250 Glaciation of British Columbia.^Dawson. 
fluent glacier which spread from the mainland and was sub¬ 
ject to minor variations in direction of flow dependent on sur¬ 
face irregularities, in the manner described in my report on the 
northern part of Vancouver Island. 4 No conclusive evidence 
was here found, however, either in the valley of the Stikine 
river or in the pass leading inland from the head of Lynn 
Canal, to show that the ice moved seaward across the Coast 
Range, though analogy with the coast to the south favors the 
belief that it may have done so. The front of the glacier must 
have passed the outer border of the Archipelago, as at Sitka, 
well-marked glaciation is found pointing toward the open 
Pacific 5 (average direction about S. 81° W. astr.). It is, how¬ 
ever, in the interior region, between the Coast Range and the 
Rocky Mountains proper and extending northward to latitude 
63°, explored and examined by us in 1887, that the most in¬ 
teresting facts have come to light respecting the direction of 
movement of the Cordilleran glacier. Here, in the valleys of 
the Pelly and Lewes branches of the Yukon, traces were found 
of the movement of heavy glacier-ice in a northerly direction. 
Rock-surfaces thus glaciated were observed down the Pelly to 
the point at which it crosses the 136th meridian and on the 
Lewes as far north as lat. 61° 40', the main direction in the first- 
named valley being north-west, in the second north-north¬ 
west. The points referred to are not, however, spoken of as 
limiting ones, for rock exposures suitable for the preservation 
of glaciation are rather infrequent on the lower portions of 
both rivers and more extended examination may result in 
carrying evidence of the same kind much further toward the 
less elevated plains of the lower Yukon. Neither the Pelly 
valley nor that of the Lewes is hemmed in by high mountain¬ 
ous country except toward the sources, and while local varia¬ 
tions in direction of the kind previously referred to are met 
with, the glaciation is not susceptible of explanation by mere¬ 
ly local agents, but rather implies the passage of a confluent 
or more or less connected glacier over the region. 
In the Lewes valley, both the sides and summits of rocky 
hills 300 feet above the water were found to be heavily glaciat- 
4 Annual Report Geol. Surv. Canada, 1885, p. 100 B. 
. 5 Mr. G. F. Wright has already given similar general statements 
with regard to this part of the Coast of Alaska, American Naturalist, 
March, 1887. 
