936 
index periods several cells may exist. Each of these 
patterns may thus be found in different years, or they 
may appear in succession as the hemispheric flow pat- 
tern changes. Consequently, isolated observations in 
different years may indicate conflicting results. 
Cyclones travel in families south-southeastward 
across the three great oceans to stagnate in the seas 
which indent the coast to greater or lesser degree: Ross, 
Amundsen, Bellingshausen, Weddell, and “‘Mackenzie”’ 
(70° to 80°E). These cyclones may go inland when the 
zonal index is low and there are several anticyclonic 
cells, especially in summer, but they are confined to 
the coast when the index is high and there is an exten- 
sive central anticyclone. Those which do go inland pro- 
vide some precipitation in the interior, but fill rapidly 
as they leave open water and lose energy. 
Exchange. Expansion of the atmosphere in summer 
by the warming of both troposphere and stratosphere 
causes a net outflow of air from Antarctica. This out- 
flow occurs chiefly in those areas where anticyclonic 
lobes or cells form during moderate- or low-index con- 
ditions: in the southeastern corners of the three oceans 
washing Antarctica, or roughly 10°W to 20°H, 130° to 
160°H, and 120° to 100°W. A fourth area is along the 
Palmer Peninsula (60° to 70°W), for topographic 
reasons. 
Outflow im these areas is reinforced by meridional 
exchange, which occurs throughout the year at a rate 
inversely proportional to the zonal index. It is not a 
continuous flow, but occurs as outbreaks of anticyclones 
from the interior; as they progress outward, these anti- 
cyclones cause ‘‘pressure waves.” Their formation, m 
turn, may be associated with the poleward motion in 
other longitudes of warm anticyclones of subtropical 
origin, so that the waves at times appear to have 
travelled across the entire continent. The warm anti- 
cyclones from the subtropics provide some of the com- 
pensating air influx. 
Even under very-low-index conditions, when the 
meridional exchange between antarctic and subantarctic 
regions is greatest, 1t is very slight compared to the 
amount of air exchanged between continents and oceans 
in the rest of the world. The first report [45] on the ex- 
treme cold of the winter stratosphere over Little Amer- 
ica suggested that “lack of circulation between Antarc- 
tica and lower latitudes would permit air there to cool 
unmolested in the winter until it attained lower tem- 
peratures than anywhere else on earth,” and that the 
subantarctic westerlies ‘“may largely confine the ant- 
arctic air.” 
This hypothesis and those of Hobbs, Meinardus, 
Simpson, and others were doubted by Ramage [64] 
insofar as they inyolved a “disconnection between ant- 
arctic and temperate circulations,” because “before 
such a circulation theory, unparalleled anywhere else 
in the world, can be put forward successfully, detailed 
and accurate observations must be made in the region 
where the discontinuity is thought to exist.’’ However, 
he conceded that “‘a temporary interruption of atmos- 
pheric interchange with lower latitudes would be suf- 
POLAR METEOROLOGY 
ficient to cause the tropopause inversion to disappear 
in winter.” 
Strong support for the hypothesis of a general lack 
of air exchange between Antarctica and the rest of the 
world is provided in an “exclusion principle” postulated 
by Starr and reported by Willett [140]: ‘‘A strong cir- 
cumpolar cyclonic vortex or zonal westerlies should nec- 
essarily repress” transport of heat and kinetic energy. 
Even at their weakest, the subantarctic westerlies are 
far stronger than the Northern Hemisphere westerlies 
[8], so that by this principle the meridional flow in high 
southern latitudes should be far less than in the north 
polar region—as Bjerknes and others had suggested 
much earlier. After several months of drawing and 
analyzing complete Southern Hemisphere maps, Wil- 
lett and his staff felt tentatively that “real polar out- 
breaks of the large-scale type involving the equatorward 
movement of large polar anticyclones, such as we ob- 
serve in the Northern Hemisphere, do not occur on the 
same scale in the Southern.’” 
Slight as this exchange is when compared to that of 
the Northern Hemisphere, variations in it affect the 
general hemispheric circulation. Seasonal changes in the 
pressure field induce less meridional exchange in some 
seasons than in others. The flow apparently is least in 
late spring and early summer (October and November) 
because of the rapid heating at all levels, and isobars 
then should be most nearly zonal without pronounced 
lobes on the vertically expanding continental cyclone. 
Consequently, the zonal westerlies are strongest in 
spring, as Gabites [48] has found. This seasonal varia- 
tion may account for the early summer cyclonic weather 
encountered by Ellsworth in an area later postulated as 
having an anticyclonic wedge or cell im late summer. 
Implications. The general hypothesis outlmed above 
differs from earlier versions of the polar anticyclone 
postulated by others in allowing two types of modifica- 
tion: seasonal and circulational. In winter it is similar 
to Simpson’s laminar model, and during low-index 
periods not markedly different from Hobbs’s glacial 
anticyclone—except that no artifices are needed to 
provide central precipitation since recent evidence in- 
dicates that little occurs. In summer it follows the 
cellular or lobar pattern suggested by Lamb and the 
Highjump aerologists, recognizing that their limited 
period of analysis was one of moderate westerly flow. 
And in recognizing the penetration of warm air masses, 
it agrees with Palmer’s contention that the anticyclone 
is not a permanent feature but a model picture, though 
somewhat more than a statistical abstraction. The only 
model with which it is completely at variance is the 
peripheral anticyclone of Meinardus, which has been 
vitiated by more recent findings that Antarctica is not 
a simple elevated plateau. 
Although the processes suggested above explain the 
upper-air temperatures and the surface pressure regime 
at Little America, they seem to offer no ready explana- 
tion for the three progressively colder temperature 
7. Private communication, 4 November 1949. 
