206 J. D. Dana on the Glacial and 
charge being still to the south) over the St. Lawrence valley. 
But as already mentioned, the scratches over the region from 
western New York and Lake Huron to eastern Canada and 
Maine point toward the watershed as the head of the flow; and 
hence, there was apparently no discharge northward into Hud- 
son Bay, and moreover the ice must have stood high above 
this region of alamecor i is — that the —_ nce 
the glacier was a little to the north of the summit o 
watershed ; but more erent that ae small eemeety i abel 
the region of the watershed had from its elevation, and from its 
contrast in this respect with the Hudson Bay depression, was 
retained throughou 
while the thickness of the ice thus increased to the 
northward and northwestward, the amount of precipitated mois- 
ture must have decreased in that direction. The rates of pre- 
cipitation for different latitudes going northward was probably 
nearly the same as now in our winters. In summer the greatest 
amount of precipitation in New England occurs over the higher 
lands of its northern half, and the amount over Canada is but 
New Rao, and (3) the main part of the St. Lawrence valley 
with the Canada watershed, being 4: 8 Thus the low 
coast slopes take the moisture in winter, not the higher 
0: ter, 
mountains of the interior. ee accords — the general 
i 
—— in the winter, would have been that the “region of 
ost abundant precipitation was situated a little farther south 
than now, and the amount of diminution—not the ratio— 
* —— the excellent rain charts of O. A. Schott recently published (Tables and 
Results of the Precipitation * re and Snow in the United States, collected by 
the eege ag — d discussed under the direction of Joseph Henry, 
ache le No. f the Souithecetae Contributions, May, 1872), we jean that in 
winter, > the. acerder region of New England, 30 to 50 miles wide, the 
amount os rain and snow is 10 aiek or more, to 12 va boke large areas ite ‘ 
2 
Blodget’s rain charts give 5 inches as the average for a large part of the St 
Lawrence valley. The amount for the Canada watershed would be the same, OF 
less, since, as New England shows, there is no increase northward even if there 18 
——— elevation. The ratio o used above supposes 3 5 inches to be the mean for 
