Surface Geology of the Ynkon Territory. — Nordenskjold. 291 
was not able to take any soundings to ascertain their depth. 
Out of the lowest of these lakes, Deep lake, the river flows for 
almost three miles through a canyon with vertical walls of 
rock, and is impassable in the summer time. These canyons, 
that are found on both sides of the pass, are very interesting. 
I think it presumable that they were formed subsequent to the 
glacial age and consequently they show \\o\v long a time 
must have elapsed even at this hight above the sea since this 
region was covered by ice. The severe decay that some of 
the dififerent kinds of rock in the sides of the valley show points 
in the same direction. The origin of the canyons is perhaps 
connected with an unequal and extreme uplift of the central 
portions of the mountain-chain. 
At the foot of the terrace-wall passed by the canyon is 
situated lake Linderman, the first of the extensive system of 
large lakes that occupy a number of parallel valleys at a higlu 
of 2,150 to 2,170 feet above the sea, joined by cross valleys 
also filled with water. The lakes are quite fjord-like and are 
enclosed by steep walls of rock on an average 3000 feet high. 
They are moreover very deep. I took a number of soundings, 
though not as a rule in the centre of the lakes ; they conse- 
quently may be taken to show too low results. In lake Lin- 
derman at a mile from the southernmost end, the depth was 
130 feet, in the middle i6o feet, and at a quarter of a mile from 
the foot of the lake it was 80 feet. In lake Bennett only two 
soundings were taken. The greatest depth discovered was 
about 200 feet. In the S. W. arm of lake Tagish I registered 
a depth of 80 feet, but the depth at the entrance to the arm 
called Windy bay is probably much greater. Lake Marsh and 
lake Nares are filled with sediment and 'are everywhere shal- 
low; the greatest depth discovered in the former was 40 feet. 
At lake Linderman the above-mentioned and elsewhere 
often described terraces are first seen. The plaiiiest of them, 
consisting of gravel and silt, is exactly 50 feet above the sur- 
face of the lake, as measured by the barometer. Above it 
however there are others almost equally clearly marked, the 
topmost one being 125 feet above the lake. At a still greater 
elevation, 280 feet above the lake, is the beginning and lowest 
part of the mountainous terrace-plateau intersected by 
the above mentioned canyon. The rock is here as a 
