204 J. D. Dana on the Glacial and 
which the cooling reached its maximum; and probable also 
that there was some increase of difference between the level of 
the north and south corresponding somewhat with the increase 
of cold. 
XII. The scratches of eastern Canada, of the high land of 
northern New England, and of eastern and western New York 
and northwestern Canada, point to a part of the Canada water- 
shed between the St. Lawrence and Hudson bay as the head of 
the glacier that moved southeastwardly over New England.* 
The large valley of the river St. Lawrence, over 800 miles wide 
between the watersheds on either side, and trending east of 
northeastward, afforded no discharge for the ice; and this is 
proof that the summit-surface of the glacier about the mouth 
of this river, or over the St. Lawrence bay, was somewhat 
higher than over the watershed to the west. 
In order that the glacier ice should have flowed over the 
whole line of the barrier or watershed bounding the St. Law- 
rence valley on the south, the level of the ice over the Canada 
watershed must have been the higher; and so also that in the 
St. Lawrence valley, for the first result of the movement would 
have been to raise the level of the valley ice to that of 
the barrier in front, the law of flow being, according to the 
generally accepted view, much like that of a stiffly viscous fluid. 
But that the glacier should have abraded the White Mountain 
the plateau about the headwaters of the Connecticut, the gene 
ral level is about 1500 feet above the sea, which would make 
the upper surface of the glacier, in that region, if it were of 
the same thickness, about 6500 feet. 
put we have the means of arriving at a more certain conelu- 
sion with regard to the last mentioned altitude. The slope of 
* See this Journal, ITI, ii, 324. 
. 36° EL: ; . : ; 
47° 44’ long. 69°-69° 12’, S. 49°-64° E.; near Lake Temiscouata, lat. 47° 35) . 
39’, and long. 68° 39” to 49’, S. 48°-54° E., with one observation of S. 27° B.; 
on : . : " + i 
° 9’, long. 69° 8’, S. 32° E. These courses are cited eee 
not to prove the convergence alluded to, but to show that the system of movemen 
was the same north of the high northern New England border as south of it. 
