608 
development and intensification of the upper disturb- 
ance. However, it is not possible to conclude anything 
about the causal connection between these phenomena 
without a detailed analysis. 
In a recent paper by Berggren, Bolin, and Rossby 
[4], the conclusion was drawn that extratropical cy- 
clones comprise a heterogeneous collection of different 
types of disturbances, including typical frontal waves 
as well as dynamically quite dissimilar major storms 
associated with the deepening of planetary wave- 
troughs in the upper west-wind belt. Our schematic 
picture is intended to include some of these types. 
The surface polar front and the associated perturba- 
tions are, in our schematic figures, marked only on the 
southeastern sides of the large, upper, cold troughs. 
However, synoptic experience both in North America 
and in Kurope indicates that cyclonic disturbances can 
often develop also along the southwestern sides of a 
cold trough. Especially over western Europe during 
the cold season such disturbances can develop into 
strong cyclones, though they are usually weaker than 
the cyclones formed on the southeastern side of a 
trough. It was not possible to include that type of 
disturbance in our scheme. 
It is not quite clear why there is a preference for the 
southeastern side for the development of large cyclones. 
The preference could be associated with the availability 
of moisture and also with the meridional variation of 
the Coriolis parameter. 
Through the “cutting-off” process associated with 
the formation of “high-level” cyclones on the south 
side of the strongest westerlies, a previous large upper 
trough is gradually elimmated and must ultimately 
disappear. The number of large upper waves then de- 
creases. Because this process of deformation and ulti- 
mate elimination of upper troughs always goes on 
there must also be an opposite process, namely, the 
formation of new major troughs. Sometimes this process 
is a rapid one, an intensifying smaller disturbance de- 
veloping into a major disturbance. In other cases the 
process ismore gradual, and the new large wave is formed 
in a region where there is always a general tendency for 
sucha formation, forexample, on the east coasts of North 
America and Asia. 
CYCLONE FAMILIES AND THE DEVELOPMENT 
OF INDIVIDUAL CYCLONES 
The general scheme presented in Fig. 3 can be con- 
sidered the “normal” pattern for the relationship be- 
tween the surface polar fronts and the long upper waves. 
Some phenomena associated with this normal pattern 
for disturbances in the westerlies will be discussed in the 
following paragraphs. For that purpose a part of the 
scheme containing two surface fronts with their cy- 
clonic perturbations and the cold anticyclone separat- 
ing them is reproduced on a larger scale in Fig. 5. In 
this figure the distance between the two large troughs 
has been somewhat increased and other small changes 
have been made. 
In Fig. 5 the surface fronts, one over the Atlantic and 
the other over the Pacific Ocean, with the separating 
MECHANICS OF PRESSURE SYSTEMS 
anticyclone over the North American continent, to- 
gether represent a synoptic situation very common to 
this part of the Northern Hemisphere. Three frontal 
perturbations representing different stages of develop- 
110 100 30. 60 
Fic. 5.—Schematie surface map showing cyclone families 
commonly observed with fairly symmetrical long waves. Heavy 
lines are 500- and 700-mb contours of polar front, and surface 
fronts; thin lines, isobars. Precipitation shown by stippled 
areas. 
ment are marked on both surface fronts. In Fig. 6 the 
corresponding 500-mb chart with schematic fronts and 
contour lines is presented. In order to emphasize the 
connection between both figures, the positions of the 
surface fronts as well as the contours of the fronts at 
110 
Fic. 6.—Schematie 500-mb map corresponding to Fig. 5. 
Frontal contours as in Fig. 5; thin lines are contours of 500-mb 
surface. Arrows starting at A, B, and C indicate approximate 
48-hour trajectories of air particles in polar air and in tropical 
air, respectively. On trajectory A are indicated pressure levels 
reached by particle in frontal layer at approximate 12-hourly 
intervals. 
700 mb and 500 mb are marked on both the surface 
chart and the 500-mb chart. 
Analyses of a great number of upper-air charts indi- 
cate, as has already been mentioned, that the polar- 
front zone, interrupted in the surface layers, appears 
as a rather well-marked zone of transition between the 
