breadth made with a pocket comparator, which, allow-
used a method of least-circle diameters to measure grain
ing for the 7.5 magnification of the contact print, per-
sizes in rubbings of ice. Alhmann and Droessler (1949)
mitted measurements to the nearest 0.05 mm. The mean
measured the shortest and longest axes of ice grains to
cross-sectional areas of crystals (mm2) were then con-
obtain average cross-sectional areas. Schytt (1958) and
Stephenson (1967) employed much the same technique
verted to root-mean-square diameters (mm). This mea-
to measure crystal sizes in Antarctic firn. These meth-
surement method was preferred to the intercept method
ods all suffer from the problem that a section seldom
because of the substantial corrections needed to account
cuts crystals at their maximum (true) diameter, leading
for intersected inclusions, both air bubbles and brine
to underestimates of the true sizes of crystals. Varia-
channels, which occur abundantly in accreted ice. With
tions in particle size also complicate the problem, which
the exception of one or two thin sections, crystal size
is further compounded by the lack of knowledge con-
measurements reported here were restricted to sections
cerning the distribution of particle sizes within the origi-
cut parallel to the accreting surface.
nal sample. Because under-sized cuts of crystals in a
Inclusion size measurements
thin section exert a disproportionate influence on the
average particle size, we decided to restrict measure-
We also attempted to measure air bubble dimensions
ments of crystal size in accreted ice to the 50 largest
in the majority of the thin sections, despite the intrinsic
crystals in a given area of the section (Gow 1987).
difficulty of distinguishing air bubbles from brine pock-
As applied to the samples of accreted ice obtained
ets trapped in the ice. Although this distinction between
on the cruise of the USCGC Midgett, an area of 10 8
these two principal types of inclusions trapped in the
cm on the enlarged print (17.5 12.5 cm) was used to
ice is difficult to establish with certainty, those we iden-
select the 50 largest crystals. This selection of crystals
tified as air bubbles tended to be found either along
generally constituted less than 25% of the total number
grain boundaries or at multigrain intersections. We con-
of crystals in the 10- 8-cm area of the contact print.
centrated our measurements of bubble size on spheri-
Actual values of the cross-sectional areas of crystals
cal inclusions, which we believe are more likely to rep-
were calculated from measurements of length and
resent air bubbles than brine pockets. Small rounded
Table 3. Crystal and bubble size measurements.
Thin
Root mean
section
Mean crystal
square
Mean bubble
Sample
Accretion surface
orientation
area (mm2)
diameter (mm)
diameter (mm)
F1
Deck
Horizontal
0.63
0.79
0.26
F2
Deck
Horizontal
0.45
0.67
0.28
F3
Superstructure vertical surface
Vertical
0.60
0.78
0.20
F4
Superstructure vertical surface
Vertical
0.38
0.62
Not measured
F5
Superstructure vertical surface
Vertical
0.40
0.63
0.23
F6
Deck hatch
Horizontal
0.59
0.77
0.25
F7
5-in. gun forward face
Vertical
0.62
0.79
0.26
F8
5-in. gun forward face
Vertical
0.46
0.68
0.25
F9
Superstructure vertical surface
Vertical
0.66
0.81
0.29
M3
Superstructure vertical surface
Vertical
0.58
0.76
0.27
M5
5-in. gun forward face
Vertical
0.31
0.56
0.27
M6
5-in. gun forward face
Vertical
0.53
0.72
Not measured
M8
Deck near 5-in. gun
Horizontal
1.33
1.15
0.29
M9
5-in. gun vertical face Port side
Vertical
0.50
0.71
0.31
M10
5-in. gun vertical face Port side
Vertical
0.58
0.76
0.24
M12
Icicles, 5-in. gun mount
Horizontal
0.73
0.85
0.27
M12
Icicles, 5-in. gun mount
Vertical
2.66*
1.63*
Not measured
0.83†
0.91†
M13
Deck near 5-in. gun
Horizontal
3.53*
1.88*
0.25
1.06†
1.03†
M14
5-in. gun forward face
Vertical
1.05
1.02
0.24
*Large crystals.
†Small crystals.
14
to contents
Previous Page
