546 
surfaces. Such estimates as can be made indicate that 
this gross average action may be of very great effec- 
tiveness. 
In view of the fact that the correlation of density 
with meridional motion is most marked near the sur- 
face, it follows that the mean meridional distribution 
of this divergence and convergence shows greatest con- 
trasts at low levels. This at once suggests that the mean 
meridional distribution of net external heating and cool- 
ing of the atmosphere is linked with the process, since 
the heating of the atmosphere is most intense near the 
surface and much of the cooling is accomplished through 
surface effects also. Thus it would be logical to suppose 
that the net external heating in the more equatorial 
latitudes causes a net horizontal expansion, while the 
net cooling in the polar caps brings about a net hori- 
zontal contraction, accounts then being balanced, as 
far as mass continuity is concerned, by differential 
advection across middle latitudes. 
If one agrees to argue on this basis, a far-reaching 
observation may be made, subject of course to the 
correctness of the premises involved. Earlier in the 
discourse it was pointed out that one of the most 
fundamental questions in meteorology relates to the 
presence of broad belts of westerlies in the middle 
latitudes of each hemisphere. Speaking now only of the 
lower levels we see that the presence of the westerlies 
coincides with the requirement that the regions of net 
external heating and divergence should coincide with 
regions of higher pressure along horizontal surfaces. 
Unless this situation exists a net production of kinetic 
energy to overcome friction would be impossible. Thus, 
if for a moment we were to visualize a hypothetical 
situation with mean easterlies in middle latitudes, the 
net heating and consequent divergence would take 
place at a relatively low pressure in the more tropical 
regions, while the convergence and cooling in the polar 
caps would take place at a relatively high pressure. 
Such a hypothetical situation would then be rapidly 
destroyed, since more kinetic energy would necessarily 
disappear in the polar caps through convergence than 
could be generated in the more tropical regions by 
heating and divergence. 
Although the net heating and cooling of the earth is 
ultimately caused by radiational exchanges with space, 
when one isolates the atmosphere and considers the 
circumstances in the winter season especially, due re- 
gard has to be given to the strong heating of the atmos- 
phere by the oceans when cold air masses flow out over 
relatively warm water surfaces. As has been pointed 
out by Sverdrup and others [13], the oceans serve in a 
manner of speaking as a hot-water heating system. It 
should thus be expected that during winter the regions 
of net heating for the atmosphere extend much farther 
into middle latitudes than might otherwise be supposed. 
6. Since variations of pressure along horizontal surfaces are 
necessary for the generation of a net amount of kinetic energy 
to overcome friction, it is indeed reasonable on general grounds 
to suppose that the atmosphere behaves in such a way as to 
take advantage of the large systematic pressure differences 
between the polar and tropical regions. 
THE GENERAL CIRCULATION 
Since we have been speaking of the meridional trans- 
fer of internal heat energy, it is desirable to mention 
also the effects of the meridional transfer of latent heat. 
Crude estimates made by the writer using such data 
as are available indicate that very considerable amounts 
of energy are transferred poleward by this means. How- 
ever, whereas the transport of internal heat energy 
cannot take place without a proportional transport of 
kinetic energy, the meridional transfer of latent heat 
energy is not related to the kinetic energy in the same 
way. It is thus possible to transfer large amounts of 
energy in the form of latent heat without a proportional 
kinetic energy flow being present. Whereas the flow of 
kinetic energy and therefore of internal heat energy 
may be regulated by the distribution of pressure and 
of divergence and convergence, this particular limita- 
tion does not enter as far as latent heat is concerned. 
One could therefore find arguments to support the view 
that the transport of latent heat furnishes a means for 
balancing the radiation requirements of the general 
circulation which bypasses the kinetic energy balance. 
For if the portion of the total meridional energy flow 
represented by the transfer of latent heat energy had 
instead to be transported as additional internal heat 
energy, a far more intense flow of kinetic energy would 
also be necessary. Whether this would imply a more 
“vigorous” general circulation is, however, a question 
which cannot be readily dealt with. 
The meridional transport of geopotential energy is 
of course immediately ruled out im the average con- 
dition discussed here, because of the assumption that 
mean meridional circulations are not significant. 
In the discussion mentioned above [11], an endeavor 
was made to study the fluctuations in the kinetic energy 
balance of the Northern Hemisphere with the aid of 
5-day mean data such as are used by the U.S. Weather 
Bureau in extended-period forecasting. Only data for 
the colder half of each of seven successive years were 
used, so that essentially winter conditions were studied. 
By approximate methods a measure was secured of the 
intensity of differential advection of air with varying 
density across 45°N. This quantity 7 is therefore a 
measure of the net volume transport northward across 
45°N and hence also a measure of the mean convergence 
in the polar cap and to some degree probably a measure 
of the mean divergence in the more tropical zones of 
the Northern Hemisphere. A mean for the layer be- 
tween sea level and 10,000 ft was used. ; 
It became immediately apparent that the quantity 
r undergoes large fluctuations, being high during those 
periods which are commonly designated as “low-index” 
periods and low during “high-index” conditions. The 
practical question involved here concerns itself with 
the cause of these vagaries in the behavior of the general 
circulation, for if this were known a better insight into 
the long-range forecasting problem might result. 
In view of the fact that 7 is also a rough measure of 
the rate of kinetic energy transport into the polar cap, 
it is interesting to note its relationship to the general 
pressure distribution there. If these pressures are low 
compared to the pressures farther south, a relatively 
— 
