DECEMBEU28, 1900.] 



SCIENCE. 



993 



tween rims of basic volcanic rock on the 

 south and granites and gneisses on the 

 north. These are regarded as Archean in 

 age. Between these rims lie two series of 

 Huronian sediments separated bj' an un- 

 conformitj^ The lower Huronian sedi- 

 ments comprise in ascending order quartz- 

 ites, dolomites and jasper. The upper Hu- 

 ronian beds are a jasper and ore formation, 

 black slates, a second ore formation and 

 gray slates. Over these unconformably 

 lie horizontal beds of Lake Superior sand- 

 stone. 



The ore formations consist of alternating 

 beds of jasper, hematite and quartzites. 

 The principal producing horizons are in the 

 upper Huronian. The lower ore-bearing 

 beds are mainly fragmental, and the upper 

 ore-bearing beds are mainly altered crystal- 

 line sediments. The ore of the latter has 

 come from iron carbonates, which have been 

 decomposed as in the Marquette district, 

 yielding cherts and hematite. 



All of the Huronian rocks are strongly 

 compressed and closely folded. The ores 

 occur in pitching synclines with impervious 

 bottoms. Geologically the Menominee dis- 

 trict bears a striking resemblance to the 

 Marquette district. The lower Huronian 

 ore measures, however, which are large pro- 

 ducers in the latter district, are scarcely 

 known in the Menominee district, in which 

 district the principal producing mines are 

 in the lower ore formation of the upper 

 Huronian. 



In a paper on ' The Still Elvers of Western 

 Connecticut,' Professor Wm. H. Hobbs de- 

 scribed the general course of the streams of 

 this region as being to the south-southeast 

 down the slope of the Cretaceous plain of 

 erosion. In a few cases, however, large 

 tributarystreams arefound flowing in nearly 

 the opposite direction. Two notable in- 

 stances of this sort have been studied, each 

 bearing the name ' Still Kiver ' ; and atten- 

 tion is thus directed to their exceptionally 



sluggish currents, due to the barely percep- 

 tible slope of their present beds. Oue of 

 these streams rises near Tarrington, flows 

 north-northeasterlypast Winsted, and, after 

 a course of about twelve miles, enters a 

 branch of the Farmington at Eobertsville. 

 The other river of the same name, some 

 twenty-five miles distant to the southwest, 

 is a tributary of the Housatonic, having its 

 source in a barrier of drift hills south of 

 Bethel, flowing north northeasterly past 

 Danbury and Brookfield, to enter its trunk 

 stream just where the latter departs from 

 the limestone valley to cut its way through 

 gneiss. 



In each case the course of the Still River 

 has been determined by a belt of limestone 

 within harder walls of gneiss and schist. 

 The Still Eiver, tributary to the Farming- 

 ton, is, furthermore, an instance of reversal 

 of drainage brought about by obstructions 

 of glacial material. 



In a paper on ' Drift Erosion, Transpor- 

 tation and Deposition,' by Mr. Warren 

 Upham, the work of the ISTorth American 

 ice-sheet is described as threefold. Its ero- 

 sion of the bed rocks, over the greater part 

 of the glaciated area, is shown to have 

 supplied far more drift than was desired 

 from the preglacial residuary claj' and 

 river sand and gravel. Only near the bor- 

 ders of the ice-sheet, or to a distance of 

 two or three hundred miles from it in the 

 interior of this continent, the successive 

 stages of fluctuating glaciation added each 

 its drift deposits without general erosion of 

 the underlying rocks or the earlier formed 

 drift. The transportation of the drift ap- 

 pears to have been chiefly within the lower 

 part of the ice-sheet, reaching in consider- 

 able amount at least 1,000 feet above the 

 land surface on the mainly plain-like region 

 of Minnesota and Manitoba. Its deposi- 

 tion for the greater part was directly from 

 the ice, yielding the till and a large propor- 

 tion of the mass of the moraines. Another 



