Time-Lag in the Spark Discharge. 219 



for a definite process which must be completed before the 

 -spark can pass. Some recent experiments seem to throw 

 further light on the matter; and though like all experiments 

 undertaken primarily for practical objects, they do not serve 

 to elucidate the matter completely, they are worth recording 

 if they give any information on a matter of some pure 

 scientific interest. 



Effect of -preliminary ionization. 



If the action of the third point in the 3-point gap is due 

 simply to the preliminary ionization of the main gap which 

 it causes, then any other source of ionization should produce 

 the same effect. A gap of 5 mm. between pointed rods 

 2*5 mm. in diameter was found to have a steady sparking 

 potential of 4800 volts. It was connected to a magneto, 

 giving a potential of which the frequency was about 5000 

 per sec. and the magnitude could be altered without change 

 of wave-form by altering the strength of the magnet. Regular 

 sparking was obtained with a peak potential of 11,000 volts, 

 while a spark sometimes passed when the peak potential was 

 as low as 6600 volts. The addition of the third point, 

 suitably adjusted, reduced the peak potential for regular 

 sparking with the same magneto to 7200 volts, while the 

 steady sparking potential and the peak potential which would 

 sometimes give a spark remained unchanged. The third 

 point was then removed and a tube containing a few milli- 

 grams of radium brought near the gap ; the peak potential 

 for regular sparking was reduced to 8100 volts, while that 

 which sometimes gave a spark and the steady sparking 

 potential again were unaltered. The action of the radium 

 is then precisely similar in nature to that of the pilot spark, 

 though it was rather smaller, doubtless because the ionization 

 produced by the radium was too small. 



It would be expected again that anything which would 

 decrease the number of casual ions present would increase the 

 peak potential required for regular sparking. Here again the 

 expectation was fulfilled. If a pair of plates charged to a 

 potential difference of 10,000 volts were placed on either 

 side of the gap and at a distance of 2 cm. from the electrodes, 

 regular sparking could not be obtained with the highest 

 peak potential which the magneto would give, 13,200 volts; 

 but again a spark sometimes passed at 6600 volts. Again, 

 if casual ions can thus be removed by a field between 

 external electrodes, they ought also to be removed by the 

 field between the sparking electrodes. It is to be expected 



Q2 



