642 
MR. W. CROOKES OK MOLECULAR PHYSICS IN HIGH VACUA. 
there is absolutely no phosphorescence when the sensitive surface is fully -within the 
narrow dark space surrounding the negative pole in imperfect vacua. As the rarefaction 
improves the dark space widens out, and phosphorescence begins to appear outside the 
luminous margin, but not inside it. On further exhausting, the luminous boundary of 
the dark space gets fainter and larger, till it disappears, and now the phosphorescence 
extends all over the sensitive surface. Several pieces of apparatus were constructed 
in order to test this fully. Negative poles of flat, convex, and concave shapes were 
experimented with, and various substances were used as the sensitive material for 
rendering apparent, by phosphorescence, the molecular rays. The result in all cases has 
been to confirm the first observation, viz. : that there is no phosphorescence within the 
dark space. 
588. Experiments previously described have, I think, shown that the molecular 
stream hypothesis is the correct one. According to this, the molecules of the residual 
gas, coming in contact with the negative pole, acquire a negative charge, and imme¬ 
diately fly off 1 by reason of the mutual repulsion exerted by similarly electrified bodies. 
Were the individual molecules solely acted on by the initial impulse from the negative 
pole, they would take a direction accurately normal to the surface repelling them, and 
would start with their full velocity. But the molecules, being all negatively elec¬ 
trified, exert mutual repulsion, and therefore diverge laterally. The negative pole, 
likewise, not only gives an initial impulse to the molecules, but it also continues to act 
on them by repulsion, the result being that the molecules move with an accelerating 
velocity the further they get from the pole. The lateral divergence of the molecules, 
owing to their negative electricity, will naturally increase with the amount of charge 
they carry; the greater the number of collisions the more the molecules lose negative 
charge, and the less divergent the stream becomes. This hypothesis is borne out by 
facts. When the vacuum is just good 1 enough to allow the shadow to be seen, it is 
very faint (owing to few molecular rays), but is quite sharp (owing to the divergence 
of the molecules laterally). The variation in mutual repulsion is shown by the fact 
that the focus projected from a concave pole falls beyond the centre of curvature, and 
varies in position with the exhaustion, being longer at high than at low exhaustions. 
589. Assuming that the phosphorescence is due, either directly or indirectly, to the 
impact of the molecules on the phosphorescent surface, it is reasonable to suppose that 
a certain velocity is required to produce the effect. Within the dark space, at a 
moderate exhaustion, the velocity does not accumulate to a sufficient extent to produce 
phosphorescence; but at higher exhaustions the mean free path is long enough to 
allow the molecules to get up speed sufficient to cause phosphorescence. At a very high 
exhaustion the phosphorescence takes place nearer the negative pole than at lower 
exhaustions ; this I consider results from the initial velocity of the molecules being 
sufficient to produce phosphorescence, their greater speed being due to the fewer 
collisions near the negative pole. 
The luminous boundary to the dark space round the negative pole is probably due 
