J. E. Todd — Revision of the Moraines of Minnesota. 473 



the south longer because it is at a lower level. About Detroit 

 City and White Earth Agency the drainage is distinctly east- 

 ward through the western moraine and down the valley south- 

 ward. This is indicated by the character of the deposits, which 

 are of a sandy nature and lying in terraces. 



Curving westward from the strongly developed north and 

 south moraines upon the divide between the Mississippi and 

 Red River basins, is a short spur south of the Wild Rice River 

 in Norman County and ending abruptly along the Upper 

 Herman beach. A similar westward branch of the moraine 

 turns westward between Hill and Lost Rivers in eastern Polk 

 County. This ends abruptly at a beach marking a lower stage 

 of Lake Agassiz. This is on Section 31, T. 150-40. More- 

 over, a broad channel with, terraces extends from a gap in this 

 moraine sonthwestward to the beach, indicating- a discharge of 

 a stream apparently from the southern edge of the lobe occu- 

 pying the Red Lake area into Lake Agassiz. The course of 

 Clearwater and Red Lake Rivers, together with a certain sub- 

 dued unevenness of a belt extending northwest from this point, 

 seems to indicate the location of the edge of the ice in Lake 

 Agassiz during the formation of the moraine. 



Having expressed thus definitely our dissatisfaction with the 

 former interpretation, we proceed to outline one that seems to 

 avoid these difficulties. 



During the occupation of this region by the ice sheet, we 

 conceive that it moved mainly in two great lobes; one coming 

 from the norcheast through the valley of Lake Superior, the 

 other from the northeast and north in the valley of Red Lake 

 and Red River. This latter divided so as to pass down both 

 the James River in Dakota and the Minnesota River. During 

 the Wisconsin stage the two principal lobes mentioned were 

 probably confluent over the whole of Minnesota, except the 

 southeast comer. This was indicated by Mr. Upham in the 

 9th annual report of the Minnesota Survey, and was followed 

 with some modification by Professor Chamberlain in the 3d 

 annual report of the LTnited States Geological Survey. The 

 main modification was that he conceived the lobes to be more 

 separated by the divide along the Mesabi Range. As the ice 

 receded and stood at intervals to form the different moraines, 

 there was the growth of a long interlobular moraine, first 

 northwest from Minneapolis to the Leaf Hills, thence north 

 and later northeast, following approximately the present divide 

 between the two main valleys, so as to finally connect with the 

 Mesabi moraine. This would evidently derive its material 

 from the western side of the Lake Superior lobe and the 

 eastern side of the western lobe. As different stages were 

 reached in the recession, this interlobular portion would 



