Classification of Glacial Deposits. — IVoodworth. 83 
kame field, and an outer edge at the head of the sand-plain, 
if that has not entirely broken down into kames. 
In the case of the water-laid drift, the ice-contact is fur- 
ther marked by textural, structural, and topographic features 
of a minor sort. As to texture, the fragments composing the 
detritus are invariably coarser in the contact zone than out- 
ward in the body of the sand-plains or terraces, because the 
currents issuing from the ice diminish as the distance increases 
from the ice-contact. Boulders and even patches of till are 
not infrequent exhibitions in the contact zone because of the 
deposition from moving ice. As for structure, the fore-set 
and top-set beds of sand-plains, as shown by professor Davis, 
spring out abruptly from this zone and decline, the one steeply, 
the other gently, towards the outer limits of construction. 
There is, even where the back-set beds hav^ not been ob- 
served, a thin talus deposit frequently showing the slidi'if^ 
down of the material? after the melting out of the ice. As for 
the associated topographic features, the ice-contact flanks the 
highest part of the sand-plain and the lateral terrace, except 
where the latter has been built out from the valley wall towaul 
the ice or by lateral drainage indifferent to its banks. 
The importance of recognizing and using the ice-contact 
in unravelling the history, relations, and genesis of the water- 
laid glacial deposits cannot be too much insisted upon. Prac- 
tically all the advance which has been made in the last ten 
years in the understanding of the stratified drift has come 
about through the tracing of this feature, as is witnessed bv 
the work of Salisbury in New Jersey, Willis in Oregon, and 
numerous other workers in America alone. Without the 
recognition of the ice-contact, expressed or implied in the 
interpretation of topographic form, the long, open hollows 
of a plain, as that of New Haven, remain to be explained 
as depressions kept open by violent eddies in powerful floods 
of water, sweeping, in this instance, down the lower Connecti- 
cut valley long after the ice had disappeared from the vicinity. 
X'iewed in the new light, the sands and gravels of this vallev, 
instead of being deposits formed long after the ice-sheet re- 
treated to the mountain valleys of the upper Connecticut, are 
seen as accumulations making in the presence of lingering 
masses of the ice, a view which breaks down much of the 
