¥ May 12, 1923] 
NATURE 
641 

whole earth. As regards the positive charge in the 
atmosphere, there is little difficulty provided that we 
can account for the maintenance of the negative charge 
ontheearth. For, even though a theory which accounted 
for the latter did not immediately imply the former, 
the known fact of the increase of atmospheric con- 
ductivity with altitude, combined with the law of 
continuity of flow of the electric current, would be 
sufficient to bring the positive charge into evidence. 
One of the earliest theories of the earth’s charge is due 
to C. T. R. Wilson, who supposed that the atmospheric 
ions would serve as nuclei for the precipitation of rain, 
and that the drops would form more readily upon the 
negative than upon the positive ions, with the result 
that rain would be, on the whole, negatively charged, 
and would thus constitute the replenishment of the loss 
by conduction. The difficulty confronting this theory 
lies in the fact that the conditions necessary for the 
precipitation of rain on ions to form drops of appreci- 
able size, do not readily occur in the atmosphere, and 
in the still more potent fact that, so far as measure- 
ments go, 90 per cent. of the rain which falls is posi- 
tively charged. Thus, while rainfall may constitute 
a factor in the replenishment of the earth’s charge, it 
is not one which operates in the right direction to serve 
as the sole cause. 
Another theory of replenishment, depending vilti: 
mately upon gravity for the separation of the charges 
in opposition to the electric field, is that due to Ebert. 
It constitutes a modification of an earlier theory due 
to Elster and Geitel. Ebert’s theory invokes the fact 
that if an ionised gas be passed through a fine tube the 
negative ions diffuse to the walls of the tube more 
rapidly than do the positive ions. Ebert supposes 
that, during periods of fall in barometric pressure, the 
air in the pores of the soil, which is ionised on account | 
of the radioactive material therein, becomes drawn 
out into the atmosphere, positively charged on account 
of its having deposited an excess of negative ions in the 
interstices of the soil. Rising currents of air are then 
invoked to explain the transference of the positive 
ions to appreciable altitudes, against the electrostatic 
attraction of the negative. This theory has been 
criticised on account of the insufficiency of the charging 
action resulting from the diffusional process, on account _ 
of the smallness of the upward convection current of 
positive electricity as measured experimentally, and 
on account of the fact that it may be shown to predict | 
a diminution of potential gradient with altitude such 
as would result in the gradient itself being practically 
negligible at an altitude of a kilometre. 
The precipitation theory, and the Ebert theory, are 
of a type in which the replenishing action takes place 
over a limited region of the earth’s surface at any one 
time, in such a manner that the positive electricity 
which is the counterpart of the negative charge on 
the earth is to be found in this limited region of the 
atmosphere. Under such conditions, the negative 
charge will be held on the portion of the earth’s surface 
which lies in the immediate vicinity of the positive, and_ 
the potential gradient will be confined to this region. 
A partial way out of this difficulty can be found, how- 
ever, if we admit the existence of a highly conducting . 
layer in the upper regions of our atmosphere. In this ° 
case, the charge separation sets up a potential difference 
NO. 2793, VOL. I1T] 



between the layer and the earth, so that the potential 
gradient, which would otherwise be confined to the 
region of replenishment, is shared as it were by the 
earth asa whole. Thus, for example, calculation shows 
that if a charged cloud is to be found at an altitude h 
above the earth’s surface, and if H is the altitude of 
the conducting layer, and R the radius of the earth, 
the hemisphere of the earth which is symmetrically 
remote from the charged cloud receives R/H times the 
number of tubes of force which it would receive in the 
absence of the layer, and h/H times the number which 
it would receive if the negative charge on the earth 
and the positive charge in the cloud were spread 
uniformly over the earth and atmosphere respectively. 
It may be remarked, moreover, that this action of the 
conducting layer provides a partial loophole for escape 
from the particular objection to the Ebert theory which 
is founded on the impossibility of the positive charge 
reaching an altitude of more than a kilometre or so. 
Even such a small separation in the region of replenish- 
ment would make its own contribution to the potential 
gradient at other places through the medium of the 
conducting layer. The contributions in these places 
would be of a perfectly normal type, the variation with 
altitude being determined only by the nature of the 
variation of conductivity with altitude, in such a 
manner as to keep the vertical conduction current 
density independent of altitude. 
In 1904 G. C. Simpson proposed a tentative theory 
-of the earth’s charge, in which it was assumed that the 
sun emitted negative and positive corpuscles of high 
penetrating power. The former were supposed to pass 
right through our atmosphere and penetrate the earth, 
while the latter were caught in the atmosphere. Such 
a degree of penetration is very much greater than any 
we are familiar with in the laboratory, for the beta rays 
of highest energy investigated will pass through only 
about ro metres of air. 
We can account for the replenishment of the earth’s 
charge if we suppose that the atmosphere emits high- 
speed negative corpuscles. The earth will then charge 
up on account of the corpuscles which come from the 
molecules of air lying within striking distance of it. 
Such a possibility was examined by the writer in 1915. 
So far as the replenishment of the charge is concerned, 
the average range of the corpuscles may be made as 
small as we please by supposing a sufficiently copious 
emission of corpuscles. It turns out, however, that 
appreciable values of the potential gradient become 
confined to altitudes comparable with the average 
range, so that for this reason a large range must be 
assumed. This difficulty is avoided in a somewhat 
similar theory suggested by the writer, and somewhat 
later, but quite independently, by von Schweidler. 
According to this theory, the emission of corpuscles 
from the atmosphere is caused by the penetrating 
radiation which, coming from above, and being of a 
very hard type, ejects the corpuscles almost completely 
in a downward direction. If we assume that only 
three corpuscles are emitted per c.c. per second, by 
the penetrating radiation, an average range of nine 
metres in air at atmospheric pressure is sufficient to 
account for the replenishment of the earth’s charge. 
Two main difficulties confront any corpuscular 
theory of the earth’s charge. The first arises from the 
