The Ice-Sheet in Xm-raf/a/isett Baij. — Woodworth. 159 
This line of retreat exhibits both phases of frontal deposit. 
A moraine occurs west of Pawtuxet station in the area delim- 
ited on the Providence atlas sheet by the 60 and 80 feet con- 
tour lines. The surface is here cast into mounds and the 
material is till with surface boulders of conglomerate from 
five to six feet in diameter. Surrounding the morainal accu- 
mulations proper are sand and gravel deposits, in the nature 
of overwash plains on the south and of sand}^ moraine ter- 
races on the north. 
Westward the southern bank of the river is clearly for the 
major portion of its extent and elevation a characteristic 
sand-plain head, with kanies and elongated hollows, these last 
marking the site of irregular projecting tongues of the ragged 
edge of the ice-sheet. Even beyond the line of contact with 
the main mass there are isolated hollows marking similar en- 
closed or buried remnants. Some of these depressions are 
drained by small brooks northward into the Pawtuxet. 
Where Elmwood avenue crosses the Pawtuxet river the 
head of the Pawtuxet ])lain and the frontal aspect of the next 
succeeding stage on the north are tolerably well contrasted. 
The Pawtuxet sand-plain is broken down into kames with 
here and there a distinct steep northward slope, all of which 
show (except at their base) no signs of the erosive action of 
the river, whose flood-plain is sharply marked topographically 
in the trough at their base. The criteria for the recognition 
of successive stages are distinctly shown. The head of the 
plain is higher than the opposite northern bank of the river, 
and the deposits on the south bank are perceptibly coarser 
than those on the north bank. 
This bank curves southwestward bej'ond the confines of the 
Providence sjieet on to the Narragansett Bay sheet. In the 
vicinity of the Rhode Island state farm, Pontiac, and Natick, 
the eastern or southern bank of the Pawtuxet is distinctly 
steeper than the opposite bank, Init it is also obvious that in 
places the slope alone is due to postglacial erosion, so that 
this criterion of the ice-contact is locally wanting; but the 
bight and conformation of the sand-i)lains confronting this 
slope and their relation to the deposits on the left bank of tlie 
river make it strongly probable that the course of the Paw- 
tuxet from wlu-re it enters the glacial plain at Natick to its 
