430 NEW YORK STATE MUSEUM 



general slope of the drumlin area is up hill, but was produced by an 

 effective thrust on the marginal ice by the pressure of the rearward 

 mass. As the ice sheet thinned by ablation there came a time when 

 the drift-loaded ice in contact with the ground was subjected to less 

 vertical pressure and to relatively greater horizontal pressure by the 

 deep ice in the rear, and was pushed forward, bodily. In this fact is 

 believed to lie the key to drumlin formation. 



It does not follow that drumlins must always have been formed 

 where the bottom ice had a sliding motion, as several other condi- 

 tions are probably requisite, but- it seems quite certain that long- 

 continued and vigorous horizontal thrust is the prime necessity. 

 Such thrustal movement would be effective only where a border of 

 the ice sheet was backed by a thick or vigorously pushing rearward 

 mass. The combination of conditions requisite for effective thrust 

 movement over a belt of country and for the considerable time 

 necessary to build up the drumlins may be rare. It does not seem 

 so strange that drumlins are uncommon features of the drift when 

 we add to the requisite dynamic factors mentioned above the several 

 others which are doubtless directly concerned with the drumlin 

 formation. 



As a working hypothesis it may be assumed that wherever the 

 ground contact ice had a vigorous movement of some duration it 

 should be indicated by the molding of the ground surface, specially 

 where that surface is comparatively smooth and composed of drift 

 or soft rocks. The form and degree of the ice molding would vary 

 according to the strength and adjustment of the several factors. 

 An application of this idea can be made to the region under 

 present study. 



Well marked drumlins are not found" on the high ground east of 

 Seneca lake, and are wanting on the low ground east of Syracuse. 

 The explanation seems to lie in the relationship of the larger topog- 

 raphy to the movement of the ice sheet. When the glacier was deep 

 over the Finger lakes region the bottom of the ice in the drumlin 

 area was probably quiescent and served as the bridge over which 

 the upper ice moved by gravity; the repose of the lower ice probably 

 being due to the opposing land slope and to the large volume of 

 drift which the ice had incorporated. Over the nearly level area 

 north of the Finger lakes the waning of the ice sheet finally sub- ■ 

 jected the ground-contact ice to a vigorous and long-continued hori- 

 zontal thrust with consequent sliding motion. But in the adjacent 



