Abstract
The current in a parallel plane gap during the long formative time lags of sparks in air has been measured from 106 amp a few μsec after application of the voltage up to 102 amp just before breakdown, for time lags between 10 and 100 μsec. Calculations for the increase of exp(0dαdx) due to field distortion by positive ions are in good agreement with an observed rapid increase of current shortly before breakdown. The Townsend discharge is shown to be spread over the whole electrode surface in contrast to the filamentary nature of the final spark. The photosensitivity of the cathode surface has been found to decrease following collection of positive ions. The conditioning of the electrodes observed by Fisher and Bederson is believed to be due to such a decrease in γp. Some gain has been made toward a theoretical solution for the growth of the Townsend discharge. Working with an integral equation and taking due care with integration limits, has established the range of validity of the existing solution and yielded another approximate solution; from these, exact solutions for the limiting cases of only one γ are obtained. There is over-all agreement between theory and experiment within limits of error except for an observed delay in the initial current rise. To explain this requires the consideration of mechanisms, involving creation and transmission of active photons, which could cause a delay of the order of an electron crossing time.