GENERAL ASPECTS OF UPPER ATMOSPHERIC PHYSICS 
above the observer would appear to him absolutely 
opaque; a line drawn from the observer towards any 
direction would pass through many molecules one be- 
hind another. If the observer continues his ascent, the 
molecules overhead will gradually thin away and he will 
reach a level where the molecules overhead will just fill 
the sky, that is, a line drawn vertically upwards will pass 
through only one molecule, whereas a line drawn in any 
other direction will pass through more than one mole- 
cule. Mounting still higher, the observer will find his sky 
overhead gradually clearing up and he will “‘see’”’ a cone 
with its axis vertical within which his sky will be clear. 
Tt is obvious that this cone will open out with height 
and the observer will finally ‘‘see” the whole sky clear. 
This cone is called by Milne the cone of escape of the 
molecules. A molecule moving within the solid angle 
of this cone will have some chance of escaping without a 
collision. The escape velocity v is given by v? > 2ga?/r, 
where v is the velocity of the molecule, g is the accelera- 
tion due to gravity at the level from which the mole- 
cule escapes, a is the earth’s radius, and r is the distance 
of the level of escape from the centre of the earth. 
Since the mean velocity of the molecules depends 
upon temperature it may be expected that a consider- 
able quantity of gas, particularly of the light variety, 
will escape from the upper atmosphere if a high tem- 
perature is assumed [49]. The problem of helium in this 
connection is particularly mteresting. Measurements 
show that the atmosphere near the surface of the earth 
contains 5 X 10+ per cent helium by volume. If, 
however, account is taken of the helium discharge from 
the earth’s crust during the geological ages, the amount 
of the helium in the atmosphere should be very much 
more than this. The reason for the scarcity of helium is 
believed to be that it is contiually diffusing upwards 
and escaping. In order that the escape may be possible 
a temperature of the order of 1000K has to be assumed 
near the limit of the atmosphere. We shall see later 
that considerations of a number of other atmospheric 
phenomena lead to a similar conclusion. 
The Fringe Region or the Exosphere. Beyond the 
limit of the atmosphere as discussed above, the mole- 
cules move freely with the velocity acquired at their last 
collision in the lower region and, being subject only to 
the force of gravity, describe elliptic, parabolic, or hy- 
'perbolic paths according to the magnitude of their 
velocities. Escaped molecules whose velocity is not 
enough to overcome the gravitational pull, will fall back 
to the atmosphere after describing elliptic paths. These 
high-flying particles constitute the fringe region or the 
exosphere of the atmosphere. The exosphere obviously 
commences at the level where the semivertical angle of 
the cone of escape approaches 90°. In the exosphere the 
particles move without collision in enormous orbits. The 
heights to which these particles will rise depend on the 
magnitude and direction of velocity acquired at the last 
collision and also on g at the point of collision. A few 
tracks of such particles are shown in Fig. 1. The merging 
of the atmosphere with interstellar space (average den- 
sity of matter one particle per cc) is estimated to take 
247 
place in the region about 2500 km above the surface of 
the earth [73]. 
If the atmospheric particles are ionized, the phenome- 
non of escape becomes complicated. This is because the 
Fie. 1.—Trajectories of a particle projected from 1000-km 
level and moving freely without collision. Velocity 3.5 km 
sec!; initial angle with the vertical—(a) 60°, (b) 45°, (ce) 30°. 
motion of an ionized particle is profoundly influenced 
by the terrestrial magnetic field. Thus, although for a 
neutral particle the critical velocity of escape is inde- 
pendent of latitude, it is not so for an ionized particle 
[73]. 
According to Vegard |103], the topmost layer of the 
atmosphere is an electric double layer formed by elec- 
trons, ions, and neutral molecules, the sun being sup- 
posed to emit radiation corresponding to soft X rays. 
The particles in the double layer are disposed of by 
the earth’s magnetic field in a manner similar to that of 
matter in the solar corona, which is also known to be 
highly ionized. The terrestrial atmosphere is thus sup- 
posed to be topped by a corona—a cloud of particles— 
similar to the solar corona [103]. 
According to Hulburt, the particles in the exosphere 
may be the scattering matter of the zodiacal pyramid 
which is seen in the evening (or early morning) to rise 
above the horizon as a faintly luminous cone after the 
disappearance (or before the appearance) of twilight 
[48]. 
In connection with the above two hypotheses of 
Vegard and of Hulburt it should be mentioned, how- 
ever, that according to some authors, the particles spend 
such short time in the transition layer or in the exosphere 
that they have not much chance of being ionized [98]. 
SOLAR CONTROL OF THE UPPER ATMOSPHERE 
The upper atmospheric regions are under strong solar 
control due, firstly, to absorption of the whole of the 
ultraviolet radiation below 2900 A and, secondly, to 
bombardment by charged particles emitted by the sun. 
The ultraviolet absorption produces dissociation, al- 
lotropic modification, and ionization of the upper at- 
mospheric gases. The bombardment by charged parti- 
cles—which is concentrated round the magnetic axis 
poles—produces auroral displays associated with optical 
excitation and ionization of the upper atmospheric 
gases. The effects of the different parts of the ultra- 
violet solar spectrum are presented in Table I. 
