G. H. Stone—Terminal Moraines in Maine. 379 
take of the nature of medial and terminal moraines. In clas- 
sifying these deposits it will be needful to distinguish accumu- 
lations of till which were formed beneath or within the ice-sheet - 
from superficial accumulations which were the result of the 
local movements sure to occur during the unequal melting of 
the retreating glacier. We know so little as to the nature of 
the work going on in the lower portion of an ice-sheet that this 
distinction is a difficult one to make. Omitting a great num- 
ber of doubtful cases, there are a few places in the State where 
the till is piled in narrow ridges which lie transverse to the 
direction of glaciation and which so clearly have the form and 
composition of terminal moraines that they must be regarded 
as having been formed at the end of a moving mass of ice. In 
some cases kames also appear to have been deposited at the ice- 
front and thus were practically a part of the terminal moraine, 
but they are omitted from this list. So far as I know the fol- 
lowing list includes all of these formations in the State con- 
cerning which the evidence is reasonably conclusive. 
1. The moraines of the local Androscoggin Glacier.—A fter the 
great ice-sheet was so far melted that the general movement 
ceased, there was still for a time sufficient ice in tne White 
Mountain region to cause a glacier to flow eastward along the 
valley of the Androscoggin River, as far as West Bethel, 
Maine. This local glaciation was first described by Professor 
G. L. Voseand Professor A. S. Packard in the first and second 
volumes of the American Naturalist, afterwards by Professor 
C. H. Hitchcock in the N. H. Geological Reports; and briefly 
by the writer in the American Naturalist for April, 1880. The 
most remarkable terminal moraine of this glacier is at the great 
bend of the Androscoggin near the boundary between Shel- 
burne, N. H., and Gilead, Me. No one who sees this can hesi- 
tate to regard it as a true terminal moraine of a local valley 
glacier. 
2. At Readfield Village—In the western part of Readfield 
village (near the woolen mill) is a confused series of hum- 
mocks and short ridges extending from north to south across 
the valley of the small stream on which the mills are situated. 
From this point the valley continues nearly west for about one 
mile when it turns northward and soon widens into a valley 
which resembles a mountain cirque. This wider valley appar- 
ently was filled with ice which flowed down the narrow valley 
to Readfield village. At several places along the valley are 
masses closely resembling lateral and semi-terminal moraines. 
At. the woolen mill the valley proper is only about one-eighth 
of a mile wide. The morainal hummocks are chiefly made up 
of till which is somewhat coarser and more sandy than the 
ordinary till of the region, as if slightly water-washed. In a 
