TROPICAL METEOROLOGY 
the greater precision that would be given to local fore- 
casting in the tropics by a complete understanding and 
description of the diurnal variation of orographical 
organized cumulonimbus well merits military and com- 
mercial support of such investigations. 
It is difficult to deny that, as far as the most general 
features of the monsoons are concerned, the climato- 
logical school seemed to have analyzed the mean data 
and to have correctly assigned astronomical and oro- 
graphical causes to the variations. But it is probable 
that the monsoons are far more complex in their details 
than that school supposed. There are several reasons 
for this conjecture. First, the so-called monsoon lows 
over the land, whatever may be their properties in the 
mean, on synoptic maps show day-to-day variations 
which could be attributed to the westward movement 
of individual cyclonic circulations through the area. 
Moreover, where the conformations of the continents 
are favorable, it can be observed that these minor 
cyclones come in from the sea, fully formed. They can- 
not under any circumstances be regarded as “heat 
lows.” These local circulations are known in India, in 
Australia, in Indo-China and China, and in Hast Africa. 
Second, and this is probably correlated with the facts 
first mentioned, the wartime synoptic observations con- 
firmed what was after all known locally for many years: 
that the weather in the Asiatic monsoons is far more 
variable than popular and textbook accounts would 
lead one to suppose. It does not rain continuously for 
three or four months mm the wet season. There are fre- 
quent spells, lasting sometimes for days, during which 
there is little precipitation and sometimes little cloudi- 
ness. Third, there are, on certain longitudes, quasi- 
permanent low-pressure areas which migrate to and 
away from the equator, lagging behind the sun in the 
same manner as the monsoon lows, and having, on their 
equatorward borders, belts of winds with a westerly 
component; they are situated, however, not over the 
land but over the ocean. There is a very persistent 
mean low of this type which migrates from eastern New 
Guinea to the eastern borders of the Philippines and 
back. There is another west of Mexico and Central 
America. If these lows had been situated over a land 
mass, we would have no hesitation in calling them mon- 
soon lows, and in accounting for their movements and 
the distribution of the meteorological elements about 
them im the usual manner of the climatological school. 
We ought, perhaps, to follow up these climatological 
hints and, combining them with an intense synoptic 
and dynamic research into subequatorial semiperma- 
nent lows in general, improve our understanding and’ 
prediction of monsoon weather. Largely because of 
historical accidents, research has been concentrated in 
the trade-wind zones of the Pacific and Atlantic. The 
time is now ripe for a more intensive investigation of 
those parts of the tropics that show large seasonal 
variations in the meteorological elements. 
We should, however, not neglect the oceanic regions. 
Particularly we should not completely reject the classi- 
cal climatological explanation of the trades, as Rossby 
has done. It is possible that that explanation is still 
873 
valid for the longitudes where the trade is really a 
trade, 2.e., where it shows, in the lower level, great 
persistence, a large meridional wind component, a com- 
paratively strong meridional temperature gradient and, 
aloft, the typical antitrade, evidence of persistent and 
marked subsidence, with a low and strong trade-wind 
inversion. These conditions obtain, as has been pointed 
out, in the eastern and southeastern branches of the 
great quasi-permanent anticyclones. Downstream, to- 
ward the west, however, factors other than those in- 
voked by the climatologica] school may dominate the 
motion. Here, Rossby’s new explanations may hold. 
Above all, our survey shows, the greatest need is for 
more complete data, especially from the upper air. At 
present the distribution of observing stations in all 
latitudes is largely determined not on scientific, but on 
economic and military grounds. This, it is to be pre- 
sumed, will not be changed in the foreseeable future. 
Nevertheless, it seems to be accepted that improvement 
in practical forecasting for periods longer than forty- 
eight hours depends on improvement in our empirical 
and theoretical knowledge of the mean circulation of 
the earth’s atmosphere. How this knowledge can be 
built upon an observing network that is half-way ade- 
quate only in North America and in Europe is some- 
thing that, to my knowledge, has never been explained. 
In particular, if scientific meteorologists are content to 
advance models of the general circulation based upon 
the North American atmospheric section and theoretical 
explanations of this model based upon an obsession with 
the circular vortex, ignoring the tremendous gaps in 
our knowledge of the meteorology of the tropics (over 
half the troposphere), of the high latitudes in the South- 
ern Hemisphere, of Asia, even of Central America—then 
they cannot blame those that hold the purse strings for 
supposing that all is well with the present setup. The 
fact that acquaimtance with new data is enough to cause 
a complete volte-face in the theoretical work of an out- 
standing leader in dynamic meteorology, at least so far 
as the tropics is concerned, should show that the filling 
of the gaps mentioned above is an absolutely necessary 
(although not sufficient) condition of the solution of 
the long-range forecasting problem. 
Achievements and Future of the Air-Mass School. 
The achievements of the air-mass school were two in 
number. First, they drew attention to the great horizon- 
tal homogeneity of the easterly and the westerly air 
streams found in tropical latitudes, described the sources 
and modifications of these masses, and accounted for 
them with varying degrees of plausibility. The work 
on the trades still goes on, at all events. There has re- 
cently been a study of the east wind in the Caribbean 
by Haurwitz and his collaborators that is a model for 
all future work of this type [12]. Similar work in the 
far eastern equatorial Pacific, based on the Panama 
Canal Zone, would be very welcome, particularly in 
giving us precise knowledge of the structure of the 
west winds in that area during the summer. Later, the 
researches could profitably be extended to the Far East. 
The second achievement of the air-mass school is that 
the earlier frontal workers discovered in the tropics, and 
