PHYSICS: A. J. DEMPSTER 
375 
would be parabolas; and in a third experiment (C) the particles passed 
between the plates of a condenser which would deflect any charged 
particles sideways out of the beam. 
The phenomena depended greatly on the pressure of the gas in the 
tube. If the pressure were very low so that the positives made practi- 
cally no collisions with the molecules of the gas, the rays, which were 
followed by the luminosity they make in the gas, could be stopped 
entirely by the directly opposing field or bent into a parabolic path 
by the inclined field or deflected sideways by the cross electric field. 
A and B indicate the appearance with high vacua. The rays could be 
detected with potentials as low as 30 volts on the tube, although they 
were then very faint. To stop the rays, or rather to reduce their speed 
so that they are unable to cause Hght, opposing potentials were required 
which were practically the same as those which originally gave the par- 
ticles their speed. We conclude that very slow positive rays are still 
able to excite light, certainly with a speed corresponding to less than 5 volts. 
A B C 
This fact needs to be considered in connection with the above mentioned 
experiments on the lower limit at which electrons excite light. 
If the pressure of hydrogen is taken a Httle higher so that some of the 
positives make colKsions, the phenomena are more complicated. If 
the positives are then deflected out by a cross field, there is an unde- 
flected bundle of light left, due to the neutral rays which have been 
formed from the positives. C indicates the appearance at about 0.005 
mm. of mercury pressure when the two bundles are of approximately 
equal intensity. With the arrangements A and B there is superimposed 
on the deflected beam a luminosity continuing up to the opposing 
electrode. If the pressure is taken higher still, the positive bundle 
becomes fainter till finally the neutral bundle alone is present. Two 
things are of special interest. First, the neutral rays can also excite 
light at very slow speeds. The lowest speed at which the neutral bundle 
has been detected is 50 volts, and it is possible that that is a lower 
limit at which the neutrals begin to excite light. With potentials of 
1500 volts, when the light is much stronger, photographs of the spectra 
