THE PHYSICAL BASIS FOR THE GENERAL CIRCULATION 
region where the strongest westerlies are actually en- 
countered. The writer rather inclines to the view that 
although very small mean meridional circulations do 
perhaps exist, their role in the horizontal transport of 
angular momentum, at least in the middle latitudes, 
is overshadowed by the characteristics of other hori- 
zontal motions. This is notin conflict with the views 
expressed by Jeffreys [5] and is reinforced by analysis 
of data to be quoted presently. 
If emphasis is placed upon the horizontal circulations 
in transporting angular momentum poleward, by and 
large the zonal and meridional components of velocity 
should be correlated at each level where such transport 
occurs. Hvidences of this correlation are then to be 
expected in the detailed structure of the instantaneous 
horizontal streamline patterns observed, as pointed out 
by the writer elsewhere [10]. Qualitatively these ex- 
pectations are amply borne out even by casual inspec- 
tion of weather maps. Except at high latitudes, closed 
horizontal circulations usually exhibit a northeast- 
southwest elongation (Northern Hemisphere) and the 
troughs, ridges, and shear lines show a tendency to 
tilt m this same sense. All these characteristics are 
recognized almost instinctively by the meteorologist as 
typical of atmospheric flow patterns. From our view- 
point they are telltale indications of a poleward flow 
of angular momentum. One may nevertheless ask 
whether an objective quantitative evaluation of the 
transport by this mechanism can be made from actual 
hemispheric data. For it is not sufficient to advance 
merely a qualitative substitute for the classical hy- 
pothesis without investigating the potency of the new 
alternative to produce the needed effect. A question 
involved here is whether the observations we possess 
are extensive enough and of sufficient accuracy. It is 
a simple matter to set up an integral expression for 
this (say) northward flow of absolute angular momen- 
tum across the vertical surface at a given latitude after 
the manner of Jeffreys, or to derive it from the atmos- 
pheric equations of motion as has been done by Widger 
[17]. The latter author proceeded to evaluate this flow 
of angular momentum by finite difference methods, as 
follows. 
During the last several years sufficient observations 
of the free atmosphere have been made to allow the 
construction of daily isobaric charts through most of 
the troposphere. In addition it is becoming possible to 
obtain fairly complete direct radiowind observations 
extending to great heights on a circumpolar basis within 
restricted latitude belts. The global wind distributions 
may be approximated from isobaric charts according to 
the geostrophic wind formula or may be taken directly 
from actual wind observations. The use of the geo- 
strophic estimates may introduce certain errors for the 
present purpose. On the other hand this use of the 
geostrophic winds automatically excludes the contri- 
butions to the angular momentum transport due to 
mean meridional circulations of the Hadley type, so 
that an advantage is gained if it is desired to study other 
modes of such transport. Several surveys of the angular 
momentum balance have been made in the past few 
543 
years (Widger [17], Mintz,? Starr and White [12]) both 
from isobaric charts making use of the geostrophic ap- 
proximation, and directly from a circumpolar network 
of actual wind observations. The survey of the angular 
momentum balance made by Widger covers the month 
of January 1946 and the results give the total geo- 
strophic transfer by latitudes for various layers during 
this period up to the 7.5 km level. In this study estimates 
were made of the surface frictional torques. The investi- 
gation by Mintz covers the month of January 1949 and 
his results give the geostrophic flux of angular momen- 
tum at various levels up to 100 mb. Starr and White 
made use of a circumpolar network of actual wind ob- 
servations at a mean latitude of 31°N for a period of 
six months from February 1949 to August 1949 up to 
an elevation of 50,000 ft. For various details of these 
investigations reference must be made to the original 
papers. 
The computations give results which are entirely 
reasonable, being quite in accord with what would be 
expected on the basis of the foregoing discussion. Thus 
the total northward transport increases in magnitude 
from low latitudes to about 30°N as the surface easterlies 
are passed, then decreases progressively northward as 
angular momentum is removed by surface frictional 
torques acting in the westerly belt. Nearer to the pole 
the transport is reversed, indicating a flow southward 
from the polar easterly zone, although the magnitudes 
involved here are small as is to be expected from the 
fact that the torque arm associated with the surface 
frictional forces is small in the polar regions. 
The main transfer across 30°N increases in intensity 
with elevation, reaching a pronounced maximum at 
about the level of the jet stream. On the basis of esti- 
mates of the surface torques during these periods it 
appears that sufficient angular momentum is trans- 
ported into the belt of westerlies to maintain them 
against friction. Let it be noted, however, that the gen- 
eral results appear to be in harmony with the thesis 
that practically all the necessary horizontal transfer of 
angular momentum could be accomplished without recourse 
to the agency of mean meridional circulations. On the 
basis of what has been said, however, we cannot make 
any statement concerning other possible functions of 
meridional circulations such as, for example, the vertical 
transport of angular momentum. 
At this point let us pause in order to take stock of 
what has been described and to see more clearly how it 
fits into the plan for research advanced in the intro- 
ductory paragraphs. Have we by this study of angular- 
momentum considerations provided a theory for or 
achieved a rational solution for the problem of the 
distribution of zonal westerlies and easterlies in the 
atmosphere? Not by a long way. We did not solve the 
equations of motion® nor did we deal with radiative 
2. “The Geostrophie Meridional Flux of Angular Momen- 
tum for the Month of January 1949.’’ Presented at the 109th 
national meeting of the American Meteorological Society, 
Jan. 29-Feb. 1, 1951, New York. 
3. Actually only one equation of motion, namely the one for 
the zonal direction, is needed to treat the balance of absolute 
