290 The American Geologist. May, 1899 
in this district, above the head of the fjords. The Dyea inlet 
has a continuation in the form of a valley filled with sand, the 
topography of which does not differ in any noticeable degree 
from that of the inlet. This sand becomes as we proceed up- 
wards gravel of increasing coarseness. Beyond the shore-line 
it descends rapidly to the bottom of the fjord (50 fathoms in 
about ^ mile). The same is the case in the majority of the 
fjords. 
About ten miles above the head of the inlet the valley just 
spoken of divides, the chief branch coming to an end somewhat 
abruptly in a steep wall, through which the river has cut its 
way down into a canyon, several hundreds of feet deep, the 
walls of which are often perpendicular. Similar terrace-like 
walls in the continuation of the valleys are not uncommon; by 
large rivers they are cut through, but small creeks often form 
waterfalls over them, pursuing abovea meandering course. The 
trail is here on the terrace and does not reach the river again 
until some miles further up at Sheep camp ; here the river val- 
ley has a beautiful U-shape, with a rocky bottom everywhere 
visible. The course of the valley is suddenly stopped by a pre- 
cipitous wall of rock, nearly 800 feet high, over which passes 
the trail of the Chilkoot pass. 
2. The Region of the Mountain Chain a?id the Inland Lakes. 
The mountain-chain itself here is not particularly lofty, 
compared with what is found farther west, but, as might be ex- 
pected considering the dampness of the climate and the abun- 
dance of glaciers, it is exceedingly wild and broken in outline 
with high peaks and narrow ridges often surrounding deep 
amphitheatric valleys. Compared with the ocean side the 
slope towards the north is very moderate. This circumstance 
is evidenced in the fact of the more extensive valleys being oc- 
cupied in their upperparts for a width of more than 100 miles 
almost entirely by lakes, which may be divided into at least 
three clearly distinct groups. Uppermost, almost immediately 
below the hight of the passes (both at Chilkoot pass and at 
White pass), there is a series of comparatively small mountain 
tarns of an irregular, elongated shape and situated approxi- 
mately at the same level ; their shores, frequently consisting of 
smooth rock, points to their having been excavated by ice. I 
