512 GEOLOGY OF OLD HAMPSHIRE COUNTY, MASS. 



Parallel with this deepening of the stream beds the unequal erosion of 

 the whole area wore back the banks of the streams in easy slopes where 

 they crossed the soft sandstones, brought out the trap ridges in sharp relief, 

 and left thus the short canyons or notches through these ridges. 



Thus on the approach of the Glacial period a surface had been reached 

 which coincided much more closely with the present cultivated surface of 

 the valley than with its present rocky floor, except along the western lateral 

 valley — that is, the broad depression Avest of the trap ranges — and north of 

 the Holyoke range, where the surface of the sandstones was then probably 

 higher than the present surface of the later sands. Then came the Glacial 

 ice, destroying the whole drainage system and removing an enormous 

 quantity of the soft sandstone. Its work was favored and localized in two 

 ways by the position of the trap ranges running down tlie middle of the 

 valley. Where these ran north and south with bold westward-facing bluffs, 

 as in the Deerfield and Mount Tom ranges, the ice coming from the north- 

 west was deflected southward and scoixred out the soft sandstone at the 

 western foot of these ranges, and where the Holyoke range runs clear 

 athwart the valley east and west the ice, by its recoil as it lifted over the 

 range, plowed out the sandstone all along its northern base down to a level 

 much lower than could have been well effected by ordinary aqueous erosion- 



Thus the river channels between the ends of the ravines in the crystal- 

 line rocks and the notches in the trap ridges were obliterated because they 

 were contained in the comparatively soft sandstone, and w^e have finally to 

 seek the reason wh}' the streams, upon the decrease of the floods which 

 accompanied the retreat of the ice, in every case found their way again 

 through their old notches instead of taking the more direct and natural 

 course down the deep western lateral valley, from which the ice had removed 

 the sandstone to so low a level. The broad river occupied then almost 

 precisely the boundaries of the earlier Triassic estuary, and the tributaries 

 entered it along the border of the western highlands. 



Across Massachusetts the great river was, indeed, rather a series of lakes 

 than a river, in that it was filled mainly from the sides by the great confluent 

 deltas of its tributaries, which were pushed out to a distance proportionate 

 to the importance of the stream that furnished each, while down its center 

 clays and fine sands were deposited in less thickness. Thus it came about 

 that the great body of sand delivered to the main stream by each tributary 

 was spread diagonally across between the western hills and the divide ranges, 



