
May 19, 1923] 
NATURE 
675 

statics, where we have to deal with the forces produced 
by positive on positive, negative on negative, and 
positive on negative, we have, in addition, for moving 
electrons, the force due to the motion of a positive 
electron on a moving positive electron, the force due 
to the motion of a moving negative electron on a moving 
negative electron, the force due to the motion of a 
negative electron on a moving positive electron, and 
the force due to the motion of a positive electron on 
a moving negative electron. The first two of these 
_ four may be taken as the basis for defining the 
measures of the two types of magnetic fields produced 
by positive and negative electrons respectively. If, 
for similar motions, these four forces are all equal, 
a moving electron, or a magnet, would be entirely 
unaffected by the rotation of the earth as a whole. 
_ If, however, the forces, due to motion, between unlike 
moving charges are suitably different from those 
between like charges in the same states of motion, it will 
immediately appear that the electrically neutral earth 
will, by its rotation, produce those forces on magnets 
and moving electrons which we associate with a magnet 
as ordinarily defined. By making the forces between 
electrons of like sign equal for both signs, the force 
due to the motion of a negative electron on a moving 
positive electron greater than, and the force due to 
the motion of a positive electron on a moving negative 
electron less than the forces between like electrons to 
the extent of about two parts in ro!*, we can account 
for the equivalent of a magnetic field of the order of 
magnitude of the earth’s magnetic field. If we wish 
to combine these alterations with suitable alterations 
in the electrostatic forces, we can also include gravita- 
tion in the complete scheme. 
The secular variation presents interesting problems 
for speculation. There is some evidence for the belief 
that the earth’s magnetic axis rotates about the geo- 
graphic axis once in about 500 years. This will result 
in induced currents, and the field we observe. will be that 
due to these induced currents (the secondary field), and 
that due to the primary causes (the primary field). 
Taking an iron sphere of the earth’s size for purposes of 
illustration, it works out that the flux of the secondary 
field through the sphere, which is, of course, related to 
that of the primary field, is of such magnitude as to 
annul almost completely the non-axial component of 
the primary flux, leaving only a small residual non-axial 
component, which lies, moreover, perpendicular to the 
primary non-axial component. Thus, in order that the 
resultant flux shall have an appreciable inclination to 
the geographic axis, it is necessary for the primary axis 
to lie very near to the equatorial plane, and yet for the 
primary flux to be so large that its axial component, 
which is small compared with it, represents the axial 
component which we observe. This example is given 
merely to illustrate the important réle which might be 
played by the induced currents due to the secular varia- 
tion in case the earth’s interior had a conductivity 
comparable with that of iron. 
The theory of the diurnal variation is in a better 
position than that of the earth’s field as a whole. The 
suggestion of Balfour Stewart, developed in detail by 
Schuster, to the effect that the diurnal variations are 
caused by Foucault currents generated in the upper 
atmosphere by the tidal motion of the atmosphere 
NO. 2794, VOL. 111] 
across the earth’s lines of force, seems well adapted to 
fit the facts, its chief difficulty being that it calls for a 
conductivity about 3x10" times that found at the 
earth’s surface. Various agencies have been invoked to 
account for this conductivity, namely, ultraviolet light, 
gamma rays, negative electrons, and alpha rays, from 
the sun, and finally charged atoms of gas, shot out from 
the sun by the pressure of light, and endowed thereby 
with velocities sufficient to give them the properties of 
low energy alpha particles. The corpuscular radiations 
have also been invoked to account for the phenomena 
associated with the aurora. 
lt is probable that ultraviolet light plays no im- 
portant réle, since it is capable of accounting for a con- 
ductivity less than one-millionth of the conductivity 
required. As regards the corpuscular radiations, the 
nature of the precipitation of corpuscles indicated by 
the aurora is of a type to correspond to a bending by the 
earth’s magnetic field such as one would not readily 
associate with particles of mass as small as that of 
electrons. The mass of an electron increases with its 
velocity ; but, so greatly has Birkeland found it neces- 
sary to draw upon this phenomenon in order to fit the 
facts, that, on the hypothesis of negative electrons, he 
is driven to assume velocities ranging from 400 metres 
per second less than the velocity of light to 4 metres 
per second less than that limit. Alpha particles have 
a mass and energy which would be better adapted to 
account for the aurora, as has been pointed out by 
Vegard ; moreover, the definiteness of their range 
ensures the characteristic feature of the sharp boundary 
of the luminescence, and the magnitude of the range 1s 
fully sufficient to account for the penetration of that 
boundary to the altitudes observed. 
The remarkable perturbations of the earth’s magnetic 
field known as magnetic storms, which occur most fre- 
quently in association with high solar activity, suggest 
the entry into our atmosphere of electrified corpuscles 
during these periods, and it is natural to look to those 
corpuscles which are responsible for the conductivity 
and the aurora for an explanation of these storms. 
While alpha rays have been suggested, some of the 
difficulties inherent in the East ton may be gathered 
from considerations put forward by Lindemann. On 
the assumption of their production by alpha rays, these 
storms would call for an incredibly large amount of 
radioactive material in the sun. Again, a conical beam 
of alpha rays, such as appears to be necessary to account 
for the storms, would, on its journey here, suffer, by 
self-repulsion, an acceleration of about 101% cm,/sec.” at 
its boundary, in such a sense as to make it spread, so 
that it could never arrive as a beam. Finally, even if 
the beam could reach our atmosphere, it would charge 
it at such a rate that the repulsion due to the charge 
which had arrived would, in a few seconds, attain a 
value sufficient to prevent the entry of any more rays. 
It is for reasons such as these that Lindemann has 
been led to favour the view that the primary agencies 
responsible for magnetic storms are atoms of gas, 
ionised by the high temperatures in the solar promin- 
ences, and shot out of them by the pressure of the sun’s 
radiation. He shows, moreover, that the velocities to be 
expected in these circumstances are such as to give the 
particles ranges in harmony with the requirements of 
auroral phenomena. 
