REQUIRED TO PRODUCE DISCHARGE THROUGH VARIOUS GASES. 



383 



pass. This first spark was followed by a torrent of others, the telephone emitting a 

 chattering noise. The resistance column was gradually shortened until these sparks 

 ceased, and then cautiously lengthened until they just began again. 



At the moment when this happened the voltmeter was read. The voltage was 

 allowed to rise slowly enough to prevent any appreciable swing of the needle. 



When an observation had been taken, the voltage was again reduced below the 

 sparking value, and then again cautiously increased by lengthening the resistance 

 column. When the telephone began to click the reading was again noted. Ten 

 observations were usually taken at each pressure. They could be obtained in fairly 

 rapid succession. 



Experiments on Atmospheric Air. 



The first set of measurements was made on atmospheric air, not specially well dried. 

 The readings will be given in full, as an example, to show the degree of concord- 

 ance obtained. 



The minimum spark potential is thus 341 volts. It will be noticed that the larger 

 potential differences, i.e., those furthest on either side from the minimum, could not 

 be measured with the precision that was possible with those in the neighbourhood of 

 the minimum. As, however, the value of the minimum is the point of chief interest, 

 this is of small importance. 



The results are plotted on Diagram No. 1. The smoothness of the curve is some 

 guarantee of the accuracy of the observations. 



