Jökull - 01.12.1953, Side 30

scale was evident everywhere. The bottom mel-
ting offers some support to the theory of
Carruthers (1943).
ARTIFICIAL MANTLING OF ICE
BY ROCK DEBRIS.
Tests were made on three areas of clean
glacier ice which were artificially covered by
dirt to a thickness of 2, 4 and 7 millimetres. The
dirt had been collected from many dirt cones
and the areas of the ice mantled were each one
metre square. Two half-inch diameter stakes
were drilled into two quadrants of each square
metre of dirt covered ice. One similar stake was
put near two sides of each square metre under
observation. The means of the ablation shown
at these test stakes indicated that the ice with
only a 2 mm. thick dirt cover ablated at a
slightly greater rate than that for clean ice,
probably due to its decreased albedo. In sun-
shine the ablation of the 4 and 7 mm. covered
ice was lower which left those metre squares
„proud“ of the ice surface. In very light rain
the 7 mm. covered ice suffered less ablation.
The other two test squares ablated almost the
same amount, the more thinly covered ice being
sligthly greater during the first day of fine rain.
These tests required such care to achieve reli-
able figures that it was not possible to carry out
tests save during a continuous period of good
weather. Rain intervened and/or wind or water
born grit contaminated the test and control
areas so that the tests had to be abandoned.
Insufficient data were collected to permit an
accurate conclusion but the thickness of a
protective dirt layer would seem to be (in the
Icelandic climate) between 4 and 7 millimetres,
though this would not survive one season due
to fluvial action removing it.
DIRT CONES.
The dirt cones were of all sizes from centi-
metres to 10 metres in height. All had a mantle
of dirt from 1 to 5 cms. thick, though the
majority had slightly less than 1 cm. on the
steepest slope. The angle of the slope varied
between 38° and 51° but a mean of the cones
considered typical and fully developed was 48°.
The ice core of the dirt cone was more dense
than the surrounding glacier ice but though
very dark in appearance had little grit present
in the ice.
Many of the cones near the nunataks of Esju-
fjöll had a vertical dirt layer sandwiched in the
ice core. This dirt centre clid not always have
an outlet to, or continuation with the outer
dirt mantle so that it was not possible to assume
that the dirt of the cone mantle had come from
the inner frozen dirt plug. In some of the cones
cut through for examination the dirt plug at
the centre was irregular and bent slightly down
glacier. Many of the dirt plugs could be lined
up with dirt bands in the glacier, some of which
had dirt cones dotted along their length.
Many of the dirt cones had patches of the
thick dirt in horizontal layers indicating fluvial
deposition of the dirt in crevasse or stream-cut
gully and subsequent melting out of the de-
posited dirt (Swithinbank 1949). Near the edges
of all streams a thin line of deposited dirt could
be seen, sometimes already producting tiny dirt
cones, but unless the stream were near the
moraine where abundant dirt loaded the melt-
water streams, there seemed few dirt cones since
there was insufficient dirt present to survive the
runoff of meltwater long enough to create its
own ’island' and thus divert the meltwater near
the base of the dirt cones.
The main supply of dirt is from stratification
planes, dirt bands, either melting back and
leaving the dirt on the surface or actually bring-
ing the dirt along the plane by differential
movement (though this latter is in doubt after
recent unpublished work).
The wind probably deposits the fine grit
initially but transport and redistribution is al-
most entirely by water action, then the cone-
bulding is similar to that outlined by Mr. Lewis
(1947), the sun melting the cléar ice but to a
much lesser degree the dirt mantled ice.
The mechanics of this procéss will evolve
after a full analysis of the observations on micro
climate ancl ablation.
ICE DENSITY AND ENGLACIAL DIRT.
The ice features apparently formed by dirt
were examined to find the extent to which the
dirt was incorporated in the ice or remained a
surface feature. Where it was possible to have
access to ice layers below the surface, samples of
ice were taken and measured for density and
dirt content.
Method. At all places where ice was sampled
the surface was chipped back and cleaned to
28