ON THE SENSITIVE STATE OF VACUUM DISCHARGES. 
G15 
due to a turning aside of the molecular streams that come from the negative terminal. 
A tube was taken in which there were a number of small objects, such as pieces of 
glass. The tube being of a very high exhaust there was phosphorescence over the 
whole of the interior of the tube, and all these small objects cast shadows, the pieces of 
glass also giving out phosphorescence. If the finger was brought in contact with the 
tube a little on the negative side of any one of these small objects (Plate 29, fig. 24), 
it was found that the molecular streams that were proceeding down the tube from the 
negative terminal were so far turned aside by the cross currents of molecules that 
proceeded from the interior surface of the tube beneath the finger that they no longer 
struck the object but passed completely over it, so that it gave no shadow at all. 
And the fact that the molecular streams did not reach it was not only shown by the 
absence of a shadow but it was also directly demonstrated in the case of the small 
pieces of glass by the fact that they ceased to manifest phosphorescence ; so that it was 
established beyond contradiction that the molecular streams that formerly reached 
them no longer did so. 
On these grounds, then, we may accept the hypothesis that the virtual shadow is 
due to the beating down of the streams of molecules that pass along from the negative 
terminal. It will be remembered that they are moving at a very slight angle to the 
sides of the tube, so that a very small deflection would suffice to account for this pheno¬ 
menon. This last point seems to be an essential condition for the formation of the virtual 
shadow, for in the case of short tubes, in which the negative terminal is a wire placed 
axially and projecting a considerable distance into the tube so that the molecular streams 
would no longer proceed at a slight angle to the sides of the tube, it has been found 
difficult to get good virtual shadows even with a fairly long positive air-spark. 
An experiment ought to be mentioned here which is serviceable in rendering more 
complete the chain of proof, viz.: the production of a double virtual shadow when the 
negative is bifurcated. This was done in the tube of Mr. Crookes, to which we have 
already referred, which was furnished with an intermediate terminal which was made 
positive, while the two terminals at the extremities of the tube were made negative 
(Plate 29, fig. 25). When the finger was placed upon the tube two shadows were 
clearly seen pointing in opposite directions, showing that these shadows point from the 
negative and not to the positive ; a law the truth of which might perhaps have been 
anticipated, but which, so far as it can, serves to show the correctness of our 
conclusions, 
XXIII.-— On the relief-effect in tubes of high exhaustion with a positive air-spark. 
III. The positive luminosity and its attendant phosphorescence. 
If a tube be watched during the process of exhaustion, and the phenomena be care¬ 
fully noted which it presents when a current with a positive air-spark is passing 
through it, the positive luminosity will be found to go through very marked changes 
4 k 2 
