328 PLEISTOCENE OF INDIANA AND MICHIGAN. 



LAKE DULUTH. 



Finally came Lake Duluth, with its outlet southward through the Brule and St. Croix 

 valleys. Lake Duluth endured for a much longer time than the earlier lakes and its lower levels 

 expanded to a lake of large size, forming beaches found even on the northern part of the Keweenaw 

 Peninsula. Its outlet was cut down about 40 feet, lowering the level of the lake and causing 

 the formation of later beaches below the highest. 



Several lower beaches, whichhave been traced by Mr. Leverett (p. 431 ) , seem to lie too low for 

 waters discharging through the St. Croix outlet and hence may not belong to Lake Duluth. 

 These suggest an outlet somewhere to the east and south around the hills south of Marquette, 

 the exact place being not yet determined. These beaches are rather weak and suggest the 

 possibility of submergence in consequence of a readvance of the ice. 



GLACIAL LAKES IN THE SLTERIOR-MICHIGAN-HURON BASIN. 

 LAKE ALGONQUIN. 



Stages of Lake Algonquin. — In the outline given above the succession of lakes in each 

 of the upper three basins — those of Lakes Huron, Michigan, and Superior — were given down 

 to the time when the glacial waters in all three were about to merge into one great lake. This 

 larger body is called Lake Algonquin and its upper beach is one of the strongest and most 

 persistent shore lines in the Great Lakes region. At its greatest extent this lake covered an 

 area considerably larger than all three of the upper Great Lakes of the present time. 



"When the lake waters fell to lower levels from Lake Warren and Lake Lundy they 

 uncovered for the first time the lowlands between Lakes Huron and Erie. This lowland divided 

 the waters into two separate lakes and inaugurated the flow of St. Clair and Detroit rivers. 



Lake Algonquin may be divided into four stages: (1) Early Lake Algonquin, confined 

 to the south part of the Lake Huron basin ; outlet at Port Huron. (2) Kirkfield stage, covered all 

 of the upper lakes except the northern part of Georgian Bay and probably the northeastern 

 part of Lake Superior; outlet at Kirkfield, Ontario; uplift in north begins. (3) Port Huron- 

 Chicago stage; outlets at Chicago and Port Huron, but diminishing at Chicago and increasing 

 at Port Huron; great uplifting, producing divergence of beaches northward; three groups of 

 beaches, Upper Algonquin, Battlefield, and Fort Brady; most rapid uplift during Battlefield 

 group. (4) Closing transition stage leading to Nipissing Great Lakes; outlet eastward to the 

 Ottawa Valley. 



Early Lake Algonquin. — For a relatively brief stage immediately following the time of 

 Lake Lundy the waters in the south half of the Lake Huron basin formed an independent 

 lake. This has been called Early Lake Algonquin. It was at the beginning of this stage that 

 St. Clair and Detroit rivers first came into existence, and the outlet during this time was through 

 these rivers to Lake Erie. The ice front then rested against the highlands which bounded 

 the two sides of the southern half of Lake Huron, as shown by the moraines in both Michigan 

 and Ontario. This stage, however, was relatively short, for the ice front stood so far north 

 at its beginning that a slight additional retreat opened passages to the east and northwest 

 where lower outlets were available — to the east to Georgian Bay and Trent Valley in Ontario 

 and to the northwest to Lake Chicago. These two passages probably opened nearly at the 

 same time. From the fact that no separate beach or outlet is known for this stage it might be 

 thought that it is wholly hypothetical and its existence entirely uncertain. But, besides the 

 local consequences of the relation of the ice sheet to the highlands, the early distributaries 

 of the St. Clair and Detroit rivers seem inexplicable except as incidents of the transition from 

 Lake Lundy to a lower stage with southward outlet corresponding to Early Lake Algonquin. 

 And, further, the five short gorges made by Niagara Kiver in the Niagara escarpment during 

 its early flow show conclusively that the river then had a volume as large or perhaps larger 

 than at present, and this it could not have had unless it received a large contribution of water 

 to Lake Erie from the north. This was clearly before the opening of the Kirkfield outlet in 

 Ontario. The discharge from Early Lake Algonquin was larger than might be expected, because 



