618 
from the south. However, the influence of the upper 
divergence field on the development of the relative anti- 
cyclonic vorticity should also be considered. 
Both the cold upper lows and the warm upper highs 
can be considered a further development of the upper 
trough and ridge described here earlier. The cold air in 
the upper cyclones represents the uppermost part of a 
sinking cold dome. The warm high-level anticyclones 
consist essentially of a body of lifted tropical air. These 
processes are closely related to the processes connected 
with the formation of cyclones of different types. 
SOME GENERAL REMARKS ON THE 
CYCLONE PROBLEM 
There is no question that our knowledge of the struc- 
ture and life history of extratropical cyclones has greatly 
improved as a result of the increased network of aero- 
logical stations which has been established during the 
last 10-15 years and especially as a result of the general 
introduction of radiosonde observations into the daily 
weather service. The extension of upper-air observa- 
tions on a hemispheric scale has undoubtedly shown 
that there exists, not only in the surface layers but also 
in the middle troposphere, a more or less distinct 
boundary layer separating air masses of polar origin 
from those of tropical or subtropical origin. This bound- 
ary has the character of a true front only over limited 
regions; in other regions it appears as a belt of strong 
meridional solenoid field. Careful analyses of different 
types of synoptic situations also indicate that this 
boundary layer of the free atmosphere can be con- 
sidered a direct continuation of the surface polar fronts. 
The polar front thus appears as a sloping boundary layer 
which extends from the surface to the upper part of the 
troposphere. This is in agreement with the ideas intro- 
duced for the first time in clear form by the Norwegian 
school of meteorologists. 
Although the existence of a more or less distinct 
polar-front surface is thus an empirical fact, the ex- 
planation of the formation of this boundary layer is 
not quite clear. It is obvious that the formation and 
maintenance of a polar front must be connected with 
the general atmospheric circulation. However, there 
still is no complete theory of frontogenesis which con- 
siders the entire dynamics of the general circulation. 
On the other hand, it is undeniable that extratropical 
cyclones frequently and predominantly form at the 
polar front although there are other birthplaces for 
cyclonic depressions. The question of the role of extra- 
tropical cyclones in the complex problem of the general 
circulation is therefore of primary importance not only 
in the cyclone theory but also for a deeper understand- 
ing of the problem of the general atmospheric circula- 
tion. 
In an earlier section it was pointed out that the polar- 
front zone in the middle latitudes is not only the zone 
of strongest solenoid concentration but also the zone of 
strongest concentration of kinetic energy. Thus the 
frontogenetic processes operating in the atmosphere on 
a hemispheric scale form not only the fronts or frontal 
layers but also the characteristic, strong concentration 
MECHANICS OF PRESSURE SYSTEMS 
of the west-wind belt into an upper wind maximum, the 
jet stream. Frontogenesis hence means at the same time 
an increase of both solenoidal and kinetic energy in the 
polar-front zone. 
On the other hand, the vertical circulation character- 
istic of cyclone development also seems to be a “direct” 
one resulting m an increase of kinetic energy. According 
to Starr [60], a system is able to produce horizontal 
kinetic energy if horizontal divergence is associated with 
relatively high pressure and convergence with relatively 
low pressure. In the previous discussion of the life 
history of a cyclone, it was pointed out that the de- 
velopment of a cyclone is characterized by low-level 
convergence and upper-level divergence. But if we con- 
sider not only the cyclone but the total region of an 
atmospheric disturbance—cyclone + anticyclone—the 
horizontal divergence (in the lower parts of the anti- 
cyclone and the upper parts of the cyclone) actually 
occurs with higher pressure than does the convergence 
Gn the lower parts of a surface cyclone and the upper 
parts of a surface anticyclone). This circulation there- 
fore means a transformation of pre-existing potential 
energy into kinetic energy. 
The combined process of frontogenesis and cyclo- 
genesis can probably be described in the followmg man- 
ner. During the time of frontogenesis the “available” 
potential energy of the atmosphere is gradually in- 
creased by nonadiabatic processes, primarily radiation, 
and through a process of concentration of the horizontal 
temperature contrasts into a relatively narrow zone. 
During this process, however, the kinetic energy also 
increases (the surface wind speed remaining relatively 
constant). The stage of frontogenesis is therefore also a 
stage with increase of zonal index, especially at upper 
levels. During the time of cyclogenesis the strong zonal 
flow has a tendency to break down; “available’’ po- 
tential energy is now transformed into kinetic energy. 
The mechanism of the breakdown of the zonal current 
into a more irregular form of movement could be at- 
tributed to some kind of instability of the zonal move- 
ment which is not known in detail.!* The fundamental 
cyclone problem would be solved in principle if one 
could find the causes and mechanism of this instability. 
Any theoretical solution of the cyclone problem is still 
very far from a solution of the cyclone problem pre- 
sented by nature. Meteorologists are still in disagree- 
ment about many fundamental aspects of the cyclone 
problem. While some meteorologists assume that the 
kinetic energy of extratropical disturbances represents 
only a concentration of the pre-existing kinetic energy 
of the west-wind belt, others regard the disturbances 
as active cells in the production of kinetic energy. The 
vertical circulation associated with the growth of the 
disturbances in the westerlies and particularly with the 
occlusion process indicates, as already pointed out, 
that the latter viewpoint is in better agreement with 
13. The ‘‘available”’ potential energy is meant to include 
only the energy available from a redistribution of the mass of 
the system such that the center of gravity is lowered. 
14. Compare, for example, Charney [13], Eady [20], and 
Godske [25]. 
