0N THE SENSITIVE STATE OF VACUUM DISCHARGES. 
649 
positive discharges from that part of the tube inwards. Not only can it do this, but 
it certainly does so, for the effect of the impulsive variations of potential in the wire 
must of necessity produce such discharges. Hence we see that the effect of producing 
a locus of such discharges in the tube is to cause the original locus to shift till it 
coincides with the new locus. Thus the creation of the new locus has made the first 
unnecessary, or, in other words, the discharge does not require any special series of 
positive centres of discharge along its course, but it is content with one series ; but 
this it must have. 
Now if we compare this with the phenomenon of positive-special-effect when a ring 
of tinfoil is used, we shall gain a valuable insight into the mode of propagation of the 
discharge. In that case we know that the inductive discharge at the tinfoil takes the 
place of the original discharge, and the latter is satisfied by the negative that is left 
behind by the former. If we suppose this effect to be more imperfectly produced and 
spread along a line, passing along the tube longitudinally, instead of surrounding it, 
we shall get an idea of how the discharge is effected when we use the special-effect in 
the way described above. And the knowledge that we now have of the identity of 
the phenomenon presented by this special effect, and the ordinary discharge (save so 
far as regards the side of the tube which the luminous column prefers) in high vacuum 
tubes, seems to point to something like the above mode of propagation of the discharge 
in all cases. 
XXIX.— General conclusions as to the electric discharge. 
IV. Molecular streams. 
There are a few questions relating to molecular streams which we are in a better 
position now to consider than we were before the phenomena of phosphorescence in 
the sensitive current had been examined by us. These we shall shortly indicate. 
In Section XVII. we stated that it was our belief that there was no essential dif¬ 
ference between the molecular streams of particles of gas which go to produce phos¬ 
phorescence and the phenomenon of the driving off from the negative terminal of small 
loose particles of conducting matter, which also occurs in rarefied media. In order to 
settle this point, we took a tube of fairly good exhaust containing a little of a mixture 
of sand and lamp-black, the sand being put there to assist in removing the lamp-black 
from the sides of the tube in case it should adhere thereto. The tube was placed ver¬ 
tically with the negative end downwards, and a current from the large 12-plate Holtz 
machine was passed through it. In a few seconds the sides of the tube were covered 
with a coating of lamp-black for about two-thirds of its length. The experiment was 
then varied by the introduction of an air-spark into the circuit. Whether this was 
placed in the positive or negative portion of the circuit, the effect was the same, the 
lamp-black was driven with such violence against the sides of the tube that it became 
caked in some places, so that it was a troublesome matter to get it off. 
Seeing, then, that these particles of lamp-black behaved in the same way as particles 
