96 THE ILLINOIS GLACIAL LOBE. 



deposits of sand and .silt that seem referable to the valley filling at the 

 Illinoian stage of glaciation and which antedate the loess by a long period, 

 as shown by weathering. The surface of the sand presents a deep red stain 

 to a depth of 3 or 4 feet below the base of the loess, and contrasts 

 strongly in color and weathering with the sand at greater depth as well as 

 with the overlying loess. The weathered zone here is apparently the cor- 

 relative of the black earth found below the loess at the "Yellow banks." 

 Excellent exposures of this weathered zone may be seen at the corner of 

 Second and Timea streets in Keokuk. 



Examinations have been extended down the Mississippi on the Illinois 

 side, and it is found that the altitude of the valley filling decreases more 

 rapidly than the fall of the present stream. At a point opposite Hannibal, 

 Missouri, where Hadley Creek enters the valley from the east, the filling 

 reaches a level only 15 or 20 feet above the broad bottom of the Mississippi 

 or scarcely 35 feet above the stream, and about 90 feet below its altitude 

 at Warsaw, 45 miles up the valley. This rapid increase in the amount of 

 filling apparently supports the view that material was swept into the valley 

 and there deposited in delta-like fashion. 



Returning to the discussion of the abandoned channel, and taking up 

 the portion northeastward from the Iowa River, it is found that it has 

 slightly lower altitude than the portion in the district south from the Iowa 

 River, much of it being below 700 feet above tide. In explanation of this 

 lower altitude it is suggested that the section to the north of the Wapsipin- 

 icon River and possibly the portion between the Wapsipinicon and Iowa 

 rivers may have been occupied by the Mississippi for a considerable period 

 after the southern portion had been abandoned. Possibly it persisted in the 

 occupancy of its channel until the Iowan ice invasion forced it out. 



The broad valley of the southwestward flowing portion of the Cedar 

 appears to have held a lake at the Illinoian stage of glaciation for which 

 Udden has recently suggested the name "Lake Calvin." The features of 

 this old lake bed are discussed by Udden in a report on the geology of 

 Muscatine County, Iowa. 1 



The abandonment of the lower end of the channel from Columbus 

 Junction southward probably occurred as soon as the ice sheet had with- 

 drawn sufficiently to uncover the present line of the stream, for the altitude 



1 Iowa Geol. Survey, Vol. IX, 1899, pp. 350-357. 



