THE SPRINGFIELD LAKE. 659 



up did not require nearly so much material as did the much deeper Hadley 

 Lake; (2) the great multitude of" elevations make the border of the lake on 

 the east very complicated. 



The steep eastern border of the valle}' is notched for the entrance of 

 only one tributary which heads l)ack of the tirst series of ridges. This is 

 the Chicopee River, whicli at the village of Three Rivers gathers all the 

 di'ainage of eastern Hampshire and Hampden counties. In the time of the 

 lake it earned certainly a much greater volume of water than at present, 

 and as the ramifications of this di'ainage cover the whole broad area of high- 

 level glacial lakes already descril^ed, their abundant sands furnished an 

 enormous volume of ah-eady sorted detritus, which is now spread in the 

 broad sand flats of South Hadley, Chicopee, Spring-field, and Longmeadow. 



The study of the basin brought to my attention several most inter- 

 esting problems, and it has been difficult to exjiress upon the map the 

 results reached. An inspection of the map will show that I have there 

 represented the lake, in contradistinction to the two northern lakes, as a 

 nearly filled-up lake. In the former, passing across the high terrace flat 

 toward the center of the valley, one comes upon a well-mai'ked scai-p of 

 deposition, or delta front, which descends to the lower plane of the lake 

 bottom. Here one goes out from the head of the Chicopee River delta, 

 264 feet above sea, and crosses the broad, gradually sloping sand plains to 

 their inner edge ovei'looking the river meadows at 240 feet, or going south 

 across Chicopee, Springfield, and Longmeadow, finds the level sinking from 

 255 feet in the north to 180 feet in the south; and yet the whole gi-eat 

 sand body, the larg-est on the river, covering a large poi-tion of four towns, 

 expands as a great, extremely flat "alluvial cone" or delta, with imper- 

 ceptible slope from the mouth of the gorge of the Chicopee at Collins 

 depot to where it is cut off" by the later erosion of the river, and shows 

 nowhere any scaq) which could justify one in separating- the central and 

 lower portion as lake bottom fi-om the higher and shoreward portion as 

 lake shore. 



A fm-ther inspection of the map will show that in the northeni portion 

 of the lake basin the deposits referred to the high terrace (1 s h, PI. 

 XXXV, D) or filled-up portion of the lake are shaped rudely like a coml), 

 with its back stretching along south of the Holyoke range and its teeth 

 extending- south between the long ridges of sandstone and till across 



