AND ON THE ELECTRIC FIELD OF THUNDERSTORMS. 
97 
thunderstorms, and the difference in the intensities of the electric fields in the 
inner and outer areas is likely to be even greater than in the example given. As 
represented in fig. 5 lines of force from the central area end on the lower charged 
portion of the cloud, those from the outer area on the upper charge, others again 
connect the upper and lower charges. 
Tims far no account has been taken of the conducting layer in the higher levels of 
the atmosphere, to the existence of which the phenomena of terrestrial magnetism 
seem to point. 
The normal potential gradient at the surface of the ground in clear weather is of 
the order of 100 volts per metre, falling off with increasing height and becoming 
negligible above 10 km. ; thus the potential in the conducting layer over regions of 
fine weather is not likely^ to exceed a value of the order of 1,000,000 volts. If we 
assume, in accordance with modern theories of terrestrial magnetism,* that the 
conductivity of the upper atmosphere is high enough to prevent any large potential 
differences within it, then even above a thunderstorm the potential in the conducting 
layer may not greatly exceed 1,000,000 volts. The potential in the head of a 
thunder-cloud probably reaches values 1000 times as great. 
One important effect of the conductivity of the upper atmosphere is to cause a 
portion of the lines of force from the head of the thunder-cloud to end in the 
conducting layer. The effect will be more marked than that which would be 
produced by a solid conducting sheet since ions of opposite sign to the charge on the 
head of the cloud will be dragged down out of the conducting layer to form an 
expansion of it extending downwards towards the thunder-cloud. The charge on 
these ions (which constitutes the induced charge on this protuberance from the 
conducting layer) will partially neutralise the electric field produced at the ground 
by the charge in the head of the cloud ; in other words lines of force from the head 
of the cloud which would otherwise have ended on the ground are now diverted 
upwards into the conducting layer.t 
The considerations brought forward in this section suggest that the electric field 
of a cumulo-nimbus cloud may be regarded as due to charges, generally unequal, in 
the upper and lower parts of the cloud (falling rain being included as part of the 
cloud) and to the charges induced by these on the ground and on the conducting 
layer of the upper atmosphere. Thus the lines of force of the cloud may be 
divided into four classes, connecting (a) the ground and the lower charge of the 
cloud, ( b ) the lower charge and the upper charge, (c) the upper charge and the 
* Schuster, ‘Phil. Trans.,’ A, vol. 208, p. 163, i907; S. Chapman, ‘Phil. Trans.,’A, vol. 218, p. 1, 
1919. 
t In the absence of a previously existing conducting layer a thunderstorm would itself produce 
ionization in the upper atmosphere; this is readily seen to follow from the values found for the electric 
moments of lightning flashes. 
VOL. COXXI.—A. 
P 
