300 
obtained with somewhat different assumptions than the 
ones on which the published calculations are based. 
The assumptions that differ markedly from those out- 
lined above are: 
1. Solar Energy. The solar emission curve in the ultra- 
violet corresponds to the results of the rocket measure- 
ments. As mentioned previously, these show consider- 
ably less solar energy than a black body at 6000K 
would radiate. 
2. Vertical Distribution of Absorbing Gases. The ver- 
tical distribution of ozone above 35 km is computed 
from photochemical-equilibrium theory [10]. This gives 
the same magnitude as Umkehr measurements and 
shows to a first approximation variations with solar 
zenith angle and temperature. 
3. Absorption Spectra. The emissivities of carbon di- 
oxide and ozone for small path lengths are obtained 
from the theoretical transmission functions rather than 
from straight extrapolation of existing measurements. 
Cooling is evaluated directly by graphical integration 
of (28). 
For the temperatures generally assumed to occur in 
the ozone layer [57], the calculated rates of heating and 
cooling are of the same order of magnitude at all levels. 
This contrasts with Gowan’s results; since his equi- 
librium temperatures are considerably higher than those 
that actually occur, his assumptions would give greater 
heating than cooling at the lower temperatures. It 
also contrasts with Penndorf’s results. In both cases, 
the difference in assumed solar energy is the principal 
reason for the discrepancy. The level of maximum 
heating is at 40-45 km, lower than that found by Karan- 
dikar. This difference is probably also a result of the 
different assumed-energy curves. The ratio of the meas- 
ured solar intensity to the black-body intensity is 
relatively smaller near the maximum of the Hartley 
bands of ozone, which heat the upper part of the ozone 
layer, than at longer wave lengths. 
The infrared cooling due to the carbon dioxide band 
at 15 w is more intense than that due to ozone or water 
vapor, at least above 35 km. This result must be con- 
sidered somewhat doubtful until further light is shed 
on the problems of pressure effects on this band, and 
of the overlapping of the 15-4 ozone band. The ab- 
solute values of the rates of heating and cooling are 
relatively low, of the order of magnitude of 0.1C per 
three hours below 30 km and of 1C per three hours 
at 35-50 km. These are no larger than might be ex- 
pected from normal atmospheric circulation processes. 
SUBJECTS FOR FURTHER RESEARCH 
It has become evident throughout this discussion 
that several fruitful avenues of research lie open to the 
interested investigator. In this concluding section, these 
are summarized and briefly discussed. 
Solar Energy in the Ultraviolet. Even though rocket 
measurements of the solar spectrum down to 2200 A 
have thus far been of great help, many more are needed. 
Kor the present problem, only a slight further extension 
into the ultraviolet is necessary, perhaps to 1800 A. 
However, many more measurements should be made to 
THE UPPER ATMOSPHERE 
check the accuracy of the information now available 
and to show whether there are any significant variations 
of the spectrum with time. 
Vertical Distribution of Absorbing Gases. The verti- 
cal distributions of all of the absorbing gases in the 
ozone layer are in doubt. In the case of ozone, the 
vertical distributions above 30 km at various latitudes 
and times of year are urgently needed. For carbon 
dioxide, some direct measurements should attempt to 
test the usual assumption that the concentration in the 
ozone layer is the same as in the troposphere. Particu- 
larly above 30 km this is desirable. Water vapor is 
probably not important to the present problem at 
levels above 30 km (unless it is present in much greater 
concentration than now assumed). However, in the 
lower isothermal part of the layer its concentration 
should be measured carefully under a variety of condi- 
tions. Recent developments in England [11] and the 
United States [2] give encouragement in this direction. 
Absorption Spectra. The absorption spectra of oxy- 
gen and ozone in the pertinent part of the ultraviolet 
seem to be satisfactorily known. For the problem of 
calculation of photochemical-equilibrium amounts of 
ozone, however, further study of the pressure depend- 
ence of oxygen absorption should be made in the 
laboratory. The spectral region where information is 
vitally needed is 1800-2200 A and the pressures range 
from 10 to 0.1 mb. 
Absorption data in the infrared are urgently needed. 
Perhaps the most practical type of laboratory measure- 
ments for the present problem would give the iso- 
thermal emissivities of carbon dioxide and ozone as a 
function of path length, pressure, and temperature, par- 
ticularly the first two. The range of path lengths of 
carbon dioxide that needs further study is from 0.01 to 1 
cm NTP at pressures ranging from 0.1 to 10 mb. Ozone 
path lengths from 10~* to 10 em NTP at pressures of 
1 to 10 mb need further study. Furthermore, emis- 
sivity measurements should be made on various mix- 
tures of ozone and carbon dioxide within these lmits 
to determine the effect of the former on the latter at 
15 p. 
Calculations with Existing Data. Further calculations 
with existing data are possible, and deed desirable, 
unless some of the above experimental information is 
forthcoming in the immediate future. The calculations 
made up to now show that useful results can be ob- 
tained. Particularly, calculations for various latitudes 
and seasons may begin to show the extent of tempera- 
ture variations in the ozone layer, a matter about which 
we now have no information. 
REFERENCES 
1. Bamrorp, C. H., ‘“‘Photochemical Processes in an Oxygen- 
Nitrogen Atmosphere” in Reports on Progress in Physics, 
9: 75-91. Phys. Soc., London, 1943. 
2. Barrett, E. W., Hernvon, L. R., Jr., and Cartmr, H. J.. 
“A Preliminary Note on the Measurement of Water- 
Vapor Content in the Middle Stratosphere.” J. Meteor., 
6: 367-368 (1949). 
3. BRASEFIELD, C. J., “Exploring the Ozonosphere.” Sci. 
Mon., 68: 395-399 (1949). 
