42 THE PREGLACIAL DRAINAGE OF OHIO. 



mantle of drift and in a deep valley cut out of the same material.. 

 The valley is narrow near its outlet but expands up stream,, 

 and near the station, Dawson, is fully a half mile wide. A 

 large portion of this valley is undulating, and the irregularities 

 found suggest that it is an old valley filled, rather than a young 

 valley cut out of the drift. It seems to the writer that this 

 valley is preglacial and that the old stream may have flowed 

 through it to the vicinity of Berlin and there have entered the 

 buried channel which has been traced to that place. 



The channel to which reference has just been made was. 

 studied during the summer of 1898, and the report published 

 in the American Geologist for March of the following year. 

 During the summer of 1899 the work was continued and the 

 mapping of the valley extended. These channels are shown on 

 the map which accompanies this report. As may be there 

 seen, they lie in Champaign, Shelby, Auglaize, Allen and Mercer- 

 counties, Ohio, and in Adams, Jay, Blackford and Grant coun- 

 ties, Indiana. 



It must be borne in mind that all surface indications of 

 these channels have been destroyed by the great ice invasions. 

 So completely have they been filled that the present streams in 

 places flow at right angles to the preglacial ones. In fact the 

 course of one of the old channels in eastern Shelby county is. 

 now the site of a watershed separating the drainage of Lake 

 Erie from that of the Ohio river. 



Our knowledge of the location of these channels is due- 

 entirely to the driller for oil and gas ; and progress in mapping 

 these is likewise dependent upon him. All that we can do is- 

 to patiently follow the drill as it moves from section to section,. 

 and tabulate the facts which it discloses. Wherever this work- 

 ceases there also the work of mapping the old channel discon- 

 tinues. For the facts relating to the greater portion of these 

 channels reference must be made to the article in the American 

 Geologist already referred to. It is proper here to discuss such 

 additions only as have been made since that article was pub- 

 lished. 



Work during the past summer has been along two lines : 

 (1) Tracing a tributary of the main channel in Auglaize and 



