Air-Blast upon the Spark Discharge of a Condenser. 691 



current almost as much as by subjecting the single gap or! 

 0*6 mm. to the air-blast. This increase appears to be due to 

 the suppression o£ the arc-discharge, as shown by the ap- 

 pearance o£ the images of the spark in the revolving mirror, 

 and hence the arcing is suppressed by the separation o£ the 

 gap into multiple gaps, almost as effectively as by blowing 

 upon the single gap. On the other hand, separating the 

 spark-gap into more than five separate gaps seems to result in 

 diminishing the total discharge current beyond a certain very 

 short length of spark. The figures in the table, however, do 

 not show what is observed in practice, namely, the greater 

 steadiness of the discharge current under the operation o£ 

 the air-blast; and the conclusion, therefore, is that in any 

 experiments in which great constancy is required in the 

 discharge current in the condenser circuit, a great ad- 

 vantage is obtained by using a short spark-gap, and by 

 subjecting the discharge spark to an air-blast, as this 

 both increases the charging voltage of the condenser and 

 steadies the discharge current by abolishing the arc, which 

 would otherwise take place even with an ordinary induction- 

 coil. 



A similar set of experiments was tried with a larger alter- 

 nating current plant and a high tension transformer, consti- 

 tuting one of the wireless telegraph sets of the Radiotelegraphic 

 Laboratory, University College. The alternating current 

 was supplied from a 5-k.w. alternator and raised in pressure 

 by a transformer discharging across a spark-gap between 

 two iron balls, the spark-gap being shunted by a condenser 

 in series with the primary circuit of an oscillation transformer, 

 the secondary circuit of which was inserted between an 

 antenna and an earth connexion, the antenna circuit being 

 tuned to the condenser circuit. As no suitable hot-wire 

 ammeter w T as available for measuring the large current in 

 the condenser circuit, readings were taken by connecting 

 the terminals of a hot-wire voltmeter to points on the earth- 

 wire connexion of the antenna, consisting of a copper strip. 

 These points being selected a few inches apart so as to give 

 a convenient reading on the low reading hot-wire voltmeter. 

 These readings, however, are approximately proportional to 

 the currents flowing in the condenser circuit. Spark-gaps 

 were then adjusted from 0*5 mm. to 3 mm. in length, and 

 the reading of the voltmeter taken, both when the spark-gap 

 was blown upon and when it was not blown upon a by a jet 

 of air. 



3 A 2 



