Wo.Klwnrlh.] 216 [Feb. 7, 
been diverted from a more suitable grade and course on to a 
somewhat level-topj^ed esker apparently in order to save to the 
agriculturist the arable land in the vicinity. The road from 
Attleboro to Briggsville, Mass., for the distance of a mile follows 
the crest of an esker. Other examples occur about Meriden 
and Bristol, Conn. 
Eskers in general exercise a sterilizing effect upon the commu- 
nities they traverse. Except for their occasional use as natural 
embankments for roads, as a reservation for chestnut, oak, and 
other hardy trees, or as borrow-pits for coarse gravel, they are 
the least desirable of the glacial deposits. 
Classification of Eskers. 
Deposits made by glacier water are usually referred to by the 
name of the stream concerned in their formation ; but eskers 
present some features requiring in our present understanding a 
different system of nomenclature. 
The normal esker is a single, well-defined ridge. In some 
instances the ridge is double, indicating, as Professor Davis has 
suggested, a superglacial origin, or the widening of a tunnel 
until the arch became deflected downward dividing the stream 
into two parts. There are also examples of eskers which are 
composed of till instead of water-laid drift, to which I have given 
the name false eskers. One occurs as a short loop in the group 
of eskers lying on the large drumlin in Hyde Park, Mass. Such 
a deposit appears to preclude the action of running water. The 
channel abandoned by the water may have in this case been filled 
by the crowding in of till from the base of the ice during a slight 
forward advance. If it is supposed that the till came in from 
above, it is difficult to account for the lack of the action of 
running water. 
Mr. Upham^ is reported as ascribing "the eskers to ice-walled 
streams open to the sky, since no boulders or till fell from an ice 
roof upon them." Opposed to this view is the fact of a large 
ano-ular block resting on the crest of the esker on the west of 
the road from Attleboro to Briggsville, Mass., and about one 
mile east of the former town. The perched position of this large 
ancrular block it seems to me must be explained, as in the case of 
' Amer. geol., v. 12, p. 177, Sept., 1893. 
