160 MR. W. CROOKES OK THE ILLUMINATION OF LINES OF 
very ill-defined and appears broken in the centre, leaving two curves of light at the 
sides. The spot of light is unsteady and ill-defined. The magnet brings it down 9'5°. 
By increasing the intensity of the spark as much as possible, the deflection caused by 
the magnet is diminished, the spot only sinking 6'5 millims. instead of 9'5 millims. 
570. The apparatus was now filled with electrolytic hydrogen, and after exhausting 
and refilling with the gas several times, the following experiments were tried :— 
Exhaustion = 427 M. Thickness of dark space = 7 millims. There is on the screen 
an exceedingly faint spot of green light, which is deflected by a magnet, but it is not 
bright enough to allow observations to be taken. 
571. Exhaustion — 147 M. Thickness of dark space =12 millims. The spot of 
light on the screen is now brighter and better defined. The magnet being brought into 
position lowers the spot 13 millims. The light in the tube shows hydrogen, both by 
the colour and to the spectroscope ; the green phosphorescence is spreading. 
572. Exhaustion = 27 M. Thickness of dark space = 15'5 millims. The spot of green 
light is much brighter and very well defined. The magnet lowers it 11 ‘5 millims. 
573. Exhaustion = 14*8 M. Thickness of dark space = 17 millims. The spot of 
light is bright and round. The magnet brings it down 10'5 millims. The tube is of 
a brighter green than usual, and there is less hydrogen light about the poles. 
574. Exhaustion =10'25 m. No dark space is to be seen. The light on the screen 
is no longer well defined, but is a mere splash of very bright green light. The rest of 
tube is of a brilliant green colour. The magnet lowers the spot of light about 
3 millims. 
575. Exhaustion = 7‘25 M. The exhaustion is getting rather too high for the green 
light, which is not quite so bright. The sparks will not always pass through. The 
magnet has very little action on the spot of light. 
576. These experiments prove several important points. In par. 559, when work¬ 
ing with an air vacuum, it is recorded that the spot of green light is visible on the 
screen at a pressure of 102 - 6 M when the thickness of the dark space is only 12 millims. 
from the pole. Assuming, as I do, that this is a measure of the free path of the 
molecules before collision, it follows that some of the molecules sufficient to cause 
green phosphorescence on the screen, are projected the whole distance from the pole 
to the screen, or 102 millims., without being stopped by collisions. It is probable 
that this would have occurred at a still lower exhaustion, for on reference to par. 566, 
it is seen that the green spot was detected on the screen when the mean path in 
carbonic acid was 8‘5 millims., and it was seen witli hydrogen (570) when the mean 
path was only 7 millims. 
577. If we suppose the magnet permanently in position, and thus exerting a uniform 
downward pull on the molecules, it is seen that their trajectory is much curved at low 
exhaustions, and gets flatter as the exhaustion increases. A flatter trajectory corre¬ 
sponding to a higher velocity of the molecules, it follows that the molecules move 
quicker the better the exhaustion. This may arise from one of two causes : either the 
