120 PLEISTOCENE OF INDIANA AND MICHIGAN. 



Knightstown, wells show 20 to 50 feet of gravel. In most places a heavy deposit of blue till 

 underlies the gravel, but in some rock underlies the gravel. The wells in the valley near New- 

 castle penetrate about 400 feet of drift, of which only a few feet is surface gravel. 



The Sugar Creek gravel plain, which heads in northeastern Johnson and western Shelby 

 counties, appears to have been an outlet for glacial waters at the time the morainic system was 

 forming. 



WHITE RIVER BASIN. 



The large gravel plain of White River does not head in the White River valley but in the Fall 

 Creek valley, White River being confined in a comparatively narrow valley above Crown Hill in 

 north Indianapolis. The valley of Fall Creek begins to expand near Millersville, about 5 miles 

 above Indianapolis, and within 3 miles its width increases from one-fourth mile to 2^ miles and 

 where it enters White River is fully 3 miles. Below Indianapolis this plain maintains a width of 

 2 to 4 miles as far as Martinsville, in Morgan County, beyond which it shortly becomes narrower 

 in its passage through the resistant limestone formations. 



Above Indianapolis the width of the White River valley is about one-half mile except at 

 some of the bends of the stream, where it is a mile or more. The abrupt enlargement of the 

 valley just above Indianapolis is near the line of the inner edge of the morainic system and seems 

 explainable as the work of a great stream issuing from the ice border. This stream must have 

 started under the ice sheet 3 or 4 miles above the moraine. 



From Indianapolis southward to the Johnson County line the moraine is aproned on the 

 west by a gravel plain with basins and other irregularities along the line of junction. That this 

 gravel plain, which stands several feet above a broad terrace of White River, is the product of 

 glacial waters and is not a valley deposit made by» White River is shown not only by its close 

 relationship to the moraine but also by its perceptible descent from the moraine riverward. 

 The material appears also to be coarser next the moraine than nearer the river. The western 

 portion of the plain has been removed by White River, but the eastern portion remains intact, 

 with basins along its eastern edge about the morainic knolls. In this portion the excavations 

 reveal comparatively little sand and a large amount of gravel, largely of limestone, like the peb- 

 bles of the till. 



West of White River, between it and Eagle Creek, is a plain which in places has what 

 appears to be outwash gravels from the moraine. The gravel, however, does not entirely cover 

 the surface, for places were observed where unmodified till crops out at the surface of the plain. 



WALNUT CREEK BASIN. 



In eastern Hendricks and southern Boone counties the outer border district is a till plain, 

 upon the most level parts of which silty clays nearly free from pebbles range in thickness from 

 a few inches to several feet. It is in this district that channels appear which have already been 

 mentioned as probable lines of glacial or subglacial drainage. The largest channel heads about 

 2 miles southwest of Lebanon and extends southwestward down one of the headwaters of 

 Walnut Creek for 7 or 8 miles, having too little descent to give the stream a good current. The 

 channel at the north end is about a mile wide, but within 2 miles it becomes narrowed to one- 

 fourth mile and maintains this width downstream for many miles, to enlarge only with the 

 increasing size of the creek. Near its head it is depressed only 10 to 15 feet below the general 

 level of the bordering plain, but it here occupies a boggy tract which was no doubt deeper when 

 first abandoned. Its bluffs are steep like the banks of a river. This channel is along the west- 

 ernmost of the Boone County streams that unite to form Walnut Creek. Two other tributaries 

 farther east are characterized by similar channels. The middle of the three passes just west of 

 the village of Milledgeville, below which it is bordered by knolls that rise a few feet above 

 the general upland level, suggesting that both the channel and the knolls may be the product 

 of a subglacial stream. 



