Search for a command to run...
The failure of W. H. Bennett working with very fine negative points and of G. G. Hudson working with larger carefully polished negative Pt points in clean dry dust-free air to observe the regular relaxation oscillator-like pulses reported by Trichel for negative points in room air led to a further study of the Trichel pulses. Certain results obtained by A. F. Kip in an earlier study of the Trichel pulses in room air have been added to the findings of Bennett and Hudson and are here reported. It was found by Kip and Bennett independently that the onset potentials and the current voltage characteristics of the corona showing Trichel pulses are independent of the metal and thus of the work function of the point. Bennett's studies indicate that the mechanism of the pulses depends entirely on the nature of the field about the point in air and that the pulses occur only in gases attaching electrons to form a space charge cloud of slowly moving negative ions whose time of movement across a critical region in the gap determines the frequency of the pulses. The absence of Trichel pulses in gases where electrons do not attach to form negative ions has independently been confirmed by G. L. Weissler in pure ${\mathrm{N}}_{2}$, ${\mathrm{H}}_{2}$ and A in studies to be published separately. It is further independently shown by Bennett and by Hudson that the regular Trichel pulses originating as they do in the gas require the presence of a source of triggering electrons. These can be furnished from fine points or by roughness on larger points through field emission, by very fine dust specks through the Paetow effect and probably by adequate photoelectric or thermionic emission from the point. Thus room air, yielding negative ions and providing ample numbers of fine dust specks for triggering, yields the regular Trichel pulses while clean, dry dust-free air, giving but rare dust specks, yields random bursts of pulses of irregular form. Observations of pre-onset phenomena by Kip and of the nature of the "active spots" and their action on the Pt point by Hudson are also reported. The theory of the phenomenon is reconsidered in the light of the findings and it is shown that the theory proposed by Trichel is applicable except as modified by the influence of the negative ion space charge and the necessity of triggering electrons. The appearance of what seems to be an incipient retrograde streamer process of the type pictured by Loeb and Meek in their study of long sparks is reported in the observations by Kip and Hudson of negative coronas from larger points well above onset.