THE SHELBY VILLE MORAINE. 211 



the Shelbyville sheet down as far as Springfield the terraces are composed 

 of a gravelly sand with a few large pebbles. It is not certain, however, 

 that these terraces should be correlated with the Shelbyville drift sheet, for 

 they seem to be quite as well developed above the point where the river 

 leaves the moraine as they do below that point. As indicated farther on, 

 these deposits may be correlated with the Cerro Gordo moraine. At the 

 point where the river crosses the Shelbyville moraine, south of Decatur, 

 and also south of Harristown, the ten-ace remnants are found to lie in a 

 valley cut to considerable depth into the Shelbyville sheet, and no over- 

 wash graA^els were found at higher levels. It should be stated, however, 

 that, owing to the absence of wagon roads following along the valley and 

 the inclemency of the weather at the time the writer was there, the valley 

 was examined only at the two road crossings south of Decatur and south of 

 Harristown. Further study at intermediate points or below Harristown 

 may bring to light gravels which can be correlated definitely with the 

 Shelbyville moraine. 



Along the valley of Salt Creek, the first stream of importance issuing 

 from the Shelbyville moraine north of the Sangamon River, a light deposit 

 of gravel was found immediately outside the moraine on the road between 

 Kenney and Hallsville, but the main deposits of the terraces along the 

 valley are sand. This stream, like the Kaskaskia and Embarras, seems, 

 therefore, to have carried only a weak discharge at the time the Shelbyville 

 moraine was forming. 



Two tributaries of Salt Creek — Kickapoo Creek and Sugar Creek — 

 which cross the Shelbyville moraine farther north than the main creek, 

 have their sources in the Bloomington morainic system and carry ten-aces 

 whose head is found in that moraine. The valleys were not given sufficient 

 study to enable the writer to determine whether they also have terraces 

 which may be connected with the Shelbyville moraine. If such terraces 

 are present, they are far less conspicuous than those which head in the 

 Bloomington moraine. 



On the Mackinaw River a terrace has been traced into connection with 

 the Bloomington drift sheet, but none has been found which can be corre- 

 lated with the Shelbyville. A deposit of" loess several feet thick caps the 

 Shelbyville moraine in the interval between the Mackinaw and Illinois 



