ON THE SENSITIVE STATE OF VACUUM DISCHARGES. 
595 
no radical difference between the nature of the discharge in high and low vacua, and it 
does away with the idea (apparently suggested by the phenomena of phosphorescence) 
that in high vacua the discharge is only derived from the negative terminal. In 
addition to this, the establishment of the fact that even in the absence of positive 
luminosity the positive terminal may be the prime source of the electric discharge, 
and the negative discharge may be only a response to the positive, sheds important 
light upon the functions of the molecular streams which accompany the negative 
discharge, and affords a strong argument in favour of the view that they do not 
represent in any sense the discharge itself, nor have any necessary connexion there¬ 
with, save as being accompanying phenomena of the passage of the negative electricity 
from the negative terminal, however such passage may be brought about. 
There is a good deal of difficulty in applying directly the results of our former 
investigations to tubes of high exhaustion. The positive luminosity, which in tubes 
of lower exhaust constitutes the main feature of the discharge, and from the behaviour 
of which we obtained the indications of the nature of the discharge, fades away, as we 
have seen, into a thin haze with outlines so vague and shadowy as to be with difficulty 
discerned; indeed, when the exhaust is very high, the positive luminosity is either so 
faint as not to be discernible in the presence of the more brilliant luminous effects of 
phosphorescence, or else is actually absent. But so long as it is present it enables us 
to obtain evidence as to the character of the discharge which, if not equally convincing 
with that obtained under the more favourable circumstances of lower exhaust, would 
at all events suffice to give great probability to the hypothesis that there is no radical 
difference in the laws of the discharge so far as its capability of possessing either sign 
is concerned. With a positive air-spark the haze (whenever it is present) is repelled 
by the finger, and is constricted by a ring of tinfoil which is touched by the finger, 
just as the positive luminosity would be in a tube of lower exhaust; indeed, the 
sensitiveness seems rather to increase with the degree of exhaust than otherwise. 
And if the finger be passed along the tube there is the same continuity in the charac¬ 
teristics of the phenomena, showing that whatever be the nature of the disturbance it 
is the same from end to end of the tube. If the air-spark be in the negative the 
appearances are markedly different from those with a positive air-spark, and there is 
the same continuity of characteristics, but it is more difficult to identify the actual 
appearances with those which we have been accustomed to see in the case of other 
tubes. This is not to be wondered at, because we have already had to notice the want 
of sharpness of the effects with negative air-sparks, and even in our previous investi¬ 
gations we were frequently compelled to work with positive air-sparks in order to 
ensure good definition. This difficulty can, however, be overcome to a certain extent 
by having recourse to special rather than to relief effects; and if it were necessary we 
have no doubt that a great amount of evidence in favour of the proposition at the 
head of this section might be obtained in this manner. 
It is possible, however, even by the use of the methods of our previous paper, to 
