AND ON THE ELECTRIC FIELD OF THUNDERSTORMS. 
79 
per metre. At 14h. 11m. 10s. distant electrical charges which were responsible for a 
portion (amounting to 150 volts per metre) of the negative potential gradient at the 
place of observation were neutralised by the passage of a lightning flash. The negative 
potential gradient at once began to be regenerated but was again suddenly diminished 
about 3 seconds later, losing 25 volts per metre by the passage of a lightning flash, 
probably at a still greater distance. This continuous production of a negative potential 
gradient and its sudden diminution at intervals by lightning discharges continues 
throughout the record. At about 14h. 13m. 40s. a sudden change of potential gradient 
of positive sign occurred, but was followed by one of negative sign and of nearly 
equal magnitude about 0‘4 second later, a small positive change again occurring after 
another almost equal interval; these changes of potential gradient amount to +240, 
—220 and +25 volts per metre respectively. Another negative change of potential 
gradient (about 60 volts per metre) is indicated 10 seconds later. A few seconds 
after 14h. 16m. the record shows two discharges to have occurred with an interval of 
2'4 seconds between them; each produced a change of potential gradient of positive 
sign, the first amounting to 840, the second to 870 volts per metre. 
The potential gradient at any moment may be regarded as being the resultant of 
several electric fields, including those due to charges concentrated in different thunder¬ 
clouds or different centres of activity in the same cloud. The passage of a lightning 
flash results in the sudden destruction of one of these constituent fields. This at 
once begins to be regenerated by processes going on in the thunder-cloud at a rate 
which is indicated by the slope of the curve. The curve of recovery of the electric 
field (approximately logarithmic) shown after the discharges of 14h. 14m. is quite 
typical; similar curves appear in most of the records, a specially striking example 
being that of fig. 11 (Plate 4). 
On account of the very short intervals between the successive peals of thunder, 
the times at which they began and ceased to be heard were not systematically 
recorded during the record reproduced in fig. 3. The first peal of thunder recorded 
is marked by the single and double dark lines as beginning at 14h. 13m. 8’9s. and 
ending at 14h. 13m. 15s., and a second one as beginning at 14h. 13m. 18'4s. and ending 
at 14h. 13m. 30s. The two peals are taken as being due to the discharges at 
14h. 12m. 47'7s. and 14h. 12m. 53'6s. respectively; this gives a distance of 7'1 km. 
for the first and of 8'3 km. for the second. The first of these discharges produced a 
total change of potential gradient of 350 volts per metre, but this took place in two 
stages of 220 and 130 volts per metre which were separated by an interval of about 
0'2 seconds; this interval is barely distinguishable in the reproduction. The discharge 
at 8'3 km. produced a change of about 95 volts per metre. The peal of thunder 
marked as beginning at 14h. 14m. 9s. probably belongs not to the flash at 14h. 14m., 
but to an earlier one, possibly the double one at 14h. 13m. 40s. 
Fig. 4 (August 9, 1917, 14h. 45m. to 15h. 2m. 30s.). 
Here the test-plate was exposed in place of the sphere. The potential gradient 
