KLONDIKE AND YUKON GOLDFIELD—CADELL. 315 
out the stones and another to attend to the engine. In order to 
give headroom for working, the gravel thus undermined is broken 
down in lumps and thrown back into the waste, and the loose stuff fills 
it nearly to the roof. In course of time the superincumbent stratum 
thaws and subsides gently like the roof of a long-wall working, and 
closes up the space above the waste, and the surface of the ground 
sinks down to the same extent 
This frozen gravel or ‘‘muck’’ provides a surprisingly strong roof 
to the working. It is, however, sometimes forgotten that water is 
a true mineral, and in its crystalline or frozen state is as much a rock 
as granite. When thus solidified in the interstices between hard 
grains and pebbles it forms a very strong and homogeneous block 
without fissures or joints to weaken it or interrupt its continuity. 
Thus it is that drift miners can work with comparative safety, 
especially in winter, and only require to leave an occasional solid 
pillar or put in a little timbering to support parts of the roof that may 
be weak. In winter time the frost is most intense and the roof not 
so liable to fall in. In one case on Dominion Creek, a ‘‘muck”’ 
roof of this kind, unsupported by pillars, is stated to have covered 
a vault 140 feet wide by 230 feet long, and remained unbroken, till 
midsummer. 
The thawing of the gravel was originally carried on by wood fires 
placed against the face like the ancient method of fire setting to dis- 
integrate the lode in metal mines before the days of explosives, but 
the use of steam points soon superseded this primitive process. There 
is, of course, considerable danger of individual stones or slices of the 
roof dropping down, and fatal accidents have often occurred from 
this cause. In the mine I have described I noticed a continuous 
shght shower of sand grains dropping on my head from the thawing 
skin of the roof, but happily no large hailstones were among them. 
The depth of the frozen ground is variable, and is less on the 
ridges than in the valleys. A shaft sunk on the ridge south of Eldo- 
rado Creek reached unfrozen ground at a depth of 60 feet, while one 
in the valley of the same creek was stopped by running water at a 
depth of a little over 200 feet. The advantage of the ice is thus 
obvious from the pomt of view of pumping, which, if it became 
necessary, would put an end to many of the poorer drift mines in 
the valleys. But for surface work the ground must be thawed arti- 
ficially when the gravel comes to be handled on a large scale at a 
depth not affected by the summer sun. 
The bed of peat, or ‘‘muck,” as it is called, that covers she valley 
bottoms acts as a nonconducting skin and prevents the sun’s rays 
from penetrating the frozen mass, but when it is cleared off the sur- 
face thaws permanently to a depth of several feet, and can be removed 
by a scraper and sluiced or otherwise treated. 
