560 
The variation in speed of the zonal circulation may 
also be treated from the standpoint of the maximum 
index, or the values of zonal wind speed taken at what- 
ever latitude the maximum speed is reached. While the 
nature of the speed variations thus treated is similar 
to those taken between fixed latitudes, this form of 
treatment brings out some important additional in- 
formation indicating that some of the highest speeds 
are reached when the jet is found at low latitudes. It 
also brings to light latitudinal variations in the position 
of the jet stream about the normal which are amazingly 
large. Even if averaged over the hemisphere for periods 
as long as a month, this variation can be as large as 20° 
of latitude during one season. For shorter periods, and 
particularly for selected regions, the latitudmal varia- 
tion of the jet stream can be much greater. Regional 
jet streams have been apparently observed in areas 
ranging from the pole almost to the equator. 
To the extent that we may speak of the maximum 
high-level zonal current averaged over the hemisphere 
as a jet stream, we may summarize the foregoing re- 
marks by saying that there are large day-to-day, week- 
to-week, and month-to-month variations in the strength 
and latitude of the jet stream. The latitudinal variations 
may be looked upon fundamentally as an expansion 
and contraction of the cireumpolar vortex toward or 
away from the pole. 
The regional variations in the latitude and perhaps 
in the strength of the jet stream are associated with the 
wave patterns in the circumpolar vortex which were, 
to some extent, treated in a foregomg section. Ideally 
these long waves are generally looked upon as a simple 
wave pattern in the upper troposphere and lower strato- 
sphere with from four to six meridionally extensive 
ridges and troughs around the hemisphere. The ob- 
served wave patterns are rarely if ever so ideal, and a 
more accurate description would be that there are 
generally two (sometimes more) different families of 
waves in different latitude bands. These two wave 
trains may, and usually do, possess different wave 
lengths (in degrees of longitude) and hence around the 
hemisphere there may be a different number of waves 
in higher latitude bands than in lower latitude bands. 
for this reason, too, the waves in the two bands are 
frequently out of phase. These remarks pertain equally 
well to daily, weekly, or monthly mean maps. 
When the jet stream of the circumpolar vortex is well 
to the north of its normal position, the ridges and 
troughs are generally strongly tilted in a northeast- 
southwest direction or even fractured as indicated in 
Fig. 11, Stage 1. When the waves get into phase (.e., 
when the troughs and ridges are more or less merid- 
ionally extensive) the latitude of the jet stream usually 
lowers but does not reach its lowest value. This posi- 
tion is reached only after the entire character of the 
hemispheric upper-tropospheric circulation is radically 
altered to a cellular rather than a wavelike structure. 
The explanation of this fundamental change in charac- 
ter, in spite of its enormous importance to meteorology, 
is not yet known. However, in following the develop- 
ment of such transitions it often appears that the 
THE GENERAL CIRCULATION 
changes go on somewhat as follows: In the initial stage 
(illustrated schematically in Fig. 11) there are extensive 
zones of confluence where deep cold polar and warm 
B 
Be 
ENO gan 
Fic. 11.—Schematic representation of successive circula- 
tion patterns of the index cycle. The time interval from 
Stage 1 to Stage 4 1s about two weeks. 
tropical air masses are forced to flow side by side, asso- 
ciated with filaments of regional jet streams to the east 
of the zones of confluence. In Stage 2 the waves, per- 
haps through some form of relative motion associated 
with a nonstationary wave length, have combined into 
deep full latitude troughs and ridges. This combination 
permits vast meridional transports of deep polar and 
tropical air. Stage 3 results from a refracture of the 
troughs at high and low latitudes, again perhaps in an 
attempt to readjust wave lengths to equilibrium values. 
The cold polar air masses brought equatorward and the 
warm currents which were delivered to polar regions by 
the in-phase troughs and ridges now have no return 
passage to their source regions—that is, the evolution 
of the flow patterns is now in the direction of trapping 
both the equatorial air m the north and polar air m the 
south. Finally, in Stage 4 the cold pools (Kaltlufttropfen) 
in the subtropics and the warm pools in the subpolar 
regions take on characteristic cyclonic and anticyclonic 
cellular circulations. In these schematic models it will 
be noted that certam asymmetry and tilt of the troughs 
and ridges are introduced where necessary to bring 
about the advection of momentum required to maintain 
the westerlies in the manner postulated. Jeffreys [27] 
and lately Starr [48] have emphasized the importance 
of such asymmetry as related to the transfer of mo- 
mentum. 
This idealized sequence of events generally does not 
take place simultaneously in all parts of the hemisphere. 
However, when such a development on a large scale 
takes place in one region, it is often followed by similar 
reactions in other regions. This transition of the cir- 
culation wherein deep cold cyclones are ultimately 
formed at low latitudes and deep warm anticyclones 
are formed at high latitudes is referred to as blocking. 
The most commonly observed behavior of blocking 
action is a progressively westward development of re- 
