ON ELECTRICAL DISCHARGES THROUGH RAREFIED GASES. 
219 
of tension at A is greater than that at B, since it compels B to give it relief. But 
however large B might be, such relief could never be greater than would amount 
to making the potential at A sink to that at B, which would by no means he tanta¬ 
mount to connecting A to earth, since the impulsive change at B will ordinarily 
be very considerable. Hence, as we get complete relief at A by joining it to B, 
it must be that B has not suffered any increase of potential at the moment that 
A seeks relief, and that thus it is not prevented from giving to A that relief which its 
size and distance from the spot where the free electricity is at the instant in question 
enable it to give—a relief which, as we have seen, is to all intents and purposes 
complete. But it does so at the expense of receiving a pulse of electricity from A of 
the same type as it would receive were it joined to the air-spark terminal, and 
we therefore get indications of non-relief-effect. We see then that the discharge 
is progressive, and that it proceeds from the air-spark terminal. For the electric 
disturbance at A makes its demand for relief on B, and finds B quite as capable 
of giving it as though no discharge had taken place, so that the discharge has not yet 
reached B. 
Another very instructive form of these experiments is obtained by placing a long 
strip of tinfoil along the tube from end to end, taking care, of course, that it is 
not connected with the terminals (Plate 20, fig. 25). Suppose the air-spark is in 
the positive, then, if the tube be sufficiently sensitive, it will be found that near 
the positive terminal there is the discharge relief-effect. A little further on there 
is repulsion, which gradually diminishes and is finally replaced by attraction of the 
luminous column which, as we have seen, is an ordinary form of non-relief-effect. * 
The portion of the tinfoil near the air-spark terminal has derived relief from the 
remainder by reason of its being the first to experience the sudden change of tension, 
and the more distant parts receive in consequence pulses of electricity sufficient 
to cause the non-relief-effects. If the tinfoil be divided a little beyond the part of the 
tube occupied by the discharge relief-effect the positive end of the further portion will 
now present the discharge relief-effect; because instead of being a part of the relieving 
system to the portion of the strip of tinfoil between it and the positive terminal, it is 
now the part of the further piece of tinfoil which is the first to feel the call for relief, 
and consequently is the part that derives the full benefit of the relieving system 
formed by the piece of tinfoil of which it forms a part. And at the same time 
the negative end of the nearer piece of tinfoil (which previously showed relief-effects) 
will show non-relief-effects, inasmuch as it has to afford relief to the part nearest 
to the positive terminal and has no longer any more distant piece of tinfoil from which 
to draw supplies of positive electricity. 
* When the air-spark is in the negative the relief-effect usually assumes the appearance of an attraction 
of the luminosity to the inner surface of the tube. In reality it is due to the luminous positive discharges 
from the side of the tube, such discharges always appearing (as we have seen) to start from the actual 
surface of the glass. The same remarks apply to the non-relief-effect when the air-spark*is in the positive. 
2 F 2 
