205 
[ VVoodwortli 
Eskers tend to lie in the direction of local glacial motion. This 
fact was long ago noted by Irish geologists, and is generally true 
of the more continuous and elongated eskers of this district. 
This coincidence of direction is obviously due to a control 
exerted by the ice-sheet. Salisbury has suggested that supergla- 
cial streams would flow towards the front of the ice in obedience 
to the slope. It seems obvious also that subglacial streams would 
be held to the same course (1) by the motion of ice in valleys, 
and (2) by the tendency of the ice to close up passages oblique 
to its own forward movement. 
Easting and westing of eskers. — Many eskers in meridional 
valleys tend to lie on one side. Generally, in the cases which I 
have observed, the easting and westing of eskers corresponds 
with a like tendency in the glacial striae. It is pronounced on 
the Meriden sheet, i the Providence^ and Burrillville^ sheets. The 
eskers in these cases are nearly or quite jjarallel with the valley 
walls. Notable instances of this distribution occur in the esker 
on the east bank of Compounce Pond, south of Bristol, on the 
Meriden, Conn., atlas sheet, and in the esker in the southern part 
of North Attleboro, Mass. This distribution follows from the 
rule of coincidence in direction with ice movement, and is in this 
district particularly noticeable with reference to the glacial lobes 
which lay in Cape Cod and Narragansett Bays, and in the Con- 
necticut River valley. In these fields, the eskers diverge south- 
wardly in the lobate axis, and converge southwardly toward the 
intermediate moraines. This phenomenon appears to be more 
pronounced in southern Massachusetts, Rhode Island, and Con- 
necticut than north of the region here referred to. This 
difference is apparently due to the fact that inland and remote 
from the more or less lobate margin of the ice-sheet, the motion 
of the ice which controlled the position of drainage lines, was 
more nearlj^ parallel across wide belts of country. 
The fact that eskers, as in tlie case of those on the Meriden 
atlas sheet, lie in positions determined by the direction of motion 
in the ice-sheet, is evidence that the period of stagnant ice was 
short, otherwise the streams in tlie ice would have had time to 
wander from courses determined by the conditions of an active 
glacier. 'I'lie presence of eskers in this position and their absence 
1 See forthcomino: descriptions of the I'ieistocene yeology accompanying these atlas 
sheets. 
