ON" ELECTRICAL DISCHARGES THROUGH RAREFIED GASES. 
229 
affected by one another, as could be clearly shown by putting the one tube at a 
greater or less distance from the other. On examining the nature of these effects, 
they seemed to follow the same law as would have prevailed had the tubes formed 
one long one which could be doubled up or bent so as to make the one portion of 
it parallel to the other portion. This law is that the discharge in the portion of 
the tube nearer the air-spark tends to produce special effects on the further portion, 
and itself to manifest relief-effects One remarkable result of this law deserves 
mention. When the air-spark is in the positive the special effect is of course accom¬ 
panied by positive luminous discharge from the inside of the glass. Accordingly, the 
tube through which the discharge passed last had positive luminous discharge 
throughout its whole length, on the side where the other tube lay, and thus the 
whole luminosity in that tube appeared to be attracted towards the other discharge. 
This was of course the case whether the discharge passed in the same or contrary 
directions through the tubes, so that the remarkable phenomenon was presented of 
a luminous discharge (which we know behaves like a current) appearing to attract 
violently a discharge going in the opposite direction. After the explanation given 
above, it will be understood that the phenomenon was not really a case of two 
currents going in opposite directions attracting one another, but was a case of the 
luminous effects of the interference of the two co-periodic discharges through the 
medium of the impulsive static induction of the pulses of which they were constituted. 
When a discharge is intermittent it can be made to pass through several tubes 
which are so connected as to offer alternative paths to the discharge. When this was 
done it was found that these discharges affected each other either by apparent 
attraction or repulsion of the luminosity which accompanied them. The authors 
have not had time to go further into the subject so as to ascertain precisely the 
law in such cases, but they have no doubt that on examination the effects could be 
accounted for by similar considerations to those which were found to hold in the 
previous case. 
An extremely interesting experiment showing how intermittent discharges can in 
this way be made to interfere when they are synchronous, but not otherwise, was 
made by using two equal small coils to produce the discharges in two parallel tubes. 
When the contact breakers were made to work in unison any effect which the dis¬ 
charges produced upon one another was steady, as would have been the case had 
their co-periodicity been due to their forming part of one discharge as in the previous 
cases. But when they were a little out of unison, so that audible beats were given 
hi addition to the two notes produced by the contact breakers, it was found that 
the luminosities in the tubes flickered (i.e., were attracted or repelled) synchronously 
with each such beat. The reason is obvious. It was only just at the time that 
the contact breakers were moving approximately synchronously that the two inter¬ 
mittent discharges were sufficiently synchronous to interfere with one another. 
