244 



SCIENCE 



[N. S. Vol. XXXVII. No. 946 



a discussion already too long, in the opinion 

 of the writer all the facts and philosophy 

 of ice erosion argue against deep glacial 

 erosion in the Finger Lake valleys. 



One interesting product of glacial ero- 

 sion is to be noted. These are some hills 

 which have the form and attitude of true 

 drumlins, but which are composed of soft 

 shale, shaped into drumlin form. These 

 rocdrumlins will be described later. 



CONSTRUCTIONAL WORK. 



Sub glacial: Bocdrumlins. — The general 

 drift sheet presents no special features 

 meriting description at this time. The im- 

 portant subglacial deposits are the roc- 

 drumlins. New York state probably has 

 the best display of these interesting hills of 

 any district in the world ; in number, vari- 

 ety of form, variety in orientation, differ- 

 ence in composition and in the clear rela- 

 tionship to the correlating moraines. 



Much space might be given to descrip- 

 tion of these singular and most beautiful 

 hills, but they have already been quite fully 

 described in a Bulletin of the State Mu- 

 seum.' 



Possibly in other regions there may be 

 drumlins produced by the ice overriding 

 and reshaping moraines, but aU the true 

 drumlins observed in New York are cer- 

 tainly constructional in their origin. The 

 New York moraines are mostly water-laid 

 drift, especially north of the divide, the 

 debris in the ice being largely grasped by 

 the glacial drainage. If the drumlins were 

 moraine accumulations they would have 

 morainal composition and structure. On 

 the contrary, they are very compact tiU, 

 distinctly bedded with concentric structure. 

 The best exhibition of the bedding is shown 

 along the shore of Lake Ontario, between 

 Sodus and Oswego, where the undercutting 

 by the waves has dissected numerous drum- 

 ■ "N. T. State Museum Bull. Ill, 1907. 



lins from top to bottom and in different 

 directions. Sand or gravel within the 

 mass of the drumlin is of infrequent oc- 

 currence, though some of the drumlins be- 

 tween Clyde and Savannah hold consider- 

 able sand in their superficial layers. 

 Many drumlins exhibit decided difference 

 between the deeper and the superficial till, 

 sometimes so pronounced as to suggest two 

 epochs of construction. 



Along the belt of outcrop of the soft Sa- 

 lina shales there are drumlins which have 

 a shale base, and perhaps some with a shale 

 core. Fifteen miles northwest of Syracuse 

 and west of Baldwinsville the drumlin 

 forms are entirely shale. The deeply 

 weathered clay rock supplied to the ice 

 sheet a plastic material similar in its be- 

 havior to the ground moraine. These hills 

 are not true drumlins. They are whoUy 

 erosional in origin, as indeed are the true 

 drumlins in their shaping. We have called 

 them rocdrumlins, using the Celtic prefix. 

 It is possible that similar forms will be 

 found in the Champlain-IIudson Valley, 

 shaped out of the softer Ordovician shales. 

 The ice sheet does not appear to have had 

 scraping force sufficient to shape into the 

 drumlin curve any rock hills of harder ma- 

 terials than soft shale, though bosses of 

 crystalline rocks in the St. Lawrence Val- 

 ley and other districts of long-continued 

 abrasion are rounded and smoothed on the 

 struck sides. 



The mechanics of drumlin construction is 

 a complex problem. The required coopera- 

 tion and balancing of several dynamic fac- 

 tors make the drumlins exceptional fea- 

 tures even in the glaciated territory. The 

 more important constructional factors ap- 

 pear to be: (1) An excessive amount of 

 drift; (2) the drift of clayey or adhesive 

 and plastic material; (3) such thickness of 

 marginal ice and with such relation to the 

 rearward ice body that the whole depth of 



