ON THE SENSITIVE STATE OF VACUUM DISCHARGES. 
641 
driving back the column of luminosity until they were diverted. But we think that 
this is only in appearance, and not in reality; and that no comparison can be made 
between the portions of the luminosity in the two cases, as the discharge has been 
affected by the alteration. Moreover, the same effect is produced by a magnet on 
discharges in low exhausts where there can be no question of molecular streams pro¬ 
ceeding right down the tube. We think that the mistake arises from the effect upon 
the imagination produced by the view of the well-known negative dark space. This 
seems to separate the region of the negative terminal from that of the positive 
terminal, and anything that affects this dark space is thought to do so by directly 
affecting the action of the negative terminal. But in reality the region of the negative 
terminal is bounded by the negative glow which is the positive end of the physical 
unit of discharge, of which Crookes’ space is the blank-space and the negative 
terminal is the negative end. The negative dark space is only the blank-space of the 
second physical unit of discharge, and the apparent advance of the positive luminosity 
shows only that this second physical unit of discharge has been affected in some way 
by the altered circumstances of the discharge. And when we consider that its 
exceptionally long blauk-space is due to its peculiarities of situation and the fact that 
its gaseous negative terminal (i.e., the haze at the back of the negative glow) is so 
unlike the gaseous negative terminals of the other stria spaces, it does not appear 
strange that an alteration which produces a great modification in these matters should 
affect the length of this exceptionally extended blank-space. 
We now come to the question of the time required for the emission of the positive 
discharge. This we have, in effect, dealt with already, when we considered the ques¬ 
tion of the dispatch of the positive discharge in positive special before the arrival at 
the tinfoil of the discharge along the tube. It must be of an order of magnitude not 
superior to that of the time required for the discharge to pass along the tube, and is 
probably of a lower order. 
The time required for the passage of the electricity along the wire outside the tube 
is, as we have seen, so short that it cannot be detected by the aid of any of the other 
phenomena of the discharge. As it is a case of conduction, it is, of course, the same 
for both electricities. 
The order of the small time-quantities of the discharge is therefore as follows; the 
groups being arranged in descending order of magnitude ;— 
A. The interval between two discharges. 
B. The time occupied by the discharge of the negative electricity from its 
terminal. 
The time occupied by molecular streams in leaving a negative terminal. 
The time occupied by the particles composing molecular streams in passing 
along the tube. 
