276 MORRIS M. LEIGH TON 



hypothesis. The northwestward movement of the ice on this side 

 of the lobe is borne out by the striae at the quarry southwest of 

 Belvidere and by the trend of the large esker. This explanation is 

 further supported by the restriction of the morainic belts to the Bel- 

 videre lobe and their disappearance to the southeast toward the 

 central axis of the lobe. The protrusion of the Belvidere ice lobe 

 from the main Michigan lobe blocked the valley of Piscasaw Creek 

 and diverted its waters temporarily westward across the divide 

 into Beaver Creek, and later during its recession impounded the 

 waters of Rush Creek Valley in western McHenry County until 

 they poured across the divide into Piscasaw Creek Valley. An 

 inspection of the Belvidere topographic map, which shows these old 

 channels, will make it clear that their explanation would be diflScult 

 on the basis of a southwestward-moving ice. 



THE GREEN RIVER LOBE 



The drift of the Green River lobe is mostly buried by the out- 

 wash sand and gravel from the Bloomington moraine and the Late 

 Wisconsin Valley train of Rock River, but strips of this drift are 

 exposed on the north and west sides and patches in the western part 

 (see Fig. 2). The boundary of this lobe on the north side is definite 

 for some 3 miles west of Eldena and 10 miles to the east and north, 

 but less so to the west. On the south side the boundary is fairly 

 definite,^ The western limits cannot be definitely mapped, due to 

 the patchy character of the drift, the heavy deposits of loess and 

 sand on the upland, and the sand and gravel outwash on the low- 

 lands; but there is reason to think that the ice blocked the old course 

 of the Mississippi River in northeastern Rock Island County and 

 caused its diversion across the rock divide south of Cordova. The 

 evidence for this will be considered later. 



The strip along the north side. — From a point near the north line 

 of Lee County and about 10 miles west of the east line, a strip of 

 relatively fresh drift up to six miles wide extends across the upland 

 to the southwest to a point south of Dixon, thence north of west 

 into the area northeast and north of Sterling, and thence south of 



' This portion of the boundary is much like that drawn by Leverett for the south 

 side of the lowan drift, on Plate XII U.S. Geological Survey Monograph XXXVIII; 

 otherwise the present mapping is in contrast to that referred to. 



