ANTARCTIC ATMOSPHERIC CIRCULATION 
ward and not, as previously presumed, southeast into 
the Weddell Sea. Both young and old cyclones occur in 
groups of twos and threes followed by an “‘injection of a 
cold air mass over the South American continent” 
which may reach Brazil, as others [27, 32, 33, 41] have 
found. 
Mean monthly pressure charts for the same area, 
based on unpublished observations of Rymill’s 1935-37 
expedition [40] and available reports from South 
America and the South Orkneys, indicated to Mirrlees 
[86] either that the main track of depressions is farther 
south than previously described or that there may be 
actually two such tracks. 
Two such separate tracks were found in the south- 
east Indian Ocean by Gibbs [73] who studied the reports 
of Palmer, Kidson, and Operation Highjump, and 
6-hourly maps of the southern Indian Ocean during the 
first year (1948) of complete weather observations at 
Marion, Heard, and Macquarie Islands. Gibbs con- 
cluded that depressions in this area originate north of 
45°S, as wave cyclones in winter and spring but fre- 
quently as frontless cyclones in summer and autumn. 
They travel southeast, intensifying between Marion 
and Heard, occlude and retard after passing Heard, 
become old cyclones by the time they pass Macquarie, 
and “probably move to the Ross Sea area, where they 
weaken and finally disappear’; a second track runs 
from western Australia to Macquarie. 
The Antarctic front is generally impelled northward into the 
latitudes of the westerlies by the approach of one of the great 
depressions to the vicinity of the Antarctic coastline. Subse- 
quently on reaching lower latitudes wave development may 
occur on it and the waves so developing may become one of 
the great southern depressions. 
Substantially similar concepts of fronts, air masses, 
and weather processes over the oceans south of Australia 
and New Zealand were reached independently by Japa- 
nese meteorologists? on board whalers in recent years. 
Circulation. In Palmer’s concept of Southern Hemi- 
sphere circulation [18], cyclones are formed in groups off 
the east coasts of continents by the interaction of polar 
maritime air of the westerlies and tropical air flowing 
southward around the three subtropical anticyclones. 
These cyclones travel southeastward, first in the gen- 
eral circulation around the highs and then with the 
westerlies} developing, occluding, and finally filling as 
they approach Antarctica. New depressions form as 
waves on their fronts and continue somewhat farther. 
This model implies that cyclones which actually reach 
Antarctica are the second or third generation of systems 
which formed originally in low latitudes several thou- 
sand miles to the northwest 
Frontal zones were of less interest to Lamb [81, 82, 83] 
4. Two Japanese reports were received after completion 
of the bibliography: Oceanographical Section, Central Meteor- 
ological Observatory, ‘‘Report on Sea and Weather Observa- 
tions on Antarctic Whaling Ground (1947-48).’’ Oceanogr. Mag. 
(Tokyo) 1: 49-88 (1949); ‘Report... (1948-49). Ibid., 1: 142- 
173 (1949). 
925 
than the general circulation of the Southern Hemisphere. 
By the end of his three summer months in subantarctic 
waters he was drawing usable maps for the entire 
hemisphere, although he was unable to obtain reports 
from whalers other than those in his own group or from 
the contemporaneous Operation Highjump. Analyses 
were based on North Atlantic experience and the postu- 
late that “stable, warm anticyclones exercise a con- 
trolling influence on the circulation patterns far beyond 
their own limits.” 
These 1947 summer maps showed groups of occlusions 
symmetrically located, with a pattern of six such groups 
spaced 60° of longitude apart (and six intervening sub- 
tropical anticyclones) characteristic of low-index condi- 
tions, four groups 90° apart with higher zonal index, and 
a hint that as winter approached the pattern would 
reduce to three groups 120° apart. 
Earlier, Kidson [75] had found ‘‘a strong suggestion 
that the moving anticyclones originate as some periodic 
function of the upper atmosphere, or of the atmosphere 
as a whole, which is continually tending to produce them 
at intervals of about 45° of longitude.” This was based on 
his daily charts of 1912-13, when the mean speed of 
anticyclones, 8.8° per day, ‘“‘wascertainly above normal” ; 
“more anticyclones passed during the year than usual,” 
and their “mean latitudes were ... unusually high.” Such 
45° spacing implies a hemispheric pattern of eight 
anticyclones and intervening occlusion groups and, by 
extension of Lamb’s conclusions, an unusually low zonal 
index with greatly increased meridional exchange of air. 
The period from 2 February 1912 to 31 January 1913, 
studied by Kidson, may well have been different in 
general characteristics from the late summer of 1946-47, 
on which Lamb based his analysis. In the Northern 
Hemisphere, the latter period was quite abnormal, 
Namias [137] found pressures in the Arctic basin some- 
what below normal during December and January and 
greatly above normal in February and March. Through- 
out this winter, the zonal index was below average and 
the total mass of air over the Northern Hemisphere 
above normal. On the other hand, during the 12 months 
studied by Kidson, the Northern Hemisphere zonal 
index varied from slightly below average to exception- 
ally above average; on the whole it was nearly 0.6 
m sec! stronger than average, or about one-third 
stronger. 
These departures from average of the Northern Hemi- 
sphere zonal index are directly opposite to those sug- 
gested by Lamb’s and Kidson’s analyses, and it is 
tempting to think that the indexes vary in opposite 
senses in the two hemispheres. If the negative correla- 
tion between zonal index and total atmospheric mass 
north of 20°N, found by Brier [128], represents inter- 
hemisphere transfer, a complementary relation of the 
hemispheric indexes is to be expected; but it may merely 
represent mass transfer from low to high latitudes. 
Opposing any inverse relation between the zonal in- 
dexes of the two hemispheres is the belief that all climatic 
changes, from ice ages to unusual years, are due basi- 
cally to similar changes in the atmospheric circulation; 
since the larger climatic variations were simultaneous 
