586 
MESSRS. W. SPOTTISWOODE AND J. FLETCHER MOULTON 
These experiments, which could be multiplied to any extent, show that phospho¬ 
rescence can be produced in tubes of all degrees of exhaust by sufficiently increasing 
the violence of the negative discharge. But there is an experimental fact which has 
been repeatedly observed and should be mentioned here, which show 7 s that it is not 
essential that there should be any increased violence of the whole discharge, but that 
a sufficient intensification of the local action at the negative terminal is all that is 
necessary. When a tube is being exhausted to a high vacuum the phosphorescence 
always appears first in tire neigbourhood of the negative terminal. But it often 
happens that one of the terminals of the tube is much smaller than the other. In 
such cases there is invariably a stage in the exhaust in which phosphorescence is 
visible wdien the smaller terminal is the negative terminal, but in which no phospho¬ 
rescence appears when the current is reversed. And similarly when there is this 
inequality of size in the terminals, a smaller air-spark will suffice to produce phospho¬ 
rescence when the smaller terminal is negative than when it is positive. Now it is 
well known that negative discharge is greatly facilitated by increasing the size of the 
terminal, so that we have a case in which, when all the other circumstances remain 
the same, we can produce phosphorescence merely by restricting the size of the 
negative terminal so as to render more violent the local action there. 
The above experiments show that the phenomenon of molecular streams can be 
produced at pressures so considerable as to deserve to be called ordinary gaseous 
pressures. We shall now endeavour to show, in the second place, that there are no 
sufficient reasons for supposing that the gaseous molecules which form the discharge 
are in any way exempt from the ordinary laws that govern gaseous media. It is true 
that their original projection is an exceptional phenomenon, and that their consequent 
motion has no analogue in ordinary gaseous media, but there are many phenomena 
which show clearly that these molecular streams are interfered with in their course by 
the circumjacent gas, much as other currents (whether of gaseous or solid matter) 
would be under like circumstances. This is, we think, made evident by the following- 
observations and experiments. 
In the first place, when a tube is being exhausted and a discharge is maintained 
through it, phosphorescence appears first in the immediate vicinity of the negative 
terminal. This is so well known that it seems to be a matter of course that such 
should be the case, and yet it is difficult to understand why it should be so, except 
on the hypothesis that the gas in the tube obstructs the path of the molecular 
streams, and lessens their velocity. The only other explanation, viz.: that it is due 
to the greater obliquity of impact on the sides of the tube farther removed from the 
negative terminal, though of course it has a very decided effect, would not in our 
opinion be sufficient to account for it. 
But we are not left to conjectural explanations to determine that the molecular 
streams are obstructed by the gaseous media through which they pass. We shall 
proceed to describe a series of experiments which put this beyond the reach of doubt, 
