100 PROCEEDINGS OF THE AMHERST MEETING 



lacustrine tills. The lower contact is more even, as far as observed, and this 

 is also true of similar tills in the Pleistocene and older cases. The thickness, 

 as far as observed, varies from 8 or 9 feet to 3 or 4 feet. 



A preliminary examination of the matrix with the microscope has been made 

 by my colleague, Prof. J. E. Wolff, and the author. The matrix contains small 

 angular and rounded grains of limestone, quartz, coarse slate, minute frag- 

 ments of limestone conglomerate, chert, secondary chlorite, secondary calcite, 

 secondary rhombohcdra of dolomite, plagioclase. sericite. and a dark isotropic 

 substance with index of refraction higher than balsam. In one fragment of 

 this substance were found small vugs filled with chalcedony and quartz. Small 

 fragments of fossils may be seen rarely. The cement is very fine and is com- 

 posed, for the most part, of minute grains of quartz, lime carbonate, and a 

 paste of sericite. The material of the matrix is mostly fresh and angular. 

 The structure of the matrix differs in no way from the matrices of other tills 

 and tillites. Very angular quartz grains constitute the most important single 

 character, and these are common to all tills and tillites. 



The shale above and below this tillite bed is most regularly banded in dark 

 brown and gray layers. The microscope proves that the light layers have very 

 tine clay material, usually homogeneous throughout, ending abruptly on top 

 with the finest material of all. These fine layers are succeeded by the brown 

 layers, composed of coarser and angular grains, mostly of quartz, but with 

 other of the minerals found in the matrix of the tillite. The brown color is 

 due to an infiltration of an oxide of iron which could find its way easily be- 

 tween the coarse grains and come to rest, but was for the most part unable to 

 penetrate the impervious clay of the fine layers. The coarse layer has the 

 characteristic alternations of coarser and finer materials found in the Pleisto- 

 cene banded clays and also as found by the author in the Squantum slate and 

 s number of other slates and shales. It would appear from the observations 

 so far made that this shale at Levis has good seasonal characters. 



The extent of the tillite bed is not fully known. In the limited time at our 

 disposal it was traced only about 400 feet. 



The Irasburg conglomerate, of Beekmantown age in Vermont, is thought by 

 Prof. C. H. Richardson, of Syracuse University, to be of glacial origin. This 

 conglomerate has been traced for 100 miles in Vermont and for 25 miles into 

 Canada. 



Read from manuscript. Illustrated with lantern slides. 



STAGES OF THE ICE ACE 

 BY WARREN UPIIAM 



{Abstract) 



In a review of the series of glacial and iuterglacial stages ascertained from 

 the North American drift deposits, the early Nebraskan stage of glaciation is 

 correlated with the first high-water stage of Lakes Bonneville and Lahontan, 

 estimated by Gilbert to have been five times longer than the second and latest 

 great rise of tbese Quaternary lakes. Intermediate nearly or quite complete 

 evaporation of these lakes, during time estimated longer than the post-Glacial 

 period, appears to be represented by the iuterglacial Aftonian stage of wide 



