TROPICAL METEOROLOGY 
during one monsoon as the other, notwithstanding that the 
North-East monsoon starts out as a cold, dry continental 
air mass and the South-West monsoon as a warm, maritime 
air mass [46, p. 27]. 
Connected with the uniformity of surface tempera- 
ture over the oceans is the well-known uniformity in 
the height of the base of tropical cumulus. This rarely 
is less than 1500 ft or greater than 2000 ft, and de- 
partures from these values are so clearly connected with 
the occurrence of precipitation that the problem of 
forecasting cloud bases reduces to that of forecasting 
the presence or absence of precipitation. 
The persistent and uniform values of the meteoro- 
logical elements are correlated. Both the trades and the 
monsoons are more than winds blowing with great 
regularity. Other elements, such as the precipitation and 
the cloud amount, the temperature, pressure, and hu- 
midity regimes, and the vertical variation of the ele- 
ments all have structures or values characteristic of 
the trades or of the monsoons. Bergeron [5] has empha- 
sized these correlated characteristics and has held up 
the trade and monsoon of the tropical climatologist as 
paradigms of concepts that should be used in the air- 
mass climatology of higher latitudes. As an example he 
pointed out that “good monsoon” in India means a 
complex of heavy rains, strong steady southwest wind, 
and a characteristic vertical temperature and humidity 
distribution. The term monsoon, therefore, is to be 
applied to a persistent kinematic and thermodynamic 
complex. Similarly numerous observations, culminating 
in the Meteor results [18], have shown that the trade is 
a persistent complex. Here again we have not only a 
characteristic and persistent wind, but also a typical 
associated cloud form, precipitation regime, and lapse 
rate of temperature and humidity. The “‘trade-wind 
inversion”’ is indeed well known and the equally well- 
known trade cumulus owes its peculiarities to the asso- 
ciation of the inversion with a typical vertical wind- 
shear. 
So far we have mentioned the representativeness of 
the statistical parameters that are derived from obser- 
vations over the great tropical oceans. At first sight, 
conditions are different over the land, particularly over 
the continents. The highly developed land and sea 
breezes of tropical islands, and the thunderstorms that 
break out with great regularity in the archipelagos of 
the East and West Indies are departures from mean 
conditions that are, through travelogue and fiction, 
known even to the layman of high latitudes. But these 
departures are clearly periodic, and synoptic observa- 
tions may be corrected for them, either quantitatively 
or qualitatively. The seasonal means for the tropical 
islands are then representative of these corrected synop- 
tic values. Moreover, the diurnal departures from the 
mean can be so clearly related to astronomical and 
orographical causes that the short-period forecast seems 
to depend only upon an adequate knowledge of the 
statistics and of geography. Upon close examination the 
same principle seems to apply to forecasting in con- 
tinental regions. Here we have the added advantage 
that the seasonal variations themselves appear to be 
861 
due to the same causes as the diurnal changes, though 
operating on a larger scale and over longer periods of 
time. On this view, we may take the slowly varying 
oceanic conditions as the ground state and regard both 
diurnal and seasonal variations over the land as per- 
turbations whose causes, orographic and astronomical, 
are known. The perturbations are thus thermal in their 
immediate origin and there appears to be no problem 
of explanation so far as synoptic variations are con- 
cerned. Similarly there ought to be no forecasting prob- 
lem for periods shorter than a season that cannot be 
solved with an adequate knowledge of statistics and of 
orography. 
In the foregoing account I have outlined an extreme 
climatological view of tropical meteorology. On this 
view, there are no synoptic problems in the tropics. 
Synoptic problems become climatological problems; the 
future advance of tropical meteorology therefore 
depends on the collection and reduction of longer and 
longer series of observations from more stations in order 
to obtain maps of stable statistical parameters and their 
seasonal variations which are to be combined with a 
more detailed knowledge of orographical peculiarities. 
The remaining problems are then dynamical and relate 
almost entirely to the “ground state” that was men- 
tioned above. If corrections for diurnal and seasonal 
variations may be made by assigning them to known 
astronomical and orographie causes, it is clear that we 
are left with the statistical picture of the ‘‘trades” 
considered as kinematic-thermodynamic complexes to 
be explained by the methods of mathematical physics. 
Since the complexes are regarded as the most persistent 
features of the atmosphere, they represent a steady state 
in the dynamic sense and the statistics give us a direct 
insight into what is known as the general circulation, 
at least for latitudes below 20°. What has to be ex- 
plained, therefore, is the tropical part of the “planetary 
circulation” or ‘‘winds of the globe” and this explana- 
tion turns out to be one of the easiest ever attempted in 
meteorology. If it should appear that I have represented 
a very extreme view and one which would not be 
defended in this form by any modern meteorologist, I 
suggest that almost all textbooks that refer to the 
tropics at all (e.g. [49]), and certainly some recent 
authors of research papers dealing with the general 
circulation [45, 56], imply this view in their treatment 
of the dynamics of the tropical atmosphere. This is 
true not only of those we may rightly classify as be- 
longing to the climatological school but also of those in 
the frontal and perturbation schools, insofar as they 
treat the general circulation. Those authors will be 
dealt with in their proper place. 
What kind of picture, then, is obtained by pursuing 
these principles to their final conclusion? First, let us 
look at the now-modified empirical laws governing the 
general circulation in the tropics. Abstracting the data 
in such a way as to omit the diurnal and seasonal 
perturbations of thermal and mechanical origin, we are 
left with a system of trade winds, blowing steadily 
from the east and toward the equator in the lower 
layers of the atmosphere. The trades of the two hemi- 
