[ 677 ] 



LXII. The Effect of an Air-Blast upon the Spark Discharge oj 

 a Condenser charged by an Induction Coil or Transformer. 

 By J. A. Fleming, M.A., D.Sc, F.R.S., Professor of 

 Electrical Engineering in University College, London, and 

 H. W. Richardson, B.Sc* 



WHEN the oscillatory discharge of a condenser is caused 

 to take place across a spark-gap in the usual manner 

 by charging the condenser by means or" an induction-coil or 

 transformer, the intermittent spark which takes place be- 

 tween the spark-balls is a complex effect. It consists partly 

 of the true oscillatory discharge of the condenser and partly 

 of an electric arc, unidirectional or alternating, which is 

 superimposed on the true condenser oscillatory spark. If a 

 hot-wire ammeter or other means of measuring the effective 

 or mean-square value of the discharge current is inserted in 

 the condenser circuit, this current will generally be found 

 to be irregular, and if a radiative circuit is coupled to the 

 condenser circuit as in radiotelegraphy, the radiation from it 

 will be found to consist of trains of waves whose initial 

 amplitude is also variable. This irregularity is a source of 

 difficulty in making radiotelcgraphie or laboratory measure- 

 ments of current, decrement, wave-length, &c, when origi- 

 nated by condenser discharges. The reason is that the 

 moment the condenser begins to discharge, and the first 

 so-called pilot spark takes place between the balls, the 

 resistance of the spark-gap falls, and an arc discharge from 

 the induction-coil or transformer commences across the gap. 

 Until this arc is extinguished the condenser cannot again 

 become charged to any high voltage, and the voltage to which 

 it is charged will depend upon the state in which the ball 

 surfaces are left as regards temperature and smoothness, 

 since these are factors in determining the spark potential, 

 and also on the condition of the air-space as regards conduc- 

 tivity. Accordingly, to produce a uniform oscillatory dis- 

 charge this true arc-discharge must be either prevented or 

 arrested at once, and the spark between the balls should arise 

 wholly from energy which comes out of the condenser, and 

 not from energy coming directly from the transformer or 

 coil. When moderate power is being employed this arc- 

 discharge can be best annulled by a blast of air thrown on 

 the spark-gap. This has the effect of blowing away the arc, 

 but does not stop the condenser oscillatory discharge. For 

 a long time past the utility of this air-blast in connexion 



* Communicated bv the Physical Society: read March t!6, 1909. 



