CHANGES OF DRAINAGE IN SOUTHWESTERN INDIANA. 101 



ing this lowland it continues westward and passes through a range of hills 

 which leads northward from Princeton past Patoka to White River at Hazel- 

 ton. This deflection lies within the glaciated region and calls for a some- 

 what different explanation from those just considered. It is probable that 

 a lower passage was afforded across a col near Patoka than the lowland to 

 the north else the stream would not have suffered this deflection. In the pas- 

 sage through this range of hills at Patoka the valley is reduced to a width 

 of scarcely one-half mile or to about one-fourth the width of the preglacial 

 valley which it occupies a few miles to the east. Upon passing this range 

 the stream soon enters the broad valley of the Wabash and takes a some- 

 what direct course into that river. 



Since Patoka River enters the glaciated region in its lower course the 

 question arises whether it found discharge beneath the ice margin during 

 the time when its lower course was covered by the ice sheet or found a line 

 or lines of discharge southward through the unglaciated region into the Ohio. 

 There is a col on the divide between Patoka and Ohio rivers crossed by the 

 abandoned Wabash and Ohio Canal about 1 mile southwest of Francisco, 

 which stands only 480 feet above tide, or about the altitude of the silt 

 deposits along the Patoka in western Dubois and Pike counties. From this 

 col there is an open line into the head Avaters of Pigeon Creek. But as 

 noted above there may have been another col in northern Warrick County 

 on the present line of Pigeon Creek to offer -resistance to the discharge of 

 water to the Ohio. It is doubtful, however, if that col stood any higher 

 than the one near Francisco, and it may have been so low as to afford an 

 easy passage for the stream. An examination of the col near Francisco 

 failed to bring to light decisive evidence that a stream had crossed it. The 

 sag crossing the divide is scarcely 100 rods in width and is not bordered by 

 banks or well-defined erosion contours. The features seem to be no differ- 

 ent from those of other sags at higher altitudes on the divide. The canal 

 cuts through about 12 feet of loesslike yellowish-brown earth before strik- 

 ing the rock surface, which is similar to the material found on border dis- 

 tricts at all altitudes. It seems less probable that cols on the Patoka-Ohio 

 divide farther east were utilized, for they stand considerably higher than 

 the col under consideration; the lowest stand probably about 525 to 550 

 feet above tide. The col crossed by the railway near Ferdinand station, 7 

 miles south of Huntingburg, is probably as low as any, standing 530 feet 



