212 The American Geologist. Vpril, 1896 
the melting point, the enclosed rock debris will, on account of 
its lower specific heat, tend to rise in temperature faster than 
the ice and thus to loosen by melting the bonds between it 
and the ice. then the comparison with debris in pitch would 
seem to hold good only in the improbable case when the im- 
bedded stones caused a local softening of the pitch. The de- 
bris in ice would not lead to extensive melting on account of 
the high latent heat of melting ice. But if the temperature 
of the debris rises ever so little above 0° ('. (32° F. ) it ceases 
to be a source of strength in the ice, the etfeetive section of 
the ice being diminished in proportion to the amount of de- 
bris. This view seems to be abundantly confirmed by Prof. 
< hamberlin's Greenland studies, for he has given us no more 
striking and significant fact than the relative facility with 
which the ice shears along innumerable lines of debris: and 
the evidence is conclusive that the ice, to a large extent, slides- 
over the enclosed debris, instead of dragging it along, as it 
would if the debris were firmly frozen into the ice. Granting, 
however, that the frozen soil would be more rigid and indif- 
ferent to gravitative stresses than the clear ice above it, the 
fact remains that the unfrozen soil at the base is more yield- 
ing and plastic than either; and hence, although we may rea- 
sonably conceive definite shearing planes as distributed 
through a considerable thickness of the ice-sheet, the lowest 
plane, and the true base of the ice-sheet, will still be at the 
lower limit of frost. 
So far as I can learn, everywhere within the range of ob- 
servation modern glaciers are either sliding over their ground 
moraines, or they rest directly upon the firm bed-rock: just 
as many rivers, at ordinary seasons, flow quietly over deep 
beds of gravel and stones or over bare ledges. In both cases 
active erosion of the bed-rock floor is nearly at a stand still: 
for where it is not protected by the stagnant debris the ice or 
water are not well suppled with stones, without which they 
can do little. In the case of the river, periods of flood, or of 
greater fall or volume in the past, must be postulated to set 
the detritus in motion and account for the effective erosion of 
their channels. Similarly for the glacier, the entire ground 
moraine must be set in motion. Increased thickness anil ve- 
locity of the ice would probably tend to accomplish this. But 
