506 
-10- 
1 
Wks Si Nis Gna (not considering the fluctuations of V due to pressure changes in the interval) 
where C is the capacity, R insulation resistance, and e the base of Naperian logarithms. {t will be seen 
from this relation that the larger the values of C and R, the more nearly will Velo = 1, i.e. the less 
charge will leak away. The following table gives an idea of the quantities involved:= 
¢ = 107° farad t = .0001 second 
100 megohms 
c = 10° farad t = .0001 second 
These values show that even ina time as short as 107” second there may be appreciable leakage 
if the capacity is small (e.g. 10° farad) and the insulation poor. The effect of such leakage on an 
actual p/t curve (taken from Mr. Hilliar's report) is illustrated in Figure 9. It will readily be seen 
from inspection of the deduced "leaky circuit’ curves that it is of great importance to secure good 
insulation. otherwise the following errors may be introduced:- 
(i) The pressure-time curve recorded is not the true curve for the explosion. 
(ii) The maximum pressure recorded Is less than the actual maximum. 
and (ili) When negative charges* are recorded, they do not necessarily indicate negative pressures. 
In all cases, therefore, where the record is considered reliable, it has been carefully observed 
that the insulation resistance is sufficiently high (when considered in relation to capacity). in general 
the capacities employed were of the order of 1078 farad and the insulation resistance not less than 40 
megohms. In 107 second the leakage is in such a Case considerably under 0.1%. Certain records have 
been obtained, however, under conditions where charges were fired when the insulation resistance was known 
to be low (.5 megohm). In these cases the record shows false ‘negative’ pressures - compare Figure 9 curves 
C and D, with records c, d and e, Sheet II. {In every explosion experiment therefore careful measurements 
of capacity and insulation were made immediately before and after firing the charge. Records f and g on 
Sheet !| are reproduced to compare with records c, d, e, aS showing the effects of good and bad insulation 
respectively. 
Calibration of Apparatus. 
Before the pressure-time curve recorded by the oscillograph can be interpreted it is necessary to 
know the sensitivity of the various parts of the apparatus and the speed at which the rays have traversed 
the plate. 
The voltsge sensitivity of the oscillograph and the pressure sensitivity of the p.e. gauge are, 
under the experimental conditions, interlinked but it is a mstter of some importance to know the oscillograph 
sensitivity independently, so that corrections may be applied under circumstances slightly different from 
the standard conditions. 
(a) ccvece 
The negative charge referred to in (iii) begins to appear when the positive charge, due 
to the increasing pressure of the explosion pulse, has leaked off, further diminution of 
pressure (in the falling pressure region of the pulse) resulting in a negative charge 
being produced, 
