96 W. UPHAM — THE SUCCESSION OF PLEISTOCENE FORMATIONS. 



Lake Agassiz and its Deltas. 



From the most southern part of the Nelson river basin, near the middle 

 of the west side of Minnesota, the glacial lake A^-assiz extended north- 

 ward across the entire width of this basin, a distance of about 700 miles ; 

 for the waning ice-sheet forming its northern boundary had doubtless 

 receded to the watershed dividing the Nelson and Churchill rivers north 

 of lake Winnipeg before the latest stage of lake Agassiz was finally per- 

 mitted to outflow along the present course of the Nelson river to Hudson 

 bay. This glacial lake, growing from south to north as the barrier of the 

 continental glacier receded, attained an area of not less than 110,000 

 square miles, exceeding the combined areas of the great lakes outflowing 

 by the Saint Lawrence ; and its depth above lake Winnipeg was about 

 600 feet. As noted in the introduction of this paper, the retreat of the 

 ice-sheet from this basin appears to have occupied no longer time than 

 1,000 years; but four large moraines are found to have been formed at 

 stages in-the glacial recession contemporaneous w^ith the extension of 

 lake Agassiz from its mouth and southern end north along only a third 

 part of its length. The marginal morainic drift is thus shown to have 

 been accumulated with a surprising rapidity, which seems to have been 

 possible only by its derivation chiefly from englacial and superglacial drift. 



Six extensive deltas of lake Agassiz are found in the southern, prairie 

 portion of the ancient lake area, which has been fully explored, wdth exact 

 mapping of its beaches and deltas, and with determinations of their 

 heights by leveling.* Each of these deltas consists largely of modified 

 drift supplied from the melting ice-sheet during its retreat past the tracts 

 which they occupy ; and one of them, brought wholly by a glacial river, 

 lies where the lake area now receives no important inflowing stream. 

 They all belong to the time of the upper Herman beach of the lake, the 

 earliest and highest in a series of about thirty well defined beaches re- 

 cording successive levels of the lake due to erosion of its outlet, to changes 

 of outlets, and to differential uplifting of the lake basin so that single 

 beaches at its southern end are represented by two, three, or several, in 

 their continuation northward. 



Preglacial epeirogenic Elevation. 



Reviewing these Pleistocene formations of the great central river basins 

 of the United States and Canada, I agree with Professor Hilgard in his 

 interpretation of the Lafayette gravel and sand beds as river deposits 



*Geol. and Nat. Hist. Survey of Minnesota, Eighth An. Rep., for 1879, pp. 84-87 ; Eleventh An. 

 Rep., for 1882, pp. 137-153, with map; and Final Report, vols, i and ii. U. S. Geol. Survey, Bulletin 

 no. 39, pp. 84, with map. Geol. and Nat. Hist. Survey of Canada, An. Rep., new serie.-^, vol. iv, for 

 l888-'89, part E, pp. 15G, with two maps, and sections. 



