Jökull - 01.12.1953, Blaðsíða 31

ensure taking debris which was in situ and not
contaminated by drainage. Pieces of ice were
chipped and put in a can insulated by straw
in a rucksack and carried to the nearest camp.
The planes of fracture of the ice pieces could
have followed cavities in the ice so that the
method of measuring density was open to sorne
doubt. As much care as possible was taken to
avoid this but doubtless slight error is present.
A wide measuring cylinder containing a flat
wire coil and stem was half filled with water
and a reading taken. The lumps of ice from a
sample were dried and immersed in the water,
holding the pieces submerged by wire and
a second reading was taken immediately. The
water temperatures were only a little abovo
freezing point so that very little melting occ-
urred in the few seconds between adding the
ice to the cylinder and taking the second read-
ing. The ice was then allowed to melt and a
third reading taken. From readings the volume
of ice and the volume of water produced by the
ice and the density was calculated. The water
was filtered through a weighed filter paper
which was weighed again on return to Eng-
land.
Results. The percentage (by weight) of grit
present is tabulated for four different depth
zones: surface, surface to 2 metres depth, 2
metres to 6 metres depth and 6 metres to 10
metres depth. The figures are given correct to
two places of decimals. Where the weight of
grit present is too small to give a figure greater
than 0.01%, the reading is shown as 0%.
There are insufficient observations to state
result below the surface with any confidence
for the range of readings is such that large
sampling is necessary. The points on the graph
(Fig. 5) have been given a size proportional to
their reliability, i. e., number of samples taken.
The observations are in two groups: —
"Clear Ice” meaning ice apperently free from
grít-
”Dirt-laden Ice” meaning ice dominated in
apperarence by tlie presence of grit, such as
moraine, dirt bands and lenses, bottom layer
(sole).
The mean density -of surface ice was 0.88 at
Esjufjöll (650 metres) and 0.92 on the snout
(50 metres) 20 kilometres distance. (Bunsen gives
0.9168 (Barnes 1928), Roth 0.9168 and Vincent
0.9160 (Kaye and Laly 1948) for the density of
Eig. 5. The content of grit in the glacier at
different depth zones.
Aur (leir, sandur, aska) i jökulisnum í mismun-
andi djpi.
ice. The density of 0.92 above agrees very well
with these figures and is some small check on the
field method used for measuring density. It is
not intended to imply that the increase in den-
sity is gradual over the 20 kms. of the falling
glacier surface. The density of 0.92 would pro-
bably be found over most of the surface of the
lower glacier. Various sizes of grit particles.
were present at all depths.
CONCLUSIONS ON ENGLACIAL DIRT.
There is a remarkably small percentage of
grit present even in the dirtiest ice. The amount
of grit present within the ice increases from the
surface down (fig. 5). The bottom layer (sole)
of the glacier, though black in appearance has
little more dirt present than ice at the dirt
bands. The penetration of grit from a heavy
dirt laden surface, dirt band or lens within the
ice falls off very rapidly away from the body of
the grit.
There is no relationship between size of par-
ticle in sand lenses, planes etc., and the depth.
29