SURE Ee Meg ee a 
y n A 
ON THE GEOLOGY OF THE ISLAND OF AQUIDNECK. 521 
the land along nearly the whole shore of the island, but the slight 
waste is not taking place in any greater measure at the mouths 
of these valleys than at other points. I am inclined to think it 
eminently probable that these valleys are the result of the am- 
plification of the original stream excavations, by the action of 
the moving sheet of ice which we shall see there is abundant 
reason to believe to have rested on this region during the drift 
period. Long before the present topography was established, prob- 
ably at a time so remote that the present surface was buried more 
than one thousand feet beneath accumulations, which have since 
been worn away, the island had already been separated from the 
main land. The ice period found certain valleys already cut in the 
rock, their general course coinciding pretty closely with the direc- 
tion of the flow of the ice, as will be seen by examining the indi- 
cated direction of the ice as determined by the glacial scratches. 
It is eminently probable that the glacial sheet was pretty nearly 
: level on its upper surface, not having that surface conforming to the 
; hills and valleys which lay beneath. Itis evident that where the 
es _ Stream lay deepest the wearing would have been the greatest, for 
the rate of motion being the same, the wearing of a glacial stream 
‘is in proportion to the weight it brings upon equal areas. Thus 
2 the valleys would have been deepened and widened until out of 
7 proportion to the magnitude of the streams of water which course 
r through them since the glacial sheet went away. 
: The importance of this consideration has been overlooked ; it 
heeds to be considered if we would form clear ideas of the cause of 
Ue irregularity in the excavations of all glaciated regions. Some- 
2 times this great difference in the erosive action may be in part 
attributed to the difference in hardness of the rock acted on. 
7 But that this is not the principal cause is shown by the great irreg- 
ularity of glaciated surfaces and underlaid by the most uniform 
| Syenites. In ordinary ærial erosion the beds of the streams are 
_ Senerally kept away from the rock by a padding of debris; but 
= ' glacial action these protective agents in running streams 
r become the very sharpening of the ice tool and do the work of 
_ĉrosion instead of protection. By reference to the diagrams (Plates 
$ and 7), it will be seen that the channels between the islands have 
a average depth of about one hundred feet below the water line. 
They gain their depth quite suddenly and then preserve their 
Sentle slopes across their whole breadth. Their variations in 
k: 
Ei 
i 
: 
ee 
