110 ILLINOIS ACADEMY OF SCIENCE 



In all of these ways occasional pebbles are now being scat- 

 tered on the surface of, and becoming imbedded in, the loess 

 accumulating upon the Wisconsin till, but each of these agen- 

 cies gradually becomes less effective and will eventually fail 

 to scatter any pebbles, as the thickness of the loess increases. 

 It is believed that by means similar to the above the pebbles be- 

 came incorporated in the lower part of the loess deposit as it 

 was in process of accumulation over the main loess covered 

 regions. 



RELIEF OF THE PRE-LOESSIAL SURFACE 



The relief of the surface of the Illinoian drift sheet at the 

 time the loess was laid down upon it was worked out in detail 

 in the Avon and Canton quadrangles in the northcentral part 

 of the State where the loess is well developed. This was de- 

 termined by finding the elevation of the upper surface of the 

 drift, and thus of the lower surface of the loess at a large 

 number of places where the contact was exposed as shown on 

 the map, (plate 2). The relief of the surface of the drift be- 

 neath the loess in the Canton quadrangle was found to exceed 

 150 feet, and the general slope of the top of the drift beneath 

 the loess in this region was found to correspond in a general 

 way to the slope of the present surface. The main divides of 

 the region at the time the loess was deposited were the same as 

 at present. In the Canton quadrangle the main highland 

 which then, as now, lay between Farmington and Norris, ex- 

 tended a little south of west past the town of Fairview. From 

 this highland the pre-loessiaj surface declined southward 

 along Copperas Creek as much as 100 feet in a distance of 

 6 miles. Along Big Creek, from its sources in the vicinity of 

 Norris to the town of St. David, the total difference in the 

 elevation of the drift surface beneath the loess exceeds 125 

 feet. From the headwaters of Put Creek to the place where it 

 leaves the quadrangle near the southwest corner, the surface 

 declines more than 100 feet. A like difference in the altitude 

 of the upper surface of the drift is shown between the head 

 waters of Turkey Creek and Coal Creek and the top of the drift 

 bordering these streams at the west side of the quadrangle. 

 From the sources of Littlers Creek to its junction with Spaan 

 River the pre-loessial drift surface declined 150 feet. 



The topography of the Illinoian drift surface beneath the 

 loess in the Avon quadrangle resembles that of the Canton 

 quadrangle described above, as does also the pre-loessial sur- 

 face of the drift in the Sumner and Hardinville quadrangles, 

 in Southeastern Illinois, and the Herrin quadrangle in the 

 southern part of the state. 



