60 



SCIENCE 



[N. S. Vol. XXXIV. No. 863 



by adding wire loops. (A series of loops of 

 varying lengths is needed.) (2) Next trans- 

 fer the mineral to the lower pan. It will lose 

 weight, so the counterpoise is moved toward 

 the fulcrum until balance is restored. The 

 specific gravity is then indicated by the posi- 

 tion of the counterpoise on the beam. 



Accuracy. — Tried with such minerals as 

 quartz and calcite, this balance is accurate to 

 about two units in the second decimal place 

 for two or three grams of material. 



A Portable. Balance. — A convenient balance 

 for rough work in the field may be made of a 

 thin strip of wood, such as a foot ruler, 

 driving a nail through for a fulcrum. To the 

 short arm is attached a thin cord with rubber 

 elastic for holding the mineral. The long arm 

 is graduated so that the specific gravity may 

 be read off directly as previously described. 



The balance upon which the above descrip- 

 tion is based was constructed by Mr. F. A. 

 Stevens, mechanician at Stanford University. 

 Austin F. Eggers 



mineealogt laboratory, 

 Stanford TjNrvEBsiTy, Gal., 

 April, 1911 



WHAT CAUSED THE DRUMLINS ? 



To THE Editor of Science: The following 

 is a concise outline of a theory offered as an 

 explanation of the process of formation of 

 the peculiar smooth-contoured hills and 

 ridges called drumlins and their allied topo- 

 graphic forms that occur in certain localities 

 within the areas of the earth's surface for- 

 merly occupied by the ice sheet, notably in 

 central New York, in southern Wisconsin, in 

 portions of New England and of Canada, and 

 in Ireland. These features of the surface 

 have been the subject of much study and 

 speculation and of a variety of theories, but 

 so far as I can ascertain from available liter- 

 ature on the subject, the explanation here 

 given has not heretofore been propQsed. 



During the period of dissolution of the ice 

 covering certain glaciated areas, commonly 

 called the period of " retreat " of the ice sheet, 

 melting took place in the upper surface as well 

 as on the front wall or slope. Owing to the 



strains in the ice mass produced by the forces 

 that caused and attended the general advance 

 of the sheet its internal structure had be- 

 come such as to modify the process of melting 

 from the upper surface. Before melting be- 

 gan there had been formed in the ice a sys- 

 tem of vertical and parallel cleavage planes 

 and fissures and the general direction of these 

 conformed to the direction of the ice move- 

 ment, owing to the forces above referred to. 

 The assumed difference between a general ice 

 sheet on a nearly horizontal surface and an 

 individual glacier with a steeper descent in 

 respect to cleavage is here to be noted. 

 Changes of temperature with the changes of 

 season may have had something to do with 

 this structure. During the melting process 

 the upper part of the ice sheet became deeply 

 pitted or honey-combed on a somewhat gigan- 

 tic scale because of the fissures and cleavage 

 planes, and the pits were more or less elon- 

 gated horizontally in the direction of these 

 fissures and planes. As the melting proceeded 

 on the internal surfaces of the pits, enlarg- 

 ing them, of course, the earthy matter in the 

 upper parts of the ice, including stones, 

 boulders, sand and gravel, dropped to the bot- 

 tom of the pits and this material was thus 

 subjected to a certain amount of water action 

 and washing while the water drained away. 

 With the enlargement and deepening of the 

 pits and the removal of water the areas of 

 ground ice and land surface beneath the pits 

 were relieved of a large portion of the vertical 

 pressure which the full thickness of the ice 

 sheet had produced, while between the pits 

 this pressure remained nearly the same as 

 before melting started. The consequence was 

 that a slow movement or flow of bottom ice 

 towards the pits and an upheaval in the bot- 

 toms of the latter took place, and this lateral 

 and concentering and upward ice flow at the 

 bottom would, of course, carry with it the 

 " till " material which was located principally 

 in the lower portions of the glacial sheet, and 

 a certain amount of the underlying material 

 as well. There may have been periods during 

 which the general melting was checked, due to 

 seasonal changes of temperature or other 



