WARREN UPHAM DISTRIBUTION OF ENGLACIAL DRIFT. 147 



find their only explanation, as to form and origin, in the drainage system of a 

 melting sheet of land-ice. The greater part of the modified drift, however, was 

 laid down outside the receding ice-margin, and occupies the avenues of drainage 

 from the ice to the sea ; and where the ice-border lay in or near the sea, or a great 

 lake, deltas of gravel and sand were formed and the finer silt was distributed more 

 widely and thinly by coastal currents. 



Influence of adjoining Lakes or tiie Sea. 



From Nantucket and ( 'ape Cod northeastward the ice-sheet at its greatest extent 

 and during a considerable part of its time of recession terminated in the ocean. 

 In the interior of the continent, too, it was bounded during its recession by vast 

 glacial lakes, filling the basins that are now partly occupied by the great lakes of 

 the St. Lawrence, Nelson, and Mackenzie rivers. During six summers of field-work 

 in examination of the shore lines, deltas, and bed of Lake Agassiz, the largest of 

 these glacial lakes, I have carefully studied the effects attributable to the influence 

 of this lake in the deposition of the drift, comparing its area, the valley of the Red 

 river of the North, with other portions of Minnesota, South Dakota, North Dakota, 

 and Manitoba, which were land surfaces during the departure of the ice. Other 

 glacial lakes of smaller size in these states and Canadian province have also come 

 under my observation, besides portions of the drift deposited in the glacial precur- 

 sors of the Laurentian lakes ; and on the Atlantic coast I have made a detailed 

 examination of the marine drift of southeastern New Hampshire. The more 

 southern parts of the New England seaboard which I have similarly examined, in- 

 cluding the coast from Boston to Plymouth, Cape Cod, Nantucket, Marthas Vine- 

 yard, the Elizabeth islands, Block island, and Long island, appear to me to have 

 stood at their present height or somewhat higher during the maximum extension 

 and the recession of the last ice-sheet. 



Upon all the areas thus studied by me where the ice-sheet was bordered by great 

 lakes or the sea, tracts of stratified sediments, as deltas of gravel, sand, and silt, and 

 somewhat more extensive deposits of finer silt and clay, are found ; and their dis- 

 tribution shows them to have been chiefly brought into these bodies of water by 

 rivers flowing down from the melting ice. But a huge portion of the englacial 

 drift, corresponding to that which elsewhere fell as wholly unstratitied till on land 

 areas, was received from the receding ice into these lakes or the sea with Little 

 change, being allowed to fall to their bottom only very slightly modified by water 

 action. Within the area of bake A.gassiz and the other associated glacial lakes very 

 extensive tracts, probably half or a larger part of their whole extent, have a surface 

 of till which differs from its characters on the adjoining tracts that were land dur- 

 ing the iee retreat in having usually slight traces of stratification within the five to 

 fifteen feel of the upper and englacial till, and in bavin-- a more smooth and even 

 contour. 



Bowlders, gravel, sand, and clay are mingled in this englacial till in the same 

 proportions as on the country outside these glacial lakes. There was generally no 

 noteworthy transportation ofbowlders or other drifl by ice floes or bergs on these 

 lakes ; nor was the line clayey pari of the englacial drifl washed away in any note- 

 worthy amount from the submerged and melting and receding ice-margin by wave 

 action, which would have covered the i ill in fronl of the ice-sheel with beds of silt, 

 [nstead, the englacial an. I finally Buperglacial drifl that escaped the stream erosion 



