658 GEOLOGY OF OLD HAMPSHIRE COU2^TY, MASS. 



middle of its length for the passage of the Connecticut and AVestfield rivers, 

 bounds on the north and west the bed of a third lake, which extended south 

 across the borders of the State to the point where the river enters the nar- 

 rows at Middletown. Its eastern boundary was the high rocky border of the 

 valle}' across the towns of Belchertowu, Granby, Ludlow, and Wilbraham. 



The two notches mentioned above were naiTOw straits connecting this 

 lake with the Hadley Lake, and only a very small portion of the sands and 

 gravels which now fill the latter came through these notches from the 

 northern lake. This was especially true of the northern passage, for over 

 a broad area in front of it the bottom of the Hadley Lake was filled up 

 only a few feet above the present level of the meadows, and that with lami- 

 nated clays capped by fine sands, while immediately south of the gorge the 

 sands are coarse and are built up to a plane 100 feet higher. 



Of course, the narrow Holyoke range on the north and west never fur- 

 nished any considerable triliutary to the lake after the ice had disappeared 

 from its north and west slope, but the ice melted away south of the moun- 

 tain much more rapidly then it did north, and there was a long time when 

 bodies of water gathered upon the ice in the northern area and swept 

 through the notches in the Holyoke range, carrjdng much sand and gravel 

 into the southern basin. I have already traced the watercourse south from 

 the Pelham liasin through the passageway between the east end of the 

 range and the crystalline rocks, and tlirough the notch next west, occupied 

 by the "Bay roacL" (See p. 588.) 



The position of the Holyoke diabase ridges detailed above had great 

 influence on the action of the ice. It plowed very deeply into the sand- 

 stones north of the main ridge in the Hadley Lake, but to the south it left 

 the sandstones over much of the basin above the level afterward maintained 

 by the lake. It seemed also to plow in long grooves, leaving several 

 parallel ridges of sandstone, which stretch south across Granby and South 

 Hadley. These ridges may he in part the uptilted western rims of the 

 great fault blocks of the sandstone. It deposited also many great drumlins 

 on and in prolongation of these ridges. Farther south also, across Hampden 

 County, a broad marginal portion of the basin is occupied by low ridges of 

 sandstone and till, which rise, for the most part, but little above the level 

 of the lake sands. This had two results of importance in the history of the 

 lake: (1) So much of the lake was from the beginning shallow that its filling 



