398 PROCEEDINGS OF THE AMERICAN ACADEMY. 



familiar photographs taken in the same way with the discharge in air 

 between cadmium terminals. The two terminals become successively 

 luminous in an oscillatory fashion. 



In Figures 1, 3, and 4 of Plate I light bright enough to affect the 

 photographic plate is obtained only from one of the electrodes, the cath- 

 ode. The exposure is too short to show what happens at the anode. 

 By the use of a larger capacity and a larger inductance the period of 

 oscillation can be increased so that the exposure is long enough to get 

 a faint impression from the anode. This was done in a case where 

 the capacity was .117 microfarads, the inductance .0071 henries, and 

 the time of an oscillation, therefore, .000178 second. The picture ob- 

 tained, which is not here reproduced, showed that, during the exist- 

 ance of the bright point of light at the cathode, there is also a weak 

 glow spread over the entire surface of the anode. 



These results indicate that the current whose action is here photo- 

 graphed is exactly of the same character as the mercury arc, except 

 that the current is reversed several times during the condenser dis- 

 charge ; for in the mercury arc with direct current the anode is covered 

 with a faint glow over its entire surface, while the cathode region is 

 dark except for a very small spot of light of extreme brilliancy. 



The fact that the several pictures of Plate I show different frequencies 

 and different damping is of no significance, as they were made with 

 various inductances, capacities, and resistances in the circuit of discharge. 



One important fact about the pictures obtained with the mercury 

 interrupter is that the impressions are so sharp (in the negative) that 

 their distance apart (cf. Figure 1, or the distance from the first to the 

 last on the left-hand electrode of Figure 3) can be measured with great 

 accuracy, which makes this form of interrupter useful in the photographic 

 measurement of the time of a condenser discharge. For example, taking 

 the stroboscopic determination of the speed of the mirror and measur- 

 ing the distances between impressions on various photographic plates, 

 the values given in Table II were obtained for the time in seconds of 

 a double oscillation. 



The measurements in the last column were made with a cadmium spark- 

 gap instead of the mercury interrupter, and show larger variations than 

 the columns obtained with the interrupter. 



Measurements similar to those of Table II have been utilized in the 

 determination of the capacity of the condensers and the inductance of the 

 leads, — quantities that are needed in the discussions of section VI of 

 this paper. The determination of these quantities was made as follows. 



