169 
most effective in melting away the ice, the surface of which 
would naturally slope toward the drumlins. The terraces are 
thus a joint produet of glacial and fluvial conditions, contempo- 
raneous with the modified drift ; and like that record the rapid 
waning of the ice-sheet when it had ceased to flow but yet lin- 
gered in the valleys. Some gullies or vertical erosion channels 
appear to have been formed on the drumlins at this time; and 
probably the many other depressions and hollows observed on 
their slopes, which are often ill-defined, but commonly some- 
what saucer-shaped, are the result, chiefly, of local landslips 
dating from this period when the till was first laid bare and was 
yet unprotected by vegetation. 
THE MODERN SHORE. 
When the sea finally retired from its highest postglacial 
level, it appears to have subsided quite rapidly to a level some- 
what below the modern beach. Evidence that the land has re- 
cently stood higher than now is found at other points in this 
region chiefly in the form of submerged peat beds and forests, 
and the former, at least, would probably be found if ехсауа- 
tions were made in the Nantasket marshes. Although the 
depressions occupied by the marshes and the winding channels 
of Weir River Bay, Strait’s Pond, ete., are clearly submerged 
land valleys, as previously explained, their origin — the erosion 
of the hard rocks — undoubtedly dates from the strong eleva- 
tion of the land at the beginning of the ice-age. In fact, it is 
not in land erosion, but in the absenee of marine erosion аё 
certain points where it might otherwise be expected to exist, 
that the evidence sought for is to be found. It is a fact familiar 
to all that the drumlins of the harbor exhibit marine erosion 
wherever freely exposed to the sea. This erosion scarp exists, 
in general, not only on the seaward, but also, in less degree, on 
the landward, sides of the drumlin islands. Being the work, 
almost exclusively, of the waves, it varies somewhat with the 
