66 Proceedings of Indiana Academy of Science 



glacial age the Wabash waters may have found their way through this 

 valley and Atherton Island was truly insular. He says that "during the 

 climax and retreat of the Wisconsin Ice Sheet, the valleys of the Wabash 

 and its tributaries were aggraded to the present terrace level. For a 

 time the present mouth of the Raccoon was blocked by ice and the 

 stream flowed over the terrace directly to the Wabash. By final with- 

 drawal of the ice margin from the Shelbyville moraine the northern 

 outlet was left a little lower than the Rosedale divide, and the present 

 conditions came into existence". 



Geographers are agreed that the region drained by Raccoon Creek 

 was uncovered by the Wisconsin Ice Sheet while the present mouth of 

 Raccoon Creek near Montezuma was still covered with ice. The Wis- 

 consin Ice Sheet came in lobes, following the valleys in the main, until 

 a great enough thickness was attained to override the divides. As 

 the ice melted and receded to the north, it probably reformed into its 

 original lobes. Thus the river was blocked after the valley was freed. 

 McBeth' advances strong argument to prove that the whole melting of 

 the Wisconsin Sheet was from east to west. 



It is evident, therefore, that Raccoon Creek must have followed 

 its southwesterly course during the time of the Wisconsin Ice Sheet. 

 It follows that the present deflection of the stream to the north is 

 Post-Wisconsin. 



It would seem probable, however, that there was a valley formed 

 before the deflection of the stream. The present sand dune divide would 

 not force the creek over a very high saddle. This valley might have 

 been foiTned in any of a number of ways. I will indicate a few of them. 



During the waning of the Wisconsin Ice Sheet it is possible that a 

 stream found its way out along the margin of the ice into Raccoon 

 Creek. Such a stream would carry a load sufficient to build up a very 

 level plain between the ice and Raccoon Creek. 



It is possible, as Dryer sugge.sts, that some of the Wabash waters 

 at one time found their way through this channel. The present width 

 and depth of gravels would not indicate much more than overflowing- 

 back water, however. 



It is possible, of course, that the original cause of deflection was 

 an ice dam — as previously suggested. I do not believe, however, that 

 it is necessary to explain the Bridgeton terrace by this dam. The 

 Bridgeton terrace might well be considered as a part of the Terre Haute 

 terrace and explained as glacial outwash with it. 



Lastly, it is possible that the divide between a north-flowing branch 

 of Leatherwood Creek and a south-flowing branch of Raccoon Creek 

 was low enough that the waters of Raccoon Creek were forced over it, 

 as the valley filled with sand, in Post-Wisconsin times. This would 

 occur gradually, of course, probably for years only at the time of 

 high water, until a permanent channel was effected. 



I have given these theories in the order of their probability, in 



• McBeth, W. A., "Development of the Wabash Drainage System and the Recession 

 of the Ice Sheet in Indiana", Proc. Ind. Acad. Sci., 1910, pp. 184-192. 



