372 THE ICE AGE IN NORTH AMERICA. 



plex action of forces often difficult to unravel, and peculiarly 

 liable, in a case like this, to lead one astray. Hence it is 

 probable that a part of the testimony in favor of the Cincin- 

 nati ice-dam, drawn at first from the terraces of the upper 

 Ohio, was irrelevant, since some of the terraces were doubt- 

 less due to the natural progress of river-erosion. For ex- 

 ample, where streams flow down from the flanks of a mount- 

 ain-chain, and have considerable opportunity to deepen their 

 channels, they leave gravel deposits at various levels, and oc- 

 casionally abandon some portion of their old bed to occupy 

 a shorter and deeper cut-off. Upon examination of the 

 Monongahela and its branches, it would seem that several of 

 the terraces at first attributed to the effect of the glacial dam 

 at Cincinnati were formed in the manner thus indicated, or 

 may have been so formed. In this case their remarkable 

 correspondence with the height of the supposed obstruction 

 at Cincinnati is one of those accidental coincidences that are 

 often met with in nature. After eliminating, however, all 

 such cases, there remains a constantly increasing residuum 

 of facts which fairly refuse to be* explained, except on the 

 theory supposed. 



Beginning with one of the clearest cases, attention is di- 

 rected to the head-waters of Brush Creek, in Pike county, 

 Ohio. By reference to the accompanying map it will be 

 seen that at the height of the Glacial period the front of the 

 ice rested at the northwest corner of this county. A more 

 accurate study of the details shows that the divide between 

 Paint Creek and Baker's Fork of Brush Creek is formed by 

 the extreme portion of the terminal moraine. Before the 

 Glacial period there was a continuous depression connecting 

 Paint Creek with the valley of Brush Creek. At a point 

 about five miles south of Bainbridge, on Paint Creek, in 

 Eoss county, this depression is filled up, from side to side, 

 to a height of about two hundred feet, with a glacial deposit 

 containing numerous northern bowlders and scratched stones. 

 On the northern side, toward Paint Creek, this deposit ex- 

 hibits every mark of a terminal moraine, with its character- 



