222 
MESSRS. W. SPOTTISWOODE AND J. FLETCHER MOULTOH 
directly operates in producing the effects observed, we are brought face to face with 
the question, How far are our conclusions relative to these discharges applicable to the 
case of ordinary or uninterrupted vacuum discharges ? or in other words, How far 
they are independent of the peculiarity of the interruptedness of the discharge ? In 
considering this most important question, it must be conceded that all due weight must 
be given to the consideration that our investigations have been carried on by means of 
experiments which would, one and all, have failed had they been applied to ordinary 
vacuum discharges ; so that any conclusions sought to be drawn from them relative to 
such ordinary discharges must be received with great caution. But this would not 
justify us in neglecting the remarkable resemblances that exist between the two 
classes of discharges, or in refusing to examine the information which our knowledge 
of the constitution of the one gives us of the probable constitution of the other. 
As we have previously remarked, it may be stated generally that there is no 
phenomenon presented by ordinary vacuum discharges which may not be shown by 
sensitive or interrupted discharges; and conversely there is no phenomenon presented 
by the latter (except such as directly depend on and are due to its sensitive and 
interrupted character) which is not also shown by the former. In illustration of this 
we will select two important features, viz.: striation and the terminal peculiarities. 
And first as to striation : Every one of the manifold shapes presented by the strife of 
the one can be found among strife produced by the other; and, moreover, the very 
same element which predisposes to the presence or absence of striation (viz. : the 
greater or less quantity of electricity discharged) in the one case predisposes also in 
the other. No doubt the sensitive discharge, in the conditions under which it has 
been here studied, is much less often striated than is the non-sensitive, but as the 
character of such a discharge is essentially variable and unsteady, it is not at all 
surprising that effects which are known to depend on complete steadiness of the 
discharge should be those least frequently met with. And that we are right in 
assigning these as reasons for the tendency in interrupted discharges to be non-striated 
is rendered more probable by the fact that, when we compel the discharges to operate 
with great regularity and within the prescribed limits of quantity and of time, as for 
example in the case of the high-speed break or of artificially produced striation, we at 
once find it easy to produce the steady effects commonly observable in ordinary 
discharges. 
Next as to the terminal peculiarities.'" These are, it will be admitted, common to 
both classes of discharges. But inasmuch as we have seen that the relief and non¬ 
relief-effects are substantially identical with terminal peculiarities, we can now go 
further, and say that they are shown wherever there is an instantaneous electric 
discharge from the space close to the surface of any fixed body. The consideration, 
* If the theory of the authors of this paper be correct, these terminal peculiarities are in reality cases of 
striation where the circumstances of the discharge are such as to maintain perforce the requisite steadiness 
and uniformity of conditions. But for the present purposes it is better to consider them separately. 
