258 
effect, it is very necessary to know how the phase of 
the tide varies with latitude. It appears that there are 
more observational stations in the Northern than in the 
Southern Hemisphere. Consequent lack of data is a 
great handicap in obtaining a complete world picture 
of the ionosphere. More ionosphere stations, partic- 
ularly in the southern latitudes, are therefore very 
necessary. It has been suggested that a sea expedition 
be sent out to secure observations at different latitudes 
and longitudes which are scientifically important but 
for which data are lacking. . 
In the night-sky spectrum there are still lines and 
bands the origin of which is uncertain. Special interest 
lies in the identification of 3471 A. If this be due to 
atomic nitrogen, then photo-dissociation of Ne» like 
that of O» will have to be assumed, though laboratory 
experiments do not show any such dissociation effect. 
It is satisfactory to note that the strong radiations 
6580 A in the red and 10,444 A in the infrared, the 
origins of which had long been matters of controversy, 
have now been identified as bands due to the radical 
OH [65]. Bands of O2 are reported to have been identi- 
fied in the night-sky spectrum. If this is confirmed, 
then there must be another luminescent layer—per- 
haps at the level of the E-layer. The presence of sodium 
D-lines has been established and the location of the 
luminescent sodium layer has been studied. But the 
presence of sodium in the high atmosphere has raised 
many questions which are still unanswered. The con- 
tinuous spectrum which forms the background of the 
lines and bands of the night-sky ight appears to have 
recelved inadequate attention and awaits further study 
regarding its intensity variations and origin [80]. 
The intensity variation of night-sky light has a 
regular part (diurnal and seasonal) and an irregular 
part. Some investigators have sought to associate the 
irregular part with the magnetic disturbances. Some 
correlation has been found, but the whole subject is 
still im a speculative state and requires more intensive 
study. 
A very plausible hypothesis regarding the excitation 
of the atomic oxygen lines and of the observed No» 
bands (first positive and Vegard-Kaplan bands) has 
been given [68], but the subject is still controversial [11]. 
Several attempts have been made to determine the 
heights at which night-sky luminescence originates. 
The values obtained vary from 60 km to a few hundred 
kilometres (21, 22, 35]. It is very likely that, like the 
several ionospheric regions, there are several layers of 
maximum luminescence. It is highly desirable that the 
many ionospheric stations which are being established 
in different parts of the world have attached to them 
observatories for study of night-sky luminescence. 
There are reasons to believe that some at least of the 
luminescent layers may be identified with some of the 
ionospheric layers [39, 69]. Close comparison of the 
variations of the night-sky intensity with those of 
ionized regions will help in the determination of the 
existence of such association. 
Of the other lights from the night sky, the mystery of 
zodiacal light is still far from solved. There is little 
THE UPPER ATMOSPHERE 
doubt that the light is due to scattering by some sort 
of dust cloud in extraterrestrial space. But the location 
of the cloud is still a matter of controversy; whether it 
belongs to the earth or to the sun is not known defi- 
nitely. Many theories as to the origin of this cloud 
have been put forward, but none of them can be con- 
sidered fully satisfactory [87, 48, 103]. 
The problems of the aurorae and of the incidence of 
the magnetic disturbances, so closely related to each 
other, are only partially solved. There is little doubt 
that the primary cause of these phenomena is emission 
of fast charged particles from the sun. It may be noted 
that direct evidence has been obtained of the existence 
of Cat ions and protons between the earth and the sun 
during magnetic storms [1, 25, 65a]. That these may be 
at least one of the kinds of charged particles emitted 
from the sun had long been suspected, but the mecha- 
nism of the emission of such particles is still only guess- 
work. Further, the fundamental dilemma still remains: 
The charged particles, to produce the observed effects, 
must arrive in bundles, whereas they are bound to be 
dispersed on their way by electrostatic repulsion. At- 
tempts at solving this problem by imagining that the 
particles constitute a neutral beam have met with 
only partial success [2, 29]. (See, however, [59a].) 
The lines and bands of the auroral spectrum may be 
classified under two heads: first, spectra excited by 
direct bombardment of charged particles—first nega- 
tive bands of N3; and second, spectra excited as a 
result of reactions amongst neutral and charged parti- 
cles produced by the bombardment (second positive, 
first positive, and the Vegard-Kaplan bands of No, as 
well as the forbidden lines of O [70]). Of these, the 
excitation processes of the first positive and the Vegard- 
Kaplan bands may be the same as those of the night- 
sky light spectrum. But there remains the problem 
that whilst the Vegard-Kaplan bands are strong in the 
night sky they are comparatively faint in the auroral 
spectrum. It has been suggested that this might be due 
to the fact that the auroral spectrum originates at 
much lower heights than the night-sky luminescent 
layer and as such the metastable N.(A) molecules 
from which these bands originate are de-excited by 
collision. Howevér, objection to this has been raised 
on the ground that the strength of the Vegard-Kaplan 
bands does not increase with height in the spectra of 
auroral streamers. The excitation of second positive 
bands has been suggested as due to radiative recom- 
bination of N+ ions and electrons. But here again it 
may be objected that the calculated intensity of such 
radiation is very small compared to the observed in- 
tensity. 
In regard to the insufficiency of laboratory data which 
is still standing in the way of upper atmospheric study, 
mention might be made of the electronic spectra and of 
the absorption coefficients of Ne, Ox, and O in the 
extreme ultraviolet region. More complete knowledge 
is very necessary because the ionization densities of the 
ionospheric regions are controlled by the photo-ioniza- 
tion of these particles by absorption in the extreme 
ultraviolet. A more detailed study by the quantum- 
