168 NEW YORK STATE MUSEUM 
It is very doubtful if any such sliding motion of thin ice is in 
accord with the known physics of ice movement. In order to get 
the matter before the reader briefly and without entering upon 
a discussion involving physics and mathematics, it will be con- 
venient to picture the glacier as made up of two major layers, 
a thin basal layer carrying most of the drift and moving slowly 
and a much thicker layer of comparatively clean, upper ice moy- 
ing more rapidly. The acceleration of velocity due to increasing 
PuESSUTe) 1s ‘Sreateriin the upper tees than in uhesbasal layer ein 
the accelerations are increased by a thickening of the ice beyond 
a certain critical ratio the upper ice practically overrides the basal - 
layer with its inclosed drift, and, forming beneath it a fluted sur- 
face bounded by streamline curves, it slowly squeezs the lower 
ice free from its burden and carries it forward, leaving the drift 
content pressed against the ground in elongated ridges. 
If this is the true explanation of drumlin fields, it follows that 
they were formed under thick ice; hence we have interpreted the 
barrier to drainage which unquestionably existed during the 
earlier stages of ablation in the Mohawk and upper Hudson val- 
leys, the Schoharie basin and over the plateau north of the Cat- 
skills, as due to the longer persistence of an overthickened portion 
of the glacier, which lay above and protected the drumlinized 
area from both erosion and sedimentation while the surrounding 
thinner ice was melting off. 
There is nothing to indicate whether this thrust of live ice was 
made before or after the general condition of stagnation of the 
elacien i betore, tne tiie 16a became stacmantanicamlediateliy 
after the drumlins were formed; if after, it was pushing into a 
body of dead ice of such extent that it “took up ” and dissipated 
the motion (mainly along a path of least resistance up the 
Mohawk valley). Because the latter is a possibility it is allowed 
to stand as a qualification of the first statement of the hypothesis 
presented in this paper (see page 159). But in either case, the 
persistence of this mass with no subsequent modification by 
advancing ice of the land surface which it produced, is proof of 
stagnation over the area which it covered. 
Because the southeastern Adirondack drainage was blocked 
by this overthickened ice, the freeing of the upper Hudson River 
basin was not accomplished until a time somewhat later than the 
freeing of corresponding levels at the southwest base of the moun- 
tains as previously noted. The first important level of the 
