100 Cady and Vinal — Electric Arc. 



of Voege* and Algermissen,t especially as the latter writer 

 shows that for short spark lengths the high-frequency dis- 

 charge potential differs but little from the static potential. 



The spark-gap used consisted of a brass plate A (fig. 4) l cm 

 in diameter, which served as anode, and as cathode a small 

 steel sewing needle B. These electrodes were mounted about 

 l mm apart in a small glass tube, and were sealed in with 

 sealing-wax. A side arm G from the tube communicated 

 with a hydrogen generator, air-pump, and manometer. The 

 copper arc in a mixture of hydrogen and illuminating-gas at 



Fig. 4. 



/ J ' V- 



/ 



atmospheric pressure was employed as the source of oscilla- 

 tions. Short, thick wires connected the arc terminals with the 

 spark-gap. 



Great difficulty was at first experienced in causing a spark 

 to pass promptly at sufficiently low potential. An iron arc, 

 and later a powerful spark between zinc electrodes, placed 

 close to the glass tube, did not materially help matters, nor did 

 ionizing the gas in the tube directly by running the discharge 

 from a small induction-coil from point to plate just before 

 each observation. 



It then occurred to us to keep the gas artificially in a state 

 of ionization while the sparking potential was being applied. 

 To this end two platinum wires D, E were sealed into the 

 tube on opposite sides of the spark-gap, about 5 mm in front of 

 the brass disc. The terminal A was kept constantly connected 

 to the positive terminals of the arc and of a variable e.m.f. 

 for calibrating. B could be connected in rapid succession to 

 the negative terminal of the calibrating e.m.f. (through a high 

 resistance), and to the cathode of the arc. Observations were 

 carried out thus : the gas in the tube was exhausted to a cer- 

 tain pressure, the auxiliary discharge from a small induction 

 coil started between D and E, and the lowest static potential 

 observed that just sufficed to start a discharge between 

 A and B. B was then quickly connected to the cathode of 

 the arc, and it was noted whether or not a discharge between 

 A and B took place. The discharge between D and E was 



* Voege, Elektrot. Zschr., xxv, p. 1033, 1904. 

 f Algermissen, Ann. Phys., xix, p. 1016, 1906. 



