436 STEVENSON— THE FORMATION OF COAL BEDS. [Nov. i. 



southwest. This stream must have been determined very quickly 

 after the region became <;Jry land, for, before the end of the Nev^ 

 River, it had made a terraced valley, at least 30 miles wide in some 

 parts of its course and showing irregularities of surface which 

 indicate existence of many minor tributaries. The main stream. 

 had its source in the Canadian highlands; an important tributary 

 drained part of the Mississippian ridge of Michigan and its course 

 lay within the area of Lake Erie; a second important tributary 

 came from the east, draining a part of southwestern New York and 

 northern Pennsylvania. In Ohio, this valley area is for the most 

 part west from the present outcrop of Pennsylvania beds; but its 

 eastern side is traced readily from 10 miles south of Lake Erie to 

 the southern border of Summit and Portage counties, about 80 

 miles, for it is filled with coarse sandstone and conglomerate, which 

 thin out abruptly at the east, permitting the overlying beds to rest 

 on Mississippian rocks. The western edge is approached in Wayne 

 and ]\Iedina counties, where the coarse beds become very thin. 

 Fragmentary exposures of the eastern side are found in counties 

 farther south until, in Vinton, Jackson and Pike, the Pennsylvanian 

 outcrop swings westward and the exposures extend to apparently 

 midway in the valley. Thence into Kentucky the outcrop again 

 trends southwardly and only the extreme eastern border is seen. 

 In northern Kentucky, another valley, now filled with New River 

 beds, begins in Carter county at 20 miles south from the Ohio 

 River and deepened rapidly southward in Mississippian rocks. Its 

 direction, west of south, is such that, within a few miles from the 

 last recorded exposure, it should unite with the main valley. In 

 southern Kentucky on the western side of the Alain Valley area, 

 M. R. Campbell discovered a valley in Mississippian rocks, deepen- 

 ing toward the east and now filled with New River beds. Farther 

 south, in Tennessee, the relations soon become indefinite, but ob- 

 servations on the, eastern side by Hayes make sufficiently clear that 

 the valley was distinct at half way to the Alabama line until midway 

 in the New River. 



Some features of this valley should be emphasized. Its exist- 

 ence is distinct for not less than 400 miles and, beyond central Ohio, 



