166 NEW YORK STATE MUSEUM 
difficulty, nevertheless the evidence is so abundant and clear that 
the general outline of the history can hardly be misinterpreted. 
The coarser drift which has been moved little if any from the 
position it held when the ice first became motionless is easily 
distinguished from the finer materials washed out from exposed 
basal drift to become superglacial; and, of the “ modified ” drift 
accumulations, those which have been deposited on land are quite 
different both in internal structure and external form, from those 
which have collapsed from a former position on the ice. 
The “ Adirondack sands” were distributed widely over the 
encircling ice. At the lower levels (400 feet down) they some- 
times form a sharp contrast with the locally derived material. 
From the marked level northeast and east from Trenton Falls 
which Professor Fairchild has called the Herkimer Lake “the 
ultimate escape of the waters must have been to the south and 
over the belt of stagnant ice”’ to Susquehanna drainage. There 
is reason to believe that the ice to the east in the Mohawk and 
Hudson basins had been thickened by a thrust of live ice, possibly 
into the stagnant glacier, by way of the Champlain portal (see 
page 108) thus barring the way to outlet in that direction. 
Where was the “ice front” at this stage of the “ recession ”? 
All the headwater streams of the Susquehanna system, from 
the Chenango river to the Schoharie basin follow rather high- 
walled well-opened valleys. All these valleys held long tongues 
of ice when forced drainage began to run through them and still 
held, as remnants of those fillings, considerable masses of dead 
ice after the last forced drainage was diverted (probably westward) 
beyond the northern divide. Such widespread stagnant ice, exist- 
ing contemporaneously, seems to the writer to demand a general 
condition of stagnation for its explanation. Though the deposits 
throughout this part of the plateau are principally concentrated 
in the valley bottoms, terraces having the form of kame-terraces 
appear here and there on the slopes and a few complexes of till 
and mostly unstratified gravels associatel with kettles indicating 
deposition over thin dead ice were found on the tops of the inter- 
stream ridges. 
That waters ponded by stagnant ice against the southwestern 
slopes of the Adirondacks, escaped somewhere into the Susque- 
hanna system while its headwater valleys were still filled with 
remnants of the glacier is reasonably certain; it is not so clear 
that we can judge by the land topography alone just where the 
