ptJBiNGTON.] HYDRAULIC MINING. 119 
four years' service. The annual cost of maintenance is $1,000, or 1<> 
per cent of the original cost. 
In ditching in bad ground it has been found beneficial not to remove 
the moss carpet, but to cut it in 4-foot lengths, transverse to the line 
of ditch, then to roll it back along the lower edge, and after the ditch 
is dug to let it drop so as to cover the side of the ditch. 
In the region adjacent to Eagle, Alaska, it has been found that 
where the topography did not admit of ditching the maintenance of 
Humes is very expensive and that the use of steel-pipe conduits is 
preferable. 
A useful suggestion for flume building was obtained on Deadwood 
Creek, in the Birch Creek district. Flat timbers obtained in the 
neighborhood were used, and the flume, 3 feet by 1 foot, was calked 
with moss inside. The cost was only 25 cents a foot, and the flume is 
said to be good for ten years. In this locality it is almost impossible 
to secure sawed lumber at any price unless the operators whipsaw it 
themselves. In all northern latitudes the moss is an excellent calking 
material. 
Through the interior country it has been found that ditches on the 
south slopes of the mountains can be made with much more success 
than on the north slopes. In fact, the north slopes are frequently 
bare rock, while on the south slopes a good deposit of earth has 
accumulated. 
The presence of crystosphenes, a which are found on the slopes as 
well as in the creek valleys in all parts of the north where perpetual 
frost prevails, is the ditch maker's greatest obstacle. Mr. Tyrrell 
saj^s: w * As a rule they [the crystosphenes, or so-called 'glaciers'] occur 
as more or less horizontal sheets of clear ice, from 6 inches to 3 feet in 
thickness, lying between layers of 'muck' or fine alluvium, usually 
where the c muck' is divided horizontally by a thin bed of silt or sand; 
and most of them, as far as my observation goes, are from 2 to 1 feet 
below the surface, though some are deeper. They approximate closely 
Ito the slope of the surface, under which they lie." These sheets are 
from 25 to 150 feet in diameter, generally somewhat oblong. They 
arc explained by Mr. Tyrrell as due to the freezing of seepage water, 
pausing a gradual accretion to the mass annually. Mr. J. P. Tol- 
machof 6 has examined similar occurrences in northeast Siberia, and 
explains them as fossil snowbanks. Whatever their explanation, they 
ire exceedingly common both in the interior and in Seward Peninsula. 
They are found not only in creek beds, but on the slopes, and show 
no regularity in distribution. 
^Tyrrell, J. B., Crystosphenes, <>i buried sheets of ice, in the tundra of northern America: Jour. 
}eol., vol. 12, 1904. 
''Ground ice of Berezovka River: Proc. K. Russ. Min. Soc., St. Petersburg, 1903. 
