888 
reached. The storm is most symmetrical during this 
period and only a relatively small area is as yet in- 
volved. 
3. The stage of maturity, when there is no further 
deepening, the isobars are gradually spreading out, and 
the area covered by gales and hurricane winds is larger 
than at any other period, but the intensity is gradually 
decreasing. 
4. The stage of decay, when the storm is dissipating 
over land or is recurving northward and assuming extra- 
tropical characteristics. 
Air-Mass Arrangements around Tropical Cyclones. 
Tropical cyclones develop in homogeneous air and dur- 
ing the early and middle stages of their existence en- 
counter only 7m, tropical maritime air (the trades), 
and Hm, equatorial maritime air (the equatorial wester- 
lies and the intertropical convergence zone). During 
or immediately following recurvature Vp, transitional 
polar air, may be encountered. There is little difference 
between all three air masses below 10,000 ft. Above 
that level Hm is the most moist and Vp usually the 
driest. The air masses around tropical cyclones are essen- 
tially homogeneous, at least until the close of the stage 
of maturity. There are no fronts until the stage of decay, 
although distortions in the wind field analogous to fronts 
may be found in cyclones having a strong northerly 
component! in the direction of movement. - 
Barometry of Tropical Cyclones. A tropical disturb- 
ance may persist for a number of days without deepen- 
ing and, particularly in the intertropical convergence 
zone, may never develop into a depression. When the 
depression has intensified to Beaufort 6-7, deepening 
takes place rather rapidly and maximum deepening 
is usually reached within 72-84 hr in the Atlantic 
and from four to five days in the Pacific. Little change 
then occurs if the storm remains over water until it 
recurves and begins to assume extratropical characteris- 
tics and the central pressure begins to rise. There is 
apparently some limiting intensity, or state of equi- 
librium, dependent upon the moisture content and pos- 
sibly the lapse rate of the lower atmosphere. 
Deppermann [2] has investigated the pressure gra- 
dients of typhoons with minima less than 28.75 inches 
(973.6 mb) with results as follows: 
Number 
Minimum pressure of storms Mean fall Mean fall 
in. hy in. mile> 
Below 27.56 in. (933.8 mb)........ 4 0.79 0.06 
27.57-27.95 in. (946.5 mb)......... 4 0.43 0.03 
27.96-28.35 in. (960.0 mb)......... @ 0.45 0.03 
28.36-28.75 in. (973.2 mb)......... 8 0.40 0.025 
He concludes that, as a rule, steepness of gradient im- 
creases with depth of the barometric minimum. How- 
ever, as Deppermann himself points out, the mean 
pressure fall per hour is determined to a considerable 
extent by the rate of movement of the storm and does 
not truly represent the pressure gradient of the storm 
but rather the steepness of the barograph trace. The 
last column does represent the actual pressure gradient, 
1. This expression should be interpreted to mean ‘‘a com- 
ponent toward the north.” 
TROPICAL METEOROLOGY 
which may or may not approximate the normal pressure 
gradient in severe typhoons since it is based on a rela- 
tively small number of storms. Cline [1] found an aver- 
age pressure gradient of 0.02 inches per mile, indicating 
the further spread of the isobars in the later hurricane 
stages which are normally found as these storms reach 
the coast line of the Gulf of Mexico. 
There are many authentic imstances of gradients much 
steeper than those in the table above, usually in storms 
in the immature stage. Several examples of steep gra- 
dients are given by Deppermann: The Pathfinder, an- 
chored at Aras, Samar, P. J., recorded a pressure fall 
of 1.07 inches in twenty-seven minutes. In a Tacloban, 
P. I., typhoon there was a drop of 1.14 inches in thirty 
minutes, with a measured fall by mercurial barometer 
of 0.49 inches in five minutes. The S.S. Virginia in the 
Central Caribbean, on September 20, 1943, experienced 
a barometer reading of 28.74 inches at 8:00 P.M. and 
27.40 inches twenty minutes later, a fall of 1.34 inches, 
and rose to 28.60 inches by 9:00 p.m. The pressure 
fell more than 2 inches in ninety minutes. The diameter 
of the circle enclosed by the 29.5-inch isobar was esti- 
mated at fifty miles; the calm center, from eight to 
twelve miles. If a flat minimum in the calm center is 
assumed, the pressure gradient was 2.10 inches (71.1 
mb) in nineteen miles or 0.11 mches per mile. This was 
a young, immature hurricane which had undergone its 
principal development during the previous two days. 
In the Labor Day hurricane over the Florida Keys in 
1935 there was evidence of a pressure gradient in excess 
of one inch in six miles. 
There has been general agreement that in the im- 
mature and mature stages of the tropical cyclone iso- 
bars are nearly circular from the center outward to 
around the 29.4-inch (995.6-mb) isobar. However, there 
has never been a sufficiently dense network of accurate 
barometer readings to verify this completely. Indeed, 
evidence in recent years tends to indicate that isobars 
are not absolutely circular except possibly quite close 
to the center. Soon after the beginning of recurvature, 
more apparent distortion of most of the isobars is 
usually evident. Isobars are more truly circular im the 
rapidly deepening stage than at any other time. 
The accepted lowest barometer reading of record in 
the world is 26.185 inches in a typhoon encountered by 
the Dutch steamship Sapoeroea about 460 miles east of 
Luzon, P.I., on August 18, 1927. The official record for 
the barometric minimum in the Atlantic is 26.35 inches 
on September 2, 1935, on Lower Matecumbe Key, 
Florida. Readings between 27.00 and 27.50 inches 
(914.3-931.38 mb) are fairly common. 
Pumping—the rapid and possibly rhythmic oscilla- 
tion of atmospheric pressure—has frequently been ob- 
served in tropical cyclones. Gherzi [7] claims to find a 
mean period of about six seconds corresponding with 
the period of typhoon swells, which would indicate that 
the oscillations in this case are produced by micro- 
seismic waves set up by the typhoon swells. Depper- 
mann believes that oscillations of such periodicity do 
not occur in the Philippines, nor have they been noted 
in the Atlantic, although they may be masked by other 
