370 Prof. E. Wiedemann on the Behaviour of Gases 



A series of researches on the laws of the total heating of 

 oases at high pressures when traversed by the discharge of a 

 Leyden battery or of an induction-coil has recently been made 

 by Villari*. They are in complete agreement with the conse- 

 quences which follow from mechanical principles, as is the case 

 with the problems treated by me. When, for example, Villari 

 finds that the heating in sparks between two spheres is nearly 

 proportional to the spark-length, that follows immediately from 

 the law established by Macfarlane and others j, that the differ- 

 ence of potential necessary to the commencement of discharge 

 between two spheres inci eases nearly proportionally to their dis- 

 tance apart, at any rate for short sparks. The same results were 

 to be expected from the laws previously established connecting 

 striking-distance with quantity of electricity discharged. Vil- 

 lari, however, is not able in his experiments to distinguish 

 between the heating at the electrodes and in the spark itself, 

 which corresponds to the capillary portion of a Geissler's tube. 



II. Further, the amounts of heating produced in the tubes 

 connecting the electrodes ivere determined when a spark of 

 10 millims. was included in the circuit, and when no spark 

 was included. The width of the capillary tube (arrangement 

 of fig. 3) was 0*4 millim. 



The tables give the results obtained in a series of measure- 

 ments. 



Air. 



Hydrogen. 



p 



F=0 



h 

 F=10 



F=0 



F=10 



P 



^=0 



F=10 



F=0 



2^=10 



11-5 

 4-2 

 04 



2-04 



0-81 



1-7 



1-48 



1-42 



24 

 1-40 



0-82 



190 



1-46 

 1-21 



22 



12-2 

 44 

 0-6 



3-76 

 2-06 

 1-02 

 0-49 



3-l6a 

 1-5 

 1-3 

 1-16 



345 

 200 

 1-08 

 0-50 



32a 

 1-75 

 1-21 

 0-92 



a The spark measured only 5 millims. 



To these observations belong determinations of the number 

 of discharges. W'hen the spark-length was 10 millims., the 

 number reduced to a galvanometer-deflection 10 was always 

 about 12 to 15 a second, whatever the pressure in the tube. 

 The number decreased a little, it is true, with increase of 

 pressure, within the limits of observation. It is otherwise when 

 no spark is included in the circuit. 



The results are given in the tables. 



* Villari, Beibl. iii. p. 713, and iv. p. 404. 

 t Macfarlane, Beibl, iii. p. 429 (1879). 



