ON ELECTRICAL DISCHARGES THROUGH RAREFIED GASES. 
193 
the positive portion of the external circuit: that is to say, between the positive terminal 
of the tube and the source of electricity, and of this case we have only attempted to 
interpret one of the two classes of effects. But, as we have already remarked, it is 
not only in this case that we get sensitiveness, or that the relief and non-relief-effects 
in the luminous discharge are observable. All these phenomena are similarly manifested 
when the air-spark or cause of intermittence is situated in the negative portion of the 
external circuit, although, as we shall see, the characteristics of the two effects are 
quite changed. We shall now proceed to examine the appearances in the four cases, 
via.: the relief and non-relief-effects when the air-spark is in the positive and the 
negative respectively, and to trace the connexions and the resemblances which exist 
between them. 
To do this satisfactorily it is necessary to examine the various forms which these 
effects assume as we vary the intensity of the causes which give rise to them con¬ 
tinuously from zero up to the highest limit practically attainable with the instruments 
used. This variation may, as we have seen, be made in several different ways, of 
which the following are the most convenient :— 
(1.) We may vary the air-spark from zero up to the limit of striking distance ; or if 
the interruptedness of the current be arrived at by other means, some 
equivalent mode of altering the character of such interruptedness may be 
adopted. 
(2.) We may vary the distance of the tinfoil which produces the effect upon the 
tube from the greatest distance at which an effect is observable to actual 
contact. 
(3.) We ma}' move the tinfoil up and down the tube so as to be at a greater or less 
distance from the air-spark terminal. 
Supposing, then, the tinfoil to be in the form of a narrow ring surrounding the tube, 
and taking first the non-relief-effect with the air-spark in the positive, we obtain the 
following results by these respective methods:— 
(1.) As we increase the air-spark from its initial value, zero, there appears first of all a 
brightening of the central portion of the positive column in the part of the tube 
close to the edge of the tinfoil on the same side of it as the positive terminal 
of the tube. As the air-spark gets bigger this becomes a more or less marked 
division of the column into two parts, the details at the place of division 
shadowing forth the effect we have already described of the hollow cone and 
bright termination of the positive column. As the air-spark is still further 
increased this effect grows clearer and clearer, till it comes to its greatest 
perfection. Increasing the air-spark still more, the terminal stria or head of 
the column gets wider, or more nearly of the same diameter as the tube, and 
the brightness is greatest round its edges, so that it in reality becomes 
annular. It then grows shadowy and gets very close to the tinfoil, and 
finally disappears or becomes from its position so difficult to observe that its 
MDCCCLXXIX. 2 C 
