22 BULLETIN OF THE LABORATORIES 



Mills. This we know, because an area of fully a square mile about C 

 has a solid deposit of drift-material rising more than loo feet above the 

 creek, composed largely of sandstone blocks whose angles are scarcely 

 worn, and masses of shale sometimes containing two or three cubic 

 feet, which disintegrate after a few weeks of exposure to the weather. 

 These could have come only from the gorge between E and the pres- 

 ent mouth of Ralston's run. They are intermingled with sand and 

 northern rock, promiscuously for the most part, but occasionally with 

 a rude stratification as if the floods had been somewhat intermittent. 

 The great apparent width of the valley below this point is due mainly 

 to the fiUing-in by drift ; but, it may be in part, also, to the earlier 

 discharge of Paint creek having enlarged it to some degree, as men- 

 tioned above. The fall of ravines and minor streams outside the 

 glaciated area is rapid; and in those filled with drift the depth to bed 

 rock may be roughly estimated by carrying downward the line of 

 slope of the hills on either side to iheir point of intersection. So of 

 the larger streams. This statement, of course, does not hold good 

 near the junction of two streams ; when they have cut down to the 

 level of those into which they flow, smaller streams can not further 

 deepen their beds, but will swing from side to side thus making 

 narrow bottoms. This is why there is always a widening of the 

 valley where two branches unite ; and in such cases the rule just 

 given will not apply. 



* * * 



From the northwestern point of the hill southeast of Frankfort, 

 one looks to the horizon northward over a practically level drift 

 covered country. The hill on which he stands bears, on one hand 

 north of east to the Scioto bottoms : on the other, it reaches a short 

 distance southward, bends toward the west, and finally sweeps away 

 northwest until it is lost to sight. This hill-land and the portion of 

 the plain adjacent to it are drained by North Fork ; the part north and 

 east of Frankfort is drained by Deer creek. Both are of post-glacial 

 date. The latter, being entirely superficial as regards the drift, need 

 not be considered : the former has a history. 



Prior to the advent of the ice, that part of the present valley of 

 North Fork between Frankfort and Paint creek was a depression with 

 an outlet in each direction, the dividing point between the two ravines 

 being near where Union and Twin townships corner. At this point 

 the shale hills are now less than loo yards apart; just below (south) 



