ANTICYCLONES 
higher, colder upper troposphere and tropopause. The 
presence of this ‘‘ecold cap” above surface anticyclones 
was pointed out by Muigge [18], who showed that 
advection from the south was not the sole cause. An 
excellent analysis of a case of anticyclogenesis, showing 
the vertical structure, vertical motions, and the cold 
cap, has been given by Fleagle [11]. Fleagle stresses the 
fact that as one follows the motion of the anticyclone, 
the upper cold cap is not composed of the same air 
particles, which is a further indication of the existence 
of dynamic processes in anticyclogenesis. 
MOTION AND TRANSFORMATION 
OF ANTICYCLONES 
The polar anticyclone in its source region is character- 
ized by a large surface inversion and a nearly isothermal 
layer above, extending usually to no more than 3 km 
above the surface which marks the top of the true 
polar continental air [35]. When the anticyclone moves 
away from its source region and travels over a warm 
surface, both the heating from below and the onset of 
mechanically induced turbulence wipe out the surface 
inversion, and may, in fact, create a steep and even 
superadiabatie lapse rate in the surface layer. Above 
this layer of steep lapse rate is usually found an inver- 
sion which increases in thickness and magnitude as the 
air in the isothermal layer subsides. The top of the 
inversion marks the top of the true polar continental 
air, while the air above, despite its subsidence, main- 
tains a greater lapse rate. The spectacular nature of the 
“subsidence inversion,” separating as it does the cool, 
moist, cloudy air below from the warm, clear, dry air 
aloft, has been studied intensively by Namias [19] 
and Hewson [17]. The more pronounced cases of sub- 
sidence are believed to accompany the transformation 
of the polar anticyclone into the warm or dynamic 
anticyclone, and may indeed be considered as one of 
the by-products of such a transformation. 
Forecasting the motion of polar anticyclones away 
from their source regions is a problem of considerable 
interest to meteorologists. Often one polar anticyclone 
follows another in a regular procession from the source 
region and at other times intense large anticyclones 
remain stationary for many days. From observations it 
has been noted that a strong belt of zonal westerlies 
equatorward of the source region tends to act as a 
“barrier” preventing appreciable motion southward of 
polar anticyclonic offshoots. But if waves develop in 
the westerlies, the troughs act as “‘spillways’”’ for the 
release of polar anticyclones. 
Another approach to the problem was recently pre- 
sented by Rossby [27] who studied the conditions neces- 
sary for the motion, away from the geographical pole, 
of barotropic vortices embedded in a motionless baro- 
tropic atmosphere of great depth. Taking into account 
the variation of the Coriolis force with latitude, he 
found that if vortices are located, or are displaced, 
away from the geographic pole, the cyclonic vortices 
will move poleward while the anticyclonic vortices will 
move equatorward with increasing speed (friction neg- 
lected). The motion of the polar anticyclonic vortex will 
625 
be accompanied by a flattening of the dome and a con- 
version of the loss in potential and in rotational kinetic 
energy into kinetic energy of translation of the dome. 
Some of this translational energy is imparted to the 
environment which thus acquires motion. Rossby 
showed that, as the anticyclonic vortex moves south- 
ward (in the Northern Hemisphere) and sets the en- 
vironment into motion, the anticyclonic center (center 
of the streamlines) is displaced westward and increas- 
ingly so with increased speed southward of the vortex. 
This process represents a dynamic movement of the 
center of the original polar anticyclone westward from 
its original position under the highest poimt of the 
polar dome. This conclusion is in excellent agreement 
with the fact long noted by synoptic meteorologists 
that the top of a southward moving polar dome is 
found about midway between the surface cyclone cen- 
ter and the anticyc one center to the west. It is not 
known whether this type of “dynamic anticyclone”’ 
leads to higher pressure than that originally found in 
the center of the polar anticyclone. At the time the 
center of the anticyclone is shifted westward, a crescent 
“moon-shaped” cyclonic center appears on the eastern 
edge of the polar dome. Considering the circular bound- 
ary of the original polar dome as the “front,” Rossby 
showed that the southward displacement of the dome 
is accompanied by frontolysis and a spreading out of 
the cold air along the western edge, and frontogenesis 
and steepening of the front along the eastern edge of 
the cold dome. Both of these latter conclusions are in 
excellent agreement with observations. 
In applying Rossby’s results to the problem of release 
of polar anticyclones initially located away from the 
geographical pole, it must be remembered that the 
thermally produced anticyclone should theoretically be 
accompanied by a polar cyclone aloft at the source 
region [36], thus indicating the baroclinicity of the polar 
atmosphere. But if one assumes that Rossby’s results, 
derived for a barotropic atmosphere, hold for this 
case, there would be a tendency for the surface anti- 
cyclone to move equatorward and the upper cyclone to 
move poleward. This “‘splitting-off” of the surface anti- 
cyclone from its upper cyclone is again quite reasonable 
in the light of aerological experience. However, as 
Rossby points out, if the concomitant sinking of the 
polar dome occurs in an actual atmosphere of normal 
stability, the resulting convergence aloft should create 
a cyclonic vortex at upper levels. This dynamically 
created cyclonic vortex may accompany the cold air 
dome equatorward, but sooner or later it will have to 
turn poleward in accordance with the increased pole- 
ward force acting on it. Thus, it is likely that, both 
at the source region and later on in the trajectory of 
the polar anticyclone, forces operate in such a way as 
to remove cyclones aloft, and pave the way for upper- 
level anticyclogenesis. 
The likelihood of transformation of a polar anti- 
cyclone into a dynamic anticyclone appears to be a 
function primarily of its position relative to the wave 
pattern of the westerlies. If the polar anticyclone is 
associated closely with the motion of a pronounced 
