AND ON THE ELECTRIC FIELD OF THUNDERSTORMS. 
81 
potential were negative—excepting the positive components of the multiple flashes— 
the largest amounting to —9600 volts per metre. The distance of this discharge, as 
is shown by the interval elapsing before the thunder began to be heard, was about 
4‘3 km. The distances of the others ranged between 4‘3 and 57 km. The peals of 
thunder, as the intervals between the single and the double black lines show, were 
very long, some lasting for as many as 40 seconds. 
Fig. 7 (June 16, 1917, 19h.'31m. 10s. to 19h. 36m. 45s.). 
This is a portion of the record taken with the test-plate next after that shown in 
fig. 6. Rain continued to fall throughout the time of exposure. The cover was 
removed from the test-plate at 19h. 31m. 30s. ; the potential gradient at that 
moment was negative, —6500 volts per metre; a lightning discharge had probably 
occurred immediately before the exposure of the test-plate. The discharge at 
9h. 33m. 20s. was really multiple, the sudden changes of potential gradient being —2900, 
— 5100 and +1300 volts per metre. The discharge at 19h. 35m. 55s. was negative 
(change of potential gradient = —4100 volts per metre) and was at a distance of 
about 5'5 km. The characteristic form of the recovery curve following the discharges 
is modified by the superposition of a general downward slope which represents an 
electric current from the ground to the atmosphere ; this current was probably mainly 
carried by falling negatively charged raindrops. The potential gradient when the 
cover was replaced at 19h. 39m. (beyond the limits of the portion of the record 
reproduced) was positive and exceeded 2500 volts. The total quantity of electricity 
transferred per sq. cm. of the test-plate to the atmosphere, during the whole 
8^- minutes of exposure, but mainly after 19h. 36m. was the equivalent of 40,000 
volts per metre, i.e., 3‘5x 10~ n coulomb. 
Fig. 8 (May 29, 1917, L9h. 4m. 10s. to 19h. 11m. 50s.). 
This is the final portion of a record which began at 18h. 44m. The sphere was 
used as the exposed conductor. The potential gradient had been +1200 volts 
per metre at 18h. 46m. when the sphere was raised; 980 at 18h. 51m. and —1400 at 
18h. 56m., at which times the sphere was momentarily lowered. The peak shown in the 
figure at 19h. 6m. represents the effect of again momentarily lowering the sphere and 
indicates that the gradient was still negative, being equal to —430 volts per metre. 
At 19h. 11m. when the sphere was finally lowered the potential gradient had again 
become positive, being now +20 volts per metre. 
These comparatively gradual changes of potential gradient accompanied the passage 
of towering cumulus clouds at no great distance. Superimposed upon them are 
sudden changes (amounting at most to 150 volts per metre) produced in the field by 
frequent discharges of more distant thunder-clouds. Some of these are positive, some 
negative ; the discharges of either sign are alike in being followed by the charac¬ 
teristic curve of recovery of the field. 
Fig. 9 (May 29, 1917, I7h. 58m. 45s. to 18h. 12m. 30s.). 
This is an enlargement of a portion of a record which extended from 17h. 56m. to 
VOL. CCXXI.-A. 
N 
