PRE-GLACIAL COXDITIONS. 509 



50 feet above the rivev a great talus of trap fraginents — a pre-Glacial 

 "Devils Garden," as these desolate slopes were called by the fathers — 

 resting against the vertical wall of trap, which here rises about 100 feet 

 above the level of the stream. The talus was exposed for a. length of 90 

 feet and for a height of 30 feet, and it apparently extends down to the 

 level of the river, 50 feet below, but this was not observed. Covering this 

 talus and extending up over the trap was a layer of very coni[)act till, 30 

 feet thick, of reddish color, made up mostly of sandstone with few liowlders 

 of mica-schist from the western hills and with none of trap. At least nlue- 

 tenths of the bowlders, down to those not above 2 inches on a sido, were 

 finely striated — a quite unusual proportion. 



A fresh vertical section of this till produce<l by caving was marked for 

 a long distance by wavy lines of apparent Ijedding so perfect that at a dis- 

 tance I had supposed the beds to be the thin-laminated Champlain clays, 

 but the lamination proved to be an unusually ])erfect pressure cleavage in 

 the till, in planes dipping 60° to 70° NW., at right angles to the direction 

 from wdiich the ice was moving iu the canyon, as marked bv the strife 

 upon the trap immediately above. These data prove that the ice breasting 

 the long westward-facing vertical wall of the Deei-field trap range was 

 pressed into this notch in the range with exceptional force, from which we 

 may deduce that the prevalent southwai-d motion of the ice in the valley 

 was due to its deflection from the normal northeast direction by the north- 

 south walls of the valley and of the divide ranges. 



A further interesting deduction is that the notch of the Deerfield range 

 is, in its present form, of pre-Glacial origin, and since the river flows 

 through without exposing rock at bottom the gorge was then of even greater 

 depth than at present. The Deerfield Indians affirmed that it was begun 

 by a squaw with a clam shell. 



One other deposit, probably of Tertiary age, is described with the 

 "Camp Meeting cutting," near end of Chapter XIX. It is a thoroughly 

 sorted, ])ink beach sand, and it appears below the glacial beds on the north 

 line of Northampton. 



PRE-GLACIAL WEATHERING. 



The only important case of the preservation of any portion of the 

 deeply decomposed surface rocks which must have characterized the 

 country before the advent of the ice, as they are now characteristic of 



