FLOOD DEPOSITS IN MONTAOUE I'.ASIN. (515 



is, outside the lake-l)Ott(im deposit. In the first case the order is, (l)lake 

 bench, (2) hike bottom, (3) 310-foot terrace (f*); in the second, (1) lake 

 bench, (2) the continuation south of t*, (3) lake bottom. 



In other words, the Hadley Lake continued through the whole period, 

 and its lake-bottom beds are, strictly speaking-, a little later in age than 

 those of the northern basin. It seems by far best to represent all the lake- 

 bottom deposits by one color, as I have done. 



DETAILED DESCRIPTION OF THE FLOOD DEPOSITS IN THE MONTAGUE BASIN. 



This description is in continuation of the interesting jiccount of the 

 terraces of the Connecticut in New Hampshire given by Mr. Warren 

 Upham in the Geology of New Ham})shire, Vol. Ill, page 19. I may men- 

 tion that I accept the criticism of Professor Dana^ of the view taken by Mr. 

 Upham, that the deltas thrust out into the main valley are often above the 

 highest "normal" terrace of the flood time, and consider these deltas as 

 marking by their levels the true height of the flood waters, and look upon 

 the lower level of the highest terrace whicli connects tliese deltas as 

 explained by a lack of material in the intermediate spaces. I can not, 

 however, accept the other criticism of Professor Dana that the esker traced 

 down the valley by Mr. Upham has no existence as an earlier structure 

 antedating the flood gravels of the open valley. 



The Montague basin is narrow — about a mile wide — where it enters the 

 Warwick quadrangle in Vernon (PI. XXXV, C), and it retains this width 

 across the area, connecting at the highest water stand westward around 

 Mount Hermon with the northern or Grreenfleld lobe of the Hadley Lake. 

 As it enters the Greenfleld quadrangle at iMillers Falls it widens to above 6 

 miles, and is connected again at flood level by several narrow passes in the 

 trap ridge with the northern lobe of the Hadley Lake at Greenfield. It 

 connects by the narrows at Sunderland with the main Hadley basin. It 

 was a nearly filled-up lake. The main stream quite filled its rather narrow 

 valley down to Millers Falls, where it widens, and here the heavy contribu- 

 tions of the Millers River filled the whole widened valley. The distinction 

 between the shore flats (1 s h) filled to the highest effectivevlevel of the 



'Am. .Jour. Sei., 3(1 series, Vol. XXII, p. 451. 



