Foshay — Preglacial Drainage of Western Pennsylvania. 397 



second sample was ignited in a closed crucible without C0 2 gas 

 and gave almost the same result, 1*71 per cent. In this sample 

 the FeO was determined, after solution in hydrofluoric and 

 sulphuric acids, and after reduction the total iron. As the 

 mineral after such strong ignition was only slowly acted on by 

 the hydrofluoric acid and in the experiment only partially 

 dissolved, the results were not thoroughly satisfactory but they 

 proved conclusively that the FeO had only been oxidized to a 

 trifling extent and the loss on ignition will represent therefore 

 very nearly the exact percentage of H 2 0. That the tendency 

 of FeO, in this combination, to oxidize is not very great is more- 

 over proved by the fact that repeated ignition did not yield 

 any increase in weight. That the H 2 is an essential con- 

 stituent of the mineral and is not the result of alteration, is 

 proved by the fact that it is very firmly united to the molecule, 

 requiring an intense heat to drive it off, and moreover it is 

 just sufficient to bring the ratio of protoxides to silica —1:1. 

 The transparency of the crystals would moreover prove that 

 the material which was analyzed was very pure and had not 

 suffered any alteration. 



Mineralogical Laboratory. Sheffield Scientific School 

 New Haven, May, 1890. 



Art. LI. — Preglacial Drainage and Recent Geological 

 History of Western Pennsylvania ; by P. Max Foshay, 

 M.S., F.G.S.A. 



The investigation into the preglacial drainage of Western 

 Pennsylvania, of which this paper is a partial record, was pri- 

 marily incited by a suggestion thrown out by Professor J. W. 

 Spencer, in a paper read before the Am. Phil. Soc, March 18, 

 1881. Professor Spencer there advanced the hypothesis that 

 the Beaver River, with part of the Ohio, had in preglacial 

 times constituted a stream which flowed up the modern Mahon- 

 ing through its now buried channel into the Erigan River 

 (Spencer) which then traversed the basin of Lake Erie. This 

 stream, now become parts of several modern rivers, 1 have 

 named Spencer River in honor of the investigator who first 

 suggested its existence and to whom is due so large a pro- 

 portion of our present knowledge of the preglacial drainage 

 of the region of the Great Lakes. Spencer River drained an 

 area nearly co-extensive with that of the Pennsylvania portion 

 of the modern Ohio with its tributaries, including the basin 

 of the Monongahela and part of that of the Allegheny, thus 



