570 
MESSRS. W. SPOTTISWOODE AND J. FLETCHER MOULTON 
carried by bursts of positive electricity, while that in the latter portion is left to go on 
its even course (Plate 25, fig. 2). 
We are brought to the same conclusion if we observe the phenomena produced by 
placing in circuit with the affected tube two other tubes, one at each end. We then find 
that the one at the negative end contains a sensitive and the one at the positive 
end a non-sensitive discharge."' To render the proof quite complete, two pieces of 
tinfoil were placed on the tube, and were connected, the one to the positive terminal, 
the other to the negative terminal of the interfering system. When both were so 
connected it was found that, although they each produced the accustomed effect in the 
tube, the discharge was not sensitive in the two auxiliary tubes in circuit, but that, if 
one of the pieces of tinfoil was disconnected from its terminal, sensitiveness appeared 
in the corresponding one of the tubes in circuit. Thus, when the negative terminal of 
the interfering system was alone connected with tinfoil on the affected tube, so as to 
produce positive discharges within the affected tube, it was found that the discharge in 
the tube at the negative end was sensitive, but this sensitiveness disappeared when 
the positive terminal of the relieving system was also put in connexion with a piece of 
tinfoil on the affected tube. The sudden in-rushes of positive electricity from under 
the one piece of tinfoil were neutralised (so far as the remaining portion of the circuit 
was concerned) by the synchronous in-rushes of negative electricity from under the 
other piece of tinfoil, and thus the discharge in the two tubes in circuit with the 
affected tube remained sensibly uninterrupted. 
It is, however, a necessary condition for these latter phenomena that the interference 
should not be too violent. If the air-spark in the relieving system is increased beyond 
a certain point, the discharges induced in the affected tube (which we will still assume 
to be of positive electricity) become so large and violent that they are more than 
sufficient to satisfy the negative electricity coming from the negative terminal; they 
consequently no longer go only towards the negative terminal but spread out both 
ways, and can even be made to pass out at both terminals of the tube. The discharge 
is then sensitive throughout the whole length of the tube ; and the sensitiveness in 
* This experiment is very nsefnl in bringing into prominence the essential difference that exists 
between the passage of electricity through conductors and its passage through gas. It shows ns that 
a current which in some portion of its circuit has to pass through gas can be rendered intermittent in one 
part of its course while it remains unaffected and continuous, or approximately so, in the other. This is 
radically different from Ohm’s law and the general theory of currents through conductors, for these 
require that at any instant of time the quantity flowing across each section of the conducting circuit 
should be the same. It is of course true that, whether a portion of the circuit be gaseous or not, the 
average quantity flowing across each section must be the same, but the equality is only in the average 
when taken over a finite and appreciable period of time, and no longer exists at each moment. The tube 
acts precisely as the air-vessel of a fire-engine. All the electricity that comes into it passes out again, but 
no longer with the same pulsations. The tube sometimes contains more and sometimes less free electricity, 
and acts as an elastic or expansible vessel would act if it formed part of the path of a stream of incom¬ 
pressible fluid. 
