ON THE SENSITIVE STATE OF VACUUM DISCHARGES. 
635 
cumstances it is clearly impossible that the revocation of the relief could be effective 
in stopping them, and preventing them from producing phosphorescence, seeing that 
they consist of material particles moving at high velocities; leaving out of considera¬ 
tion the fact that in all probability the largest portion of them would have actually 
arrived at the opposite side of the tube before the revocation arrived. 
It must not be supposed that this theory requires that absolutely the whole of the 
negative discharge at the first piece of tinfoil (i.e., the one that is nearest to the 
positive terminal) should be prevented by the revocation. This would not be in 
accordance with what the luminous phenomena would lead us to expect, and is not 
countenanced by the theory itself. There has been a discharge of positive electricity 
from beneath the further piece of tinfoil, and this has left a quantity of negative 
electricity free on the surface of the glass there. This in itself must render the 
impulse of the revocation less than that of the original demand. But the two are 
sufficiently nearly equal to prevent the negative discharge at the first piece of tinfoil 
having the violent character which would be necessary to produce relief-phospho¬ 
rescence and to cause the negative discharge there to become a differential effect, and 
pass off in a gentle and continuous manner. 
The existence of this revocation and its efficacy in preventing the emission of mole¬ 
cular streams in the ordinary way is further shown by there being no virtual shadow 
formed at the first piece of tinfoil. Now the absence of a virtual shadow is a very 
much sharper test of the absence of molecular streams than the non-appearance oi 
relief-phosphorescence on the opposite side of the tube, for in the one case the mole¬ 
cular streams produce their effect close to the spot whence they proceed, while in the 
other case they have to force their way across the tube and must then possess sufficient 
velocity to produce phosphorescence. In order to show the effect of the revocation on 
the production of a virtual shadow, one end of a narrow slip of tinfoil a few inches in 
length was cemented to the side of the tube (through which was passing a discharge 
with positive intermittence) a little way from the positive terminal, and the slip was 
allowed to hang downward. A clear virtual shadow appeared, starting from the point 
where the tinfoil was in contact with the tube. The strip of tinfoil was then laid 
along the tube, stretching towards the positive terminal, and the virtual shadow 
disappeared. Now we know that, so far as the effects pn luminosity at the cemented 
end were concerned, the two positions must have been practically equivalent. But 
in the one case there was a revocation which, though it did not come in time to stop 
the effects of the relieving system upon luminosity, was yet in time to prevent the 
emission of molecular streams so that no virtual shadow appeared. This experiment, 
though very interesting, is not so conclusive of the truth of pur theory as is the one 
first given, for it might be objected that the virtual shadows could not be formed 
until the discharge in the tube had reached the negative terminal and obtained a 
response in the form of negative discharge, and a sufficient time had then elapsed 
to permit the molecular streams which accompanied that response to arrive at the first 
