160 The American Geologist. Sei>tt'inber, i896 
mouth is throughout determined by the stand of the ice-front. 
The sand-plains of this stage between Natick and the State 
Farm have their head from 50 to 60 feet above the sea and 
slope to the southward and eastward toward Greenwood and 
Hills Grove. (Norton's pond marks the site of a large outlier 
of the ice which had not yet melted when deposition at the 
Pawtuxet stage was practically completed. Duck and Little 
Sand ponds also mark the site of lingering ice masses. At 
Hills Grove there are traces of the intraglacial bouldery or 
till deposits of a jirevious stage which were so high as not to 
be covered by the sands of the Pawtuxet stage. 
The intraglacial deposits of this stage are fairly well shown 
in the valley of a small brook north of the villages of Natick 
and Pontiac, where the gravels and sands of the next suc- 
ceeding stage were not carried southward to suffuse them. In 
this valle}^ are short eskers and kames, evidently marking de- 
posits made inside the ice while the large Pawtuxet j^lain was 
being deposited outside of the glacial sheet. 
Retreatal Formations on the C'entral and Eastern 
Shores of the Bay. 
Deposits of stratified drift, either in the form of sand-plains 
deposited against the edge of the ice-sheet or in the form of 
kames built up within its decaying margin, are not well devel- 
oped either on the islands in the axis of the bay or on the 
eastern shore from. Sakonnet point northward, though in the 
region of the Taunton river and the countr}^ westward 
through Barrington the development of stratified drift quite 
equals that of the west coast. 
Definite frontal deposits of any kind are unknown on the 
island of Aquidneck. The area of this island appears to have 
been bared of ice and above the level of deposition by glacial 
streams at the time drift deposits were laid down in the re- 
treat across the bay. At the northern end of the island occurs 
the remnant of a sand-plain, the connection of which with the 
ice retreat will appear clearer when considered with the de- 
posits of the east coast. 
On the island of Conanicut the surface deposits are mainly 
upland till, except for a small area near Sand point, opposite 
Wickford cove. The contour of this deposit is not sufficiently 
clear to correlate it with the class to which it belongs, but it 
