under the Influence of the Electric Discharge. 



367 



much as a decrease of the quantity of heat produced must 

 result, the results found for the number of discharges and the 

 quantity of heat produced are in agreement. 



If the quantity of electricity e of potential V sinks to poten- 

 tial 0, the quantity of heat produced is proportional to eY. If, 

 instead of one discharge of quantity e, n such discharges occur 



. e . Y 



each of quantity — , then these sink from potential — to poten- 

 tial 0. The quantities of heat produced are proportional to 



V e 1 '■■•.". 1 



n. — - = - Ye, and therefore only - as great as in the first 

 n n n J n 



case. 



Warren De La Rue and Hugo Miiller* have also found a 

 minimum value for the potential necessary to discharge with 

 decreasing pressure, but only with air. The same conclusion 

 may be drawn from the experiments of Morren and De La Rive f 

 and others. Thus, for example, De La Rive found a maximum 

 for the intensity of the induced current when the discharges 

 from an inductorium were passed through a gas at continually 

 decreasing pressure, and hence concluded that the resistance 

 was a minimum. In the same way we may conclude there is 

 a minimum of potential necessary to discharge-^-since the less 

 electricity discharges itself backwards through the coil, and 

 consequently the more passes through the gas, the smaller does 

 the potential at the ends of the secondary coil become. 



An exactly similar series of experiments was made with 

 electrodes at a distance of only 1-J millim. from each other in 

 air. The water- equivalent was 28 gr. The following values 

 were obtained: — 



V 



+ 



- 



673 



017 



0-28 



63 



0090 



0083 



1-7 



0-107 



0071 



0-6 



0-13 



0-13 



X 



0-94 



0-80 



These numbers show the same result as the first series: 

 the heating at first decreases sloivly, and then increases very 

 rapidly as the pressure diminishes. 



The number of discharges was, for p = 0, 90; for p = 63 

 they were not to be counted; for p = 7QQ, about 250; so that 

 here again we have a maximum. 



* Warren De La Rue and Hugo Miiller, Proc. Rov. Soc. xxix. p. 281 

 (1879). 

 t Morren and De La Rive, Wied. Galv. [2] ii. p. 316 &c. 



2D2 



