J. D. Dana— Flood of the Connecticut Fiver Valley. 101 



(4184 feet), a level almost 3000 feet above any glaciers now 

 in that vicinity, and near the parallel of 63° N., in the vicinity 

 of Kuvnilik and Bjdrnesund, at heights of 940 to 1100 meters 

 (the latter 3609 feet) ; but none on the upper part of Nukag- 

 piarsuak, northwest of Kuvnilik, whose height is 1520 meters 

 (4987 feet). Mr. Kornerup also states that in the era of ex- 

 treme glaciation when the glacier was 3000 feet higher than 

 now, the movement of the ice was nearly east-and-west, but subse- 

 quently, as the scratches at lower levels show, it followed the 

 direction of the fiords or valleys : — a change evidently due to 

 the thinning of the ice, the pitch of the upper surface being 

 great enough when the glacier was at its maximum to cause 

 ice to move independently of the courses of valleys or depres- 

 sions beneath, and not so after the thickness had been much 

 reduced. This fact had its parallel all over glaciated North 

 America. 



(3.) In addition, the hypothesis makes the submergence of 

 the Coast r< c ' elevated beaches) to. have taken 



place durin ■. and to have passed its maxiunjm 



in the height of the period ; when, according to the facts, what- 

 ever the condition in the Glacial era, the submergence was a 

 prominent feature of the era when melting was going for- 

 ward and the ice finally disappeared-— the Ciiamplain period. 

 This point needs no special remark after the descriptions already 

 given of the Connecticut Kiver terraces, and the explanations 

 in the following part of this paper. 



(4.) But the Glacial era was not for the higher latitudes 

 generally one of less elevation in the land than now, and was 

 probably one of somewhat greater elevation for large portions. 



The arguments in favor of such elevation I here briefly 

 review, in order to test them by a reference to recent discoveries. 



(a) One of these arguments is based on the depth to which 

 many river channels are excavated below the present bed of 

 the stream. It has been strongly urged by Dr. Newberry and 

 others. The facts supporting it have been drawn from New 

 England and the States of New York, Pennsylvania, Ohio, 

 Indiana, Illinois, Wisconsin, and from British America; and 

 new cases are annually becoming known. The Pennsylvania 

 Report for 1881 by 'Mr. I. C. White, treating of 

 Erie and Crawford Counties (among the western counties of 

 the State), remarks that a boring for oil on French Creek. 

 below Meadville, descended for 285 feet through the drift; and 

 another, in Conneaut Creek'valley, 180 feet in drift; and in 

 connection with his account of these and other cases, he speaks 

 of it as a general fact that "-the present water courses meander 

 along the upper surfaces of drift-deposits which fill the ancient 

 valleys to various heights above the old rock beds." Other 



