924 
Of the eleven air masses affecting South America, 
distinguished by Serra and Ratisbonna [42] in their 
pioneer study, “the originally continental mass, A, 
changes into At (transitional Antarctic) and finally into 
Pm (polar maritime)” as the air is humidified and 
warmed in leaving the continent, generally west of 
the Weddell and Ross Seas. In winter, Pm stagnating 
over Patagonia becomes continental polar air (Pec); 
active polar masses (Pk) moving northward were dis- 
tinguished from returning polar masses (Pw) returning 
poleward over South America. 
Although basing his discussion to some extent on 
this study, Coyle [27] did not use these names or in- 
dicators, preferring to consider currents rather than 
air masses. Apparently without reference to this or 
other previous Southern Hemisphere classifications, Ro- 
bin [38] identified three air masses in the area south of 
Cape Horn: ‘‘Antarctic air (A)...over the conti- 
nental and barrier ice of Antarctica, Polar maritime 
air (Pm) ...in the southern half of the westerly wind 
belt of the southern hemisphere, [and] a more temperate 
type of maritime air (7’m) was also occasionally experi- 
enced, this air coming from a more northerly portion of 
the westerly wind belt.” The antarctic front separated 
A and Pm air, and was strongest when it lay along the 
edge of the pack ice. 
Generalizing at second hand, Haurwitz and Austin 
[183] concluded: 
The air masses of the continental antarctic resemble the cP 
air of northeastern Asia. The principal differences are the 
high frequency of steep lapse rates near the stirred surface 
layer and the very cold temperatures at all levels over the 
southern continent. During summer, antarctic cP air is still 
very cold and, except for a gradual rise in temperature, 
appears to resemble the winter air mass. Throughout the 
year, low specific humidities are characteristic of the southern 
cP air masses. ... 
Because of the high topography, mP air masses affect only 
the periphery of Antarctica, where the cyclonic circulation is 
reasonably intense in summer and winter. Consequently, the 
summer air mass probably lacks the stability of the Northern 
Hemisphere air mass. Therefore, the lapse rate is likely to 
be approximately moist adiabatic throughout the year. Be- 
sides increasing the temperature and cloudiness of the coastal 
area, mP air must also supply the moisture for the light 
precipitation of the interior plateau. 
Six air masses were deduced for the Southern Hemi- 
sphere by Gentilli [9] by applying to Shaw’s classic 
maps of mean temperature and pressure [111] the air- 
mass criteria in use in the Northern Hemisphere: 
iAirbiMass) eile heakkerctiecaenarciecie tn ote A Pm Tm Eq Te Pe 
Area of source in summer 8 20 34 34 ? 
millions of square miles \winter 11 30 31 16 i 
The greatest contrasts, and hence most active fronts, 
occur between the polar maritime air (found between 
40° and 68°S in summer and between 34° and 65°S in 
winter) and tropical continental air (found typically in 
Australia in summer). Polar continental air occurs only 
in South America in winter, equatorial air is formed as 
far as 24°S in summer, and superior (S) air is presumed 
to descend under certain conditions. 
POLAR METEOROLOGY 
While conflicting in notation and definition, these 
various air-mass classifications can be harmonized, De- 
spite Kidson’s warning [75] that there is no “‘air mass to 
which the application of the term ‘antarctic’ could be 
justified,” there seem to be three such masses: conti- 
nental antarctic (cA) in the interior, transitional ant- 
arctic (nA) along the coasts, and maritime antarctic 
(mA) over the pack ice and the ocean south of the 
pressure trough, convergence zone, or antarctic front. 
To the north is a vast mass of maritime polar (mP) air. 
The symbols of various authors apparently fit this 
classification as follows: 
cA nA - mA mP 
TBM. . pcccencveoadoecoooe cP cA mA mP 
Serra and Ratisbonna......... A A At Pm 
Riobintae ae ar een eens A A Pm Tm 
Gentiles. ores A A A Pm 
Haurwitz and Austin.......... cP cP mP. mP. 
Proof of the existence of these masses (perhaps nA is 
really cAk), and better definitions and criteria, must 
come from meteorologists drawing regular synoptic 
maps of extensive antarctic and subantarctic areas. 
Storms. Waves formed on the antarctic front west of 
the Ross and Bellingshausen Seas, the Highjump aerolo- 
gists found, moved eastward and occluded, often merg- 
ing with deeper systems which had approached from 
the northwest. Some antarctic frontal waves dissipated 
near the Balleny Islands, others curved southeastward 
into the Ross Sea and dissipated before reaching its 
southeastern corner; in the Bellingshausen Sea they 
merged into old Pacific Ocean occlusions. These storms 
were minor compared with those from the north: 
Wave formations developing on the polar fronts of the 
Atlantic, Indian, and Australia-New Zealand areas occlude 
and approach the continent usually on an east-southeast 
track [at 20 to 25 knots in January, 35 to 45 in March]. 
During periods of low circulation index these lows, upon 
approaching the continent, tend to stagnate in. . .semi-perma- 
nent low pressure areas. . .in the Ross Sea, near 120°H along 
the edge of the ice pack, in the western part of the Mac- 
Kenzie Sea [60°H], and in the eastern part of the Weddell 
Sea. During high index conditions these lows stagnate only 
intermittently in these areas. ... 
When a migratory low pressure center stagnated in the 
Bellingshausen Sea, a new center developed off the northern 
tip of the Palmer Peninsula,. . .deepened slowly but moved 
rapidly off to the eastward.... A northward outbreak of 
cold continental air [occurred] along the east coast of the 
Palmer Peninsula and behind the low as it moved into the 
Weddell Sea. 
These summertime conclusions are reinforced by 
those reached during the ensuing months by Robin [38], 
meteorologist in charge of the British station at Signy 
Island, South Orkneys, during 1947—48 (and currently 
with the Norwegian-British-Swedish expedition): “Cy- 
clones observed in the Antarctic front were either well- 
developed cyclones which approached Graham Land 
and from the west, or were young cyclones which formed 
near the north tip of Graham Land, the latter being 
more in evidence in late autumn and spring.” Both types 
moved rather rapidly (about 600 miles per day) east- 
