ON THE SENSITIVE STATE OF VACUUM DISCHARGES. 
GO 9 
or turn aside those proceeding from the other, for just over each piece of tinfoil there 
is a patch of glass, of about the same shape as the tinfoil though rather larger than it 
(inasmuch as it overlaps it a little on all sides), which is wholly devoid of phosphor¬ 
escence. The best example of this is got by putting a ring of tinfoil on the tube and 
connecting it with earth. No phosphorescence appears beneath the tinfoil, showing 
that no streams proceeding normally from the surface of the glass have reached the 
opposite side. All the phosphorescence is arranged in two rings, one on each side of 
the ring of tinfoil and parallel to it but separated from it by a space whose breadth is 
pretty uniform and varies in different cases from one-eightli to half an inch. Here it 
would seem that the streams that started normally, or nearly so, have interfered with 
ODe another with the result either of neutralising one another or deviating one another 
from the normal course and causing incidence on the glass at some distance on one 
side or the other of the tinfoil ring. 
This capability of interfering with one another possessed by these molecular streams 
is one of great importance, both as giving us an insight into their nature, and also as 
an assistance in examining the mechanism of vacuum discharges. But we shall not 
dwell further upon it now as it belongs more properly to the next section, and is only 
mentioned here incidentally as one of the causes at work in determining the shape of 
the relief-phosphorescence. 
The third cause which is probably at work to distort the phosphorescent image of 
the tinfoil is the influence of the position of the exciting positive electricity upon the 
negative discharge that responds to it. It has been thought that the direction of 
these molecular streams depends solely on the shape of the negative terminal, and is 
wholly independent of the position of the positive terminal, that is to say, of the 
direction from which the demand for negative electricity corues. This may be approxi¬ 
mately so in the comparatively mild action that accompanies the continuous current 
(which may be compared to a case of steady motion in dynamics), but it certainly is 
not so in the more violent actions which accompany the impulsive discharges of the 
intermittent current. But, just as in the former case of the interference of molecular 
streams during their flight, the examination of this point belongs more properly to 
another branch of the investigation, so that we shall not notice it here at any length. 
Leaving the question of the causes which determine the direction of these molecular 
streams, there is no doubt of their identity of nature with those that accompany the 
continuous current. They cause shadows in the same manner as do those proceeding 
from the negative terminal. In this way we can obtain shadows of any loose object 
in the tube and even of the positive terminal itself. The negative terminal, however, 
does not ordinarily cast a shadow properly so called, inasmuch as it is itself giving off 
like streams, and thus the streams that proceed from it turn aside any other streams 
that would otherwise impinge on it. These shadows due to relief are wholly indepen¬ 
dent of those that are due to the discharge from the negative terminal. If the object 
casting the shadow be sufficiently near the negative terminal to cast a shadow in the 
