576 
MESSRS. W. SPOTTISWOODE AND J. FLETCHER MOULTON 
more about the nature of the intermittence, for we have no longer the power of 
detecting slight changes in the effects produced in the standard-tube as is the case 
when the sharply defined positive effects are under examination. 
It is obvious that if the standard-tube be at a considerable distance from the tube 
that is to be tested, the relieving system, composed of the tinfoil and the wire, will 
have considerable capacity, and thus the impulse on the tinfoil will lose proportionally 
in intensity. It is necessary, therefore, that they should not be too far apart. But 
there is a certain difficulty in placing them near together, for the tube containing the 
intermittent discharge acts by induction on the discharge in the standard-tube and 
makes it behave as though it were intermittent, even when there is no metallic 
connexion between the surfaces of the two tubes. But this is not of very serious 
importance, for the effect is of the same nature as that produced by the tinfoil, and 
differs from it only in being feebler and more diffused. 
It is often useful to place a second ring of tinfoil round the standard-tube uncon¬ 
nected with the tube that is being examined (Plate 26, fig. 5). By touching this we 
shall discover the nature of the intermittence produced in the standard-tube by the 
inductive discharges which are caused (either through the medium of the tinfoil, 
or directly as mentioned in the last paragraph) by the influence of the tube under 
examination. If the intermittence in the standard-tube be of a positive type, then it 
is clear that such also was the original intermittence, and vice versd. In cases where 
the electrical disturbances are very violent this is a useful addition to the tests already 
described, as the results it gives are of a milder type than those produced at the ring 
of tinfoil which is in direct connexion with the other tube. 
This method is not confined to cases in which the electric pulsations in the system 
under examination are actually in the form of discharges. It enables us to contrast 
any systems in which there are electric variations of a suitable type. Thus, for 
instance, we can by its aid compare the relative intensity of the disturbances at the 
terminals of the tube, at a point on its surface and at an intermediate terminal not in 
connexion with any portion of the external circuit. The effect at the air-spark 
terminal is always by far the greatest. But no general rule holds as to the others, 
save that there is but little difference between the intensity of the disturbance at an 
intermediate terminal (wherever it be situated) and that upon the surface of the tube. 
Nor is there, as a rule, much difference between the disturbance at the non-air-spark 
terminal and an intermediate terminal. On the contrary, it would seem that tube and 
terminals are all on much of an equality so soon as the electricity has once launched 
itself into the tube from the air-spark terminal. 
XVI .—On the Ley den-jar effect of vacuum tubes. 
We are about to touch on a subject which merits a much more complete inves¬ 
tigation than we have as yet given it, not only because the influence of this property 
