862 
spheres approach one another at the equator, as is clear 
from the statistics. There, they become light and varia- 
ble and this weakening is obviously part of the observed 
horizontal velocity convergence in the two currents, 
with a correlated increased vertical component in the 
motion. The lapse rates of temperature and humidity 
confirm this conception. At some distance from the 
equator, the lower atmosphere is characterized by the 
presence of an inversion of temperature, the trade- 
wind inversion, above which the humidity drops to 
very low values, both absolutely and relatively. But 
this inversion, under the influence of the predominantly 
vertical motions, becomes higher and weaker as the 
equator is approached, finally disappearing in the zone 
of intense upward convection at or near that boundary. 
At high levels, the motion of the air is predominantly 
and necessarily (on continuity considerations) away 
from the equator, and at relatively small latitudes it 
takes on a westerly component to form the antitrades. 
The antitrades are thus westerly winds with a poleward 
component and a downward vertical component that 
reaches its maximum values in the neighborhood of 
30°. A circulation cell is thus completed and the com- 
position of the air im it agrees well with the distribution 
of the vertical component of the steady-state motion. 
In particular the region of light variable winds near 
the equator, which we shall now call the doldrums, is, 
as the statistics show, also a region of large cumulo- 
nimbus clouds, with attendant heavy precipitation, the 
“squalls” of the early navigators. On the other hand 
the trade-wind zone is the region of trade cumulus, 
with only light showers or none. The horse latitudes, 
finally, are regions in which fair-weather cumulus or 
stratocumulus predominate; sometimes they are clear 
of cloud. 
When we come to explain this abstracted picture of 
the tropical circulation we find the necessary concepts 
ready to our hands. The equatorward component of 
the trades was first explained by Halley in 1686; this 
constitutes, in fact, the first attempt to explain any 
feature of the general circulation. According to Halley, 
the equatorward component is to be attributed to the 
replacement of air that has risen by upward convection 
at the equator. Surface heating at the equator, sup- 
posedly the region receiving the greatest amount of 
heat from external sources, is sufficient on this view to 
drive the whole meridional circulation of trades and 
antitrades. Halley’s explanation of the zonal compo- 
nents of the motion, however, was not accepted. In 1735 
John Hadley gave the accepted theory, invoking the 
conservation of absolute angular momentum. Since that 
time the combined explanation has been handed on 
from text to text. It was, for example, embodied in the 
earlier dynamic theories of the Chicago group. Rossby 
[56] specifically has used it under the guise of the ‘‘direct 
cell” in the general circulation. Although little quanti- 
tative work has been done, the qualitative use of the 
explanation, it is assumed, solves the dynamic problem 
for the tropical atmosphere. On this assumption, any 
further work would be directed to improving the theory 
TROPICAL METEOROLOGY 
quantitatively, by referring to better and better clima- 
tological data. 
Summing up the extreme climatological view, we have 
the following conceptual apparatus: the tropical atmos- 
phere is composed of persistent wind systems, with their 
associated composition and thermodynamic properties, 
the trades and the monsoons. Diurnal variations of the 
meteorological elements are due to perturbations of 
these systems, attributable to thermal and-orographic 
disturbances of known origin. The monsoons are sea- 
sonal perturbations of the trades due to the same causes. 
The trades themselves are thermally driven and all 
features of that circulation are explicable in terms of 
what happens in the doldrums, a region of intense con- 
vection and light variable winds coinciding with the 
equator in the ideal model. Astronomical variations, 
combined with the different absorptive properties of 
land and sea, therefore account for everything—trades, 
doldrums, monsoons, land and sea breezes, the regime 
of precipitation, temperature and humidity, cloudy and 
fine weather, and lastly (by dynamic functional re- 
lations embodied in the equations of motion) the pres- 
sure distribution. There are no problems left to solve. 
On this basis, a large part of the more successful short- 
range tropical forecasting in World War II was im- 
plicitly conducted; the little long-range forecasting that 
was done was based explicitly on this statistical method; 
on this basis, also, most present-day models of the 
general circulation incorporate the tropical atmosphere. 
Two unfortunate circumstances mar the beauty of 
the picture. The first is a fact that has been known to 
western science for four hundred years or more. Oceanic 
tropical regions are not only the seat of steady 
trades and monsoons, they are also, in their season, 
notorious for the presence of the most violent storms 
dealt with by synoptic meteorologists—typhoons, hur- 
ricanes, and tropical cyclones. How are these storms to 
be accounted for in terms of the causal theories of the 
climatological school? The first impulse is to explain 
their origin in terms of the standard astronomical 
and orographical variables previously invoked. During 
the last century, the tendency was to regard tropical 
cyclones as being perturbations due to local intensifica- 
tions of convection in the doldrums [43]. Those who 
wished to advance more explicit causes attributed the 
convection to local maxima of heating [27]. But this 
explanation was soon seen to be untenable, since the 
region of origin of at least some typhoons was discov- 
ered to be over the parts of the Pacific Ocean where 
the temperature gradients in the lower layers were very 
weak. It must be confessed that the climatological 
school never successfully dealt with the problem of the 
origin of tropical storms. The best that could be done 
was to chart the regions and seasons in which they oc- 
curred, to show that the mean tracks generally follow- 
ed the march of the sun and that there was some 
justification in the statistics for regarding tropical 
storms as connected in some way with the only re- 
gion, in the view of the school, that showed the nec- 
essary high variability of the meteorological elements: 
