ON THE SENSITIVE STATE OE VACUUM DISCHARGES. 
579 
resting upon it at a spot near to the non-air-spark terminal of the tube. A piece of 
tinfoil was then laid on the tube near the air-spark terminal and connected with the 
former piece' by a fine wire. The consequence of this is, as we know, that the piece 
near the air-spark terminal derives relief from the other piece ; and, in the experiment 
above referred to, it was shown that the efficacy of the system to give relief was 
greater as thus arranged than it would be if the glass rod with the tinfoil were moved 
into a position at right angles to the tube, so that the tinfoil would be as far as 
possible from the tube. The conclusion deduced from this was that the further end 
of the tube was in the act of charging up in the contrary sense to the air-spark pulse 
at the moment that the discharge occurred. Now if this be tried with the Leyden-jar 
arrangement (i.e. , with the earth connexion behind the air-spark) it will be found the 
effect is intensified, and that the tinfoil gives very much greater relief when upon the 
tube than in any other position ; showing that the further end of the tube is rapidly 
charging up at the time of the air-spark discharge. Thus if the air-spark is in the 
positive the relieving tinfoil will supply more negative when on the tube than other¬ 
wise ; in other words, the negative end of the tube must be receiving a rapidly growing 
negative charge. And this is exactly what must be the case if the explanations given 
above are correct. 
The other experiment is a very remarkable one. We take a tube of moderately 
high exhaust through which is passing a discharge of considerable quantity, with the 
air-spark in the negative, and place, as before, the earth connexion behind the air-spark 
by connecting the negative terminal of the machine to earth (Plate 26, fig. 7). The 
consequence of this arrangement is that we get a violent negative intermittence. If 
we place upon the tube a piece of tinfoil and connect it with earth we shall get the 
usual negative-relief effects, i.e., positive luminosity on the inner side of the tube 
beneath the tinfoil. But if we withdraw the earth connexion from the tinfoil, and 
hold it at long sparking-distance from it, bright phosphorescence will appear opposite 
the tinfoil. The explanation of this is that during the interval between two air-sparks 
the tube is rapidly charging-up with positive electricity. Rapid, however, as this 
charging-up is, it is not of a sufficiently impulsive type to give rise to a relief-discharge 
from the tinfoil sufficient to produce luminous effects. If, however, we hold a wire 
connected to earth at such a distance from the tinfoil that the accumulated effects of 
this charging-up are able at intervals (either once, twice, or even more frequently in 
each interval of the main air-spark) to draw a spark from the earth wire, we have 
corresponding to each such spark an impulsive relief-discharge of negative electricity 
from under the tinfoil which produces the phosphorescence observed. 
This Leyden-jar effect is found with all lengths of air-spark and in all kinds of tubes. 
When we come, however, to tubes of very high vacuum the whole becomes so compli¬ 
cated by other considerations that, though there are traces of the effect, it is so 
masked as to require special examination to detect it. In fact, thu tube represents in 
many cases a much greater resistance than the air-spark interval, so that the above 
