TROPICAL METEOROLOGY 
posed to call the new entity they had introduced into 
tropical meteorology, the equatorial front. 
The work of Brooks and Braby seems to have been 
overlooked by the Norwegians, for they proceeded more 
or less along the same path without referring to it. 
However, their researches, in which they used the same 
method of analyzing surface mean winds and rainfall, 
were more extensive, ultimately covering all the tropics. 
They describe the “‘intertropical front” in the Pacific 
[10] and Atlantic and Indian Oceans, and in addition 
attempt to trace its course over the land in Africa and 
southeastern Asia. Seasonal movements of the front are 
attributed to different mean temperature conditions in 
the trades of the winter and of the summer hemispheres. 
It is clear that, up to 1933, the new tropical meteorol- 
ogy still followed, at least implicitly, the assumptions of 
the climatological school. It was implied that the equa- 
torial or intertropical front, bemg found on maps of 
mean winds, pressure, and rainfall, would therefore be 
found also on synoptic maps. As far as I can discover, 
this assumption was not questioned at any time; indeed, 
many authors did not recognize that it was an assump- 
tion. After 1933, therefore, there was a concerted 
attempt to apply frontal and air-mass analysis to syn- 
optic maps in the tropics. The movement started slowly, 
being most active in the Pacific and the Caribbean. By 
1939 the standard high-latitude methods of frontal 
analysis had been adopted by the Philippine Weather 
Service, by the meteorologists of Pan American Airways 
[17], im New Zealand [3], Australia [50], and in Mar- 
tinique and other parts of the Caribbean [81]. World 
War II brought an immense increase in the amount of 
forecasting required in the tropics and, with it, a large 
influx of meteorologists whose training and experience 
had been largely in high latitudes. The latter applied 
standard air-mass analysis to their new problems, al- 
most without question; the literature at that time gave 
them every reason for doing this. The only paper that 
might have thrown some doubt on the accepted view 
[25] had at that time attracted little attention. Today, 
as the survey mentioned in the introduction shows, 
almost half the meteorologists who have had experience 
in the tropics still adhere to the tenets of the air-mass 
and frontal school. We are, therefore, entering the most 
controversial part of this review and it is well to pause 
and examine our fundamental conceptions closely. For 
it is often found that an irreconcilable conflict of scien- 
tific views concerns not the facts but the explanations; 
most of the difficulties may be semantic in origin. 
It must be realized that the complex of ideas 
advanced by Norwegian meteorologists may be divided 
into two sections. The first section is concerned with 
certain readily verified facts. The term front, in empir- 
ics, 1s applied to a set of atmospheric phenomena that 
are frequently discovered by our instruments and by 
visual observation to be closely associated. The most 
striking feature is the occurrence of an organized system 
of clouds, extending over immense horizontal distances; 
one type of cloud system is called a warm front, another 
a cold front, and combinations of the two are termed 
occlusions. However, these terms are only applied if, 
865 
accompanying the cloud system, there is a narrow zone 
in which the meteorological elements humidity, tem- 
perature, and wind show rapid changes whose sense is 
defined for each of the three different types of cloud 
systems. Further, in high latitudes, the pressure 
gradient should have an oppositely directed component 
on either side of the system. In this sense an observer 
on the ground or a pilot in the air may identify a front 
without ever seeing a synoptic map. It is an observable 
entity. The Norwegians, however, did not stop at the 
descriptive stage of meteorology. They had also an ex- 
planation, a theory about fronts. This theory accounts 
for the observed fronts as being boundaries, approxi- 
mating atmospheric discontinuities, between air 
masses which possess characteristic and more or less 
homogeneous values of the meteorological elements. 
Further, the theory requires that the fronts can be 
maintained for any length of time only if there is a 
density contrast accompanied by cyclonic shear at the 
discontinuity. Now this theory is not a priori true. 
There is no general theory of the atmospheric circulation 
by which we can derive from the principles of physics 
the result that fronts must form in this manner. The 
confidence we have in the theory as an explanation 
applicable to high latitudes comes not from dynamic 
meteorology but from experience with synoptic maps, 
by means of which we check the contention that the 
fronts observed at a single station or by the airman are 
in fact boundaries between air masses. This experience 
leads us to give the name front to regions where certain 
meteorological elements show steep gradients even in 
the absence of the typical cloud and weather pattern. 
This is done because such systems preserve the two 
most essential features of the theoretical front, viz., 
the density contrast and the cyclonic wind-shear. 
The density contrast and the wind shear are essential 
to the further development of the theory. For, accord- 
ing to the Bergen school, the frontal surface is 
not always in stable equilibrium. Waves develop in it 
and some of these waves are dynamically unstable. 
Such unstable waves give rise to vortical circulations 
which roll up the surface and form the cyclones 
of middle and high latitudes. The kinetic energy of these 
cyclones is derived from the potential energy of the 
mass distribution at the undisturbed fronts. It is now 
obvious why both the density contrast and the cyclonic 
shear are needed at the theoretical front. The first pro- 
vides the necessary source of energy for the storm, and 
the second is the destabilizing influence that leads to 
the occlusion process. They, of course, play other roles 
in the theory, but this is their part in the wave theory . 
of cyclones. To emphasize the independence of the 
empirical and theoretical fronts, I should like to remind 
the reader that line squalls were known long before the 
frontal theory was promulgated and that very accurate 
descriptions of cold fronts exist, dating back to the last 
century [42]. If the theory of extratropical cyclone 
formation should ever be upset and discredited, cold 
fronts, warm fronts, and occlusions would continue to 
be observed and to be identified on synoptic maps, 
