Pre-glacial Channel in Ohio and hidiana. — Bownocker. 1 8 1 
ty, passing two or three miles north of Pennville, and enter- 
ing into Blackford county. West of Geneva numerous wells 
show depths ranging from 400 to 440 feet, while just outside 
of the channel the drift ranges from forty to eighty feet in 
depth. The length of this channel is nearly 90 miles. 
Two tributaries to the channel have been located. The 
smaller one, extending from the northwest, unites with the 
main channel near the southv;est corner of Harrison town- 
ship, Blackford county, Indiana. This has not been traced 
outside of Blackford county. Its depth is as great as that of 
the main channel. The other tributary unites with the main 
channel at the west end of Grand reservoir. Running to the 
southeast from this point successive wells in Mercer county 
show^ depths of drift of 260, 300. 382, 315, 298, 402, 400, 455, 
and 340 feet. Outside of this channel the depths in the same 
locality range from 45 feet to 135 feet. The channel next 
cuts across the extreme southwest corner of Auglaize county, 
furnishing depths of 440, 380, 380, 370, and 400 feet. The 
<hift here outside of the channel does not exceed 90 feet. En- 
tering Shelby county the channel turns to the south, passing 
just west of the village of Berlin, where a depth of 350 feet 
was found, vrith a depth of 100 feet on either side. 
The depth of these channels is ir.ore than 300 feet below 
tlie surface of the rock floor. Usually the walls appear to be 
steep, and in one case a vertical one was found. This was 
near St. Marys, Ohio, where, after drilling through i8i feet 
of drift, rock was struck, but this was fovmd to be less than one 
foot thick, and then drift was again found. Several project- 
ing ledges of the limestone with drift between, were passed 
through in this manner, until a depth of 370 feet was reached, 
when the well was abandoned. .\s the channel is approached, 
the drill usually shows a gradual increase in depth of drift 
until the wall of the channel is met when the increase is very 
rapid. 
The width of the channels can be only apj^roximately 
given, because the oil wells are not usually sufficiently close 
together. Besides they are not ordinarily at right angles to 
the channel. The width appears to be usually about one 
mile. The greatest found are in the Grand reservoir an.d at 
Rockford, where widths of a mile and one-half api)arentl\' 
