ON THE SENSITIVE STATE OF VACUUM DISCHARGES. 
571 
both directions, having been produced by the same interfering impulses, will be of the 
same character, and will exhibit the same features ; thus in the case described above, 
viz. : positive impulses, we have a double hollow cone at the tinfoil. But the effects 
are no longer pure, and would not be specially worth notice were it not that some of 
them resemble very strikingly some of the positive-special effects obtained in the way 
described in our last paper, with a very long air-spark. It is probable that this same 
explanation is applicable in both cases. In the present case the power that we possess 
of altering independently the interfering system without otherwise altering the current 
affected, enables us thoroughly to test the truth of our explanation. By slightly in¬ 
creasing the air-spark, the limit of the effect of the induction discharges may be made 
to'ad vance as slowly as we please towards the positive terminal; and in a well striated 
discharge the authors of this paper have succeeded in putting out one by one the 
striae lying between the tinfoil and the positive terminal. The portion of the dis¬ 
charge that remained striated was still continuous and non-sensitive, while that 
between the striated portion and the tinfoil (as well as that between the tinfoil and 
the negative terminal) became highly sensitive, showing positive interrnittence. 
If the wire from the tinfoil be connected to the positive terminal of the machine in 
the interfering system, we shall of course have negative discharges within the tube, 
and all the interrnittence will be of a negative type. With this arrangement we are 
able to produce the well known phenomenon of repulsion of the positive luminosity ; 
and, with proper adjustments, to get excellent examples of what we have called the 
ring-terminal form of negative effect, and of all the forms with which we are familiar 
in positive-relief or negative-special effects.* Again, as in the case of the positive air- 
spark, if the air-spark in the interfering system be long, so that the intervals between 
the impulses are very considerable, or if in any other way we deviate far from the 
conditions which give us the best effects, we are apt to find mixed results, belonging 
partly to the original discharge and partly to the interfering system. This fact makes 
us incline more strongly to accept the explanation given above of the ease and com¬ 
pleteness with which the typical positive effect can be obtained, and the strong- 
resemblance which the results of positive interference bear- to those of the positive- 
special through a great range of air-spark. The negative is very inferior to the 
positive in this respect, and naturally must be so, since the structure set up by the 
negative impulses, though suited to the needs of their own circumstances, is wholly 
unlike the structure that the discharge would shape for itself in order to facilitate its 
passage through the tube ; and hence there is a greater tendency for it to reassert its 
former shape and appear superimposed upon the visible results of the interference.! 
* See Phil. Trans., 1879, Plate 17, figs. 12, 13, 14. 
t This distinction between the positive and the negative effect of an interfering system is so marked 
that when the interfering system is of small quantity it becomes very difficult to recognise the effects pro¬ 
duced at the tinfoil by negative impulses. The best method of showing the effect of these negative 
impulses in causing interrnittence in the main discharge is to place the finger on the tube between the 
