which shows — what, indeed, hardly admitted 

 of doubt — that the di-umlins were formed beneath the thick ice of the 

 general glaciation; and the position of this last group, carried with north- 

 south axes right up to the foot of the steep Holyoke range, which itself 

 is striated in the same direction, bears strong evidence against the exist- 

 ence of a separate Comiecticut River glacier which should explain the 

 north-south striation of this valley. Indeed, these north-south drumlins 

 are carried up so high on the sides of the valley that when one imagines 

 ice of the smallest thickness needful to build them and compress them to 

 their present rock-like density, one sees that the ice would have risen 

 above the boundaries of the valley and have overflowed fan-like, as in the 

 great lobes found in the Western States. The facts seem, then, to accord 

 better with the theory ])roposed by Professor Dana of a differential 

 motion of the lower portion of the ice in the valley, and the long line 



