TROPICAL METEOROLOGY 
only empirical conditions so far discovered which are 
necessary and sufficient are the existence of horizontal 
velocity convergence in the lowest layers of the atmos- 
phere plus the existence of an asymptotic lme in the 
streamline field. This statement, of course, in no sense 
gives a causal explanation of the lines; it merely gives 
descriptive correlations. 
With these conclusions, we must dismiss the whole 
body of theoretical explanation developed by the air- 
mass school in the tropics, but at the same time we 
must incorporate their empirical findings with those of 
N WIND ANALYSIS 1500 FT. 
I100B JULY 9,1946 
10° 
5° 
STATION MODEL 
TIME 
dd°vv 
160° EAST 165° 
875 
trough in the western Pacific. He comes to the conclu- 
sion, a correct one in my opinion, that the western part 
of the equatorial trough is the seat of a large number of 
eddies which for the most part move from east to west, 
either recurving ultimately as hurricanes and tropical 
storms or passing on to the continent of Asia. But his 
method of analysis is incapable of showing the relation 
of the “fronts” that develop pari passu with the vortices 
to the early stages of their development. From Figs. 16 
and 17 in his paper, for example, and from the text, it 
is clear that Riehl envisages the lines of convergence as 
=[@ o° 10° 20° 
400 
KWAJALEIN O200B 
(PRESUMPTIVE NORTHERN 
HEMISPHERE AIRMASS) 
---— TARAWA 02008 
(PRESUMPTIVE SOUTHERN 
HEMISPHERE AIRMASS) 
JULY 9, 1946: 
600 
700 
MILLIBARS 
800 
900 
Fic. 2.—Streamline and isovel analysis at 1500 ft of an equatorial vortex passing the Marshall Islands, 1100 Bikini Time July 
9, 1946, with soundings from Tarawa and Kwajalein. Speed in knots. (Reprinted from the 7th quarterly report of the Tropical 
Pacific Project, U.C.L.A., a research supported by funds provided by the Geophysical Research Directorate, Air Materiel Com- 
mand.) 
the perturbation school; this we shall attempt in the 
next section. 
Achievements and Future of the Perturbation School. 
The outstanding achievement of this school is undoubt- 
edly the discovery and description of the easterly wave 
and its relation to Caribbean hurricanes. Its fault lies 
in an almost complete neglect of the empirical findings 
of the air-mass school. This is to be attributed to the 
fact that members of the school have consistently used 
a streamline technique that is inadequate for the study 
of detailed features of the wind field. As a result, the 
school has placed great emphasis on the “trough” in 
the wind field and has marked it with a heavy line on 
the map, but at the same time has completely missed 
the asymptotes of the field which, indeed, are accurately 
located only by using the standard methods described 
by V. Bjerknes. This weakness in the synoptic methods 
of the school has recently been emphasized by Riehl’s 
investigation [54] of the structure of the equatorial 
existing before the formation of the vortices. Further, 
the equatorial westerlies are conceived as being part of 
the general circulation, so that convergence lines be- 
tween them and the trades, though not continuous 
throughout the equatorial low-pressure trough, may 
nevertheless exist independently of the vortex series. 
The result is a very complex picture of the normal state 
of equatorial circulation in the wet season- 
When the standard methods of streamline analysis 
are applied to maps from this area, however, a much 
simpler picture is seen. Upstream, in the central Pacific, 
the trade winds of the two hemispheres lie side by side 
for the most part, forming an extensive easterly current 
that overlaps the equatorial low-pressure trough in 
both the Southern and the Northern Hemispheres. 
This easterly current is subject to perturbations of the 
same type as the easterly waves of the Caribbean save 
that they are at much lower latitudes and extend across 
the low-pressure trough. No lines of convergence are 
