Electric Discharge in Gases. 45 



and anode attract each other with great force. If they have 

 been made to touch, and then we try to separate them, sparks 

 leap across and the kathode vibrates to and fro. 



If the machine works very slowly, then, as soon as the blue 

 negative layer has been traversed, irregular discharges set in, 

 the green light of which is seen at different points of the tube, 

 just as with a tube which has been exhausted so far that the 

 discharges traverse the gas only with difficulty. 



A conclusion as to these last phenomena is furnished by 

 experiments in which the poles of the machine are brought 

 within a small distance, some 3 or 5 millim., of each other, 

 and then the lower electrode is allowed to approach the other ; 

 as soon as the two electrodes are sufficiently near, a stream of 

 sparks passes between the poles of the machine, whilst the 

 tube remains dark. If, then, the distance of the electrodes is 

 increased, the discharge in the tube suddenly commences, 

 whilst no electricity passes between the poles of the machine. 

 If, now, besides the experimental tube, we include also the 

 comparison-tube in the circuit, then, when sparks are passing 

 between the poles of the machine, we see the walls of the 

 tube shine with bright green phosphorescence at the points 

 where positive light appears in the ordinary passage of the 

 discharge. The reason of this is, that the small distance 

 between the plates of the principal discharge-tube causes this 

 to act as a condenser ; they become gradually charged until 

 suddenly the discharge between the poles of the machine 

 commences, and the whole quantity of (negative) electricity, 

 which is accumulated on the one electrode of the tube, dis- 

 charges itself suddenly backwards through the small tube, 

 and thereby produces the green light. 



It follows from the experiments described that, in highly 

 exhausted spaces, the discharge between a positive and a nega- 

 tive electrode, if their distance apart is less than a certain 

 limit, takes place with greater difficulty the nearer they are 

 to each other. This result was confirmed and completed by the 

 following experiment : — 



Two discharge-tubes (figs. 4 and 4b) were connected at the 

 same time with the pump, which in this case was a Geissler's 

 pump, instead of the Topler's pump usually employed. The 

 tubes were each 26 millim. wide and 39 millim. long ; electrodes 

 of aluminium wire were melted into each, which were covered 

 with glass except about 1 millim. In the tube a the electrodes 

 were 1 millim., in the other tube b, 20 millim. apart ; the tubes 

 were included in parallel arc in the circuit of the machine. At 

 high pressures the discharge took place only in a ; at low pres- 

 sures of about 5 millim. and less, in both tubes at the same 



