ANTARCTIC ATMOSPHERIC CIRCULATION 
is a comparatively superficial effect attributable to the 
surface cold.”’ 
Others, however, postulated a vast continental anti- 
eyclone around which rotate a succession of huge oceanic 
eyclones [103, 105]. To this anticyclone Hobbs gave a 
unique character and a distinctive name [104]: a ‘‘fixed 
glacial anticyclone, modelled on the form of an hour- 
glass, ...roughly centered near the South Pole,” in which 
a continuous downdraft supplies air which flows out 
radially in bursts of strong winds. The vast snowfall 
presumably required to maintain the icecap against 
its own slow outflow and the outward sweeping of the 
winds comes from “the ice spicules of the cirrus and 
other closely related cloud forms . . . adiabatically vapor- 
ized in the downdraft and reprecipitated as they 
approach the glacier surface,” which is radiatively 
cooled. This precipitation mechanism did not appear 
quantitatively sufficient to most other meteorologists 
studying Antarctica, and the analogous permanent 
anticyelone of Greenland is not accepted in toto by 
many meteorologists today.” 
Meinardus [107] interpreted the strong easterly winds 
common at the Gauss, frozen in for a year some 40 
miles off the coast at 90°H, on the 1902-4 German 
expedition, as due to passages of cyclones to the north, 
and felt that “If there is an Antarctic anticyclone it 
can exist only in the inner portion.” 
The Antarctic anticyclone, so much discussed in the past, 
is a pressure distribution peculiar to the lower atmospheric 
strata only, appearing with distinctness only in the sea-level 
pressure distribution. On the other hand the low Antarctic 
temperature must produce such a rapid vertical decrease in 
pressure that above a certain level the Antarctic pressure 
must be lower and not higher than that of surrounding regions. 
Thus the sea-level anticyclone must be overlain by a cyclone, 
the so-called ‘polar whirl” in the general circulation of the 
globe. 
From available information, Meinardus estimated 
that this transition occurs at about 2 km above sea 
level, and that most of the continental interior is higher, 
at the level at which cyclonic conditions prevail, with 
moisture-laden air flowing in east of the Ross and Wed- 
dell Seas, and a cold dry outflow on their western sides. 
This concept was attacked by Simpson, meteorologist 
of the last Scott expedition of 1911-13, who insisted 
[112] that a continental mass like Antarctica would 
establish its own regime above its surface, and not 
penetrate into the circulation like an isolated moun- 
tain peak: 
A statement of the general air circulation over the Antarctic 
is now quite simple. Over the snow-covered surface of the 
Antarctic whether at sea-level or at the height of the plateau 
radiation is so strong that the air is abnormally cooled es- 
pecially in the layers of air immediately above the surface. 
This cooled air is heavier than the surrounding air and there- 
fore the pressure increases from the exterior to the interior 
of the Polar area; in other words the pressure distribution is 
2. Consult’ ‘“‘Some Climatological Problems of the Arctic 
and Sub-Arctic”’ by F. K. Hare, pp. 952-964 in this 
Compendium. 
919 
anticyclonic and the air motion is in general outwards. Above 
each anticyclone a cyclone forms on account of the relatively 
rapid vertical pressure change caused by the cold dense air. 
These cyclones convey air from higher latitudes over the 
Polar region and supply the air which passes outwards near 
the surface. In the normal steady state the air circulation 
takes place slowly and th descending air is warmed up 
dynamically so dissolving cloud and giving clear cloudless 
skies, thus accounting for the decreasing cloud amounts ob- 
served as one penetrates the Antarctic. 
The clear skies in their turn facilitate radiation as also does 
the small absolute humidity of the air. In consequence the 
air and the snow surface become abnormally cold and there 
is a great tendency to the formation of temperature inversions 
especially in the lower atmosphere .... The abnormally cold 
surface air is forced upwards in these [forced ascending] 
currents, rapidly cooled in the ascent, and the water obtained 
is precipitated as snow, which when combined with the high 
surface winds produces the typical Antarctic blizzard. 
Simpson also disputed Meinardus’ interpretation of 
the Gauss winds, considering them too constant in 
force and direction to be due entirely to cyclonic pas- 
sages. His general model was followed, in preference to 
that of Meinardus, by Barkow, meteorologist on the 
Deutschland during its year-long drift in the Weddell 
Sea, and Knoch, editor of the report [101] after Barkow’s 
death: 
The circulation over the Antarctic continent is dominated 
by an anticyclonic cap of air which probably flows outward 
in a series of waves moving in a south to north direction. 
Above this cap of cold air there is a cyclonic stratum connected 
with the circulation of temperate latitudes and in this 
wandering depressions travel in a west to east direction. The 
two air-systems mutually influence each other; on the whole 
the lower system dominates the upper at least so far as 
atmospheric pressure is concerned. The surface winds are 
dominated on the whole by the lower system for the most part 
on the continent and in less degree as the distance from land 
increases. In the Cirrus region, and especially in the strato- 
sphere, cloud movements appear to show that there isa 
current of air moving right across the continent from the 
Indian Ocean to the region of the West Antarctic. . .. ; 
However, Shaw [111] evened the score by abandoning 
his thin anticyclone of twenty-five years earlier in 
favor of Meinardus’ model rather than that of his own 
colleague, Simpson, or that of Hobbs: 
The circulation would correspond with a permanent anti- 
eyclone covering the continent if the space which the anti- 
cyclone should occupy were not already filled by a huge mass 
of land.... We should not expect to find evidence of the 
so-called [glacial] anticyclone on the top of the land-mass; 
the conditions there may be the reverse of those which are 
indicated by the margins at sea-level. 
In turn, Kidson, although opposing Simpson’s theo- 
ries on some points (see p. 927), accepted his surface 
anticyclone rather than Meinardus’ peripheral sea-level 
anticyclone or Hobbs’s glacial anticyclone. In his dis- 
cussion [75] of the results of Mawson’s 1911-13 expedi- 
tion (he had previously edited and analyzed the data of 
Shackleton’s 1907-9 expedition), he said: 
