Experiments on Positive Rays. 237 



can go some way before making the first collision, and then 

 make up for the loss of energy caused by the delay in making 

 the first collision by making the second collision a little 

 earlier. There is thus a much wider scope for adjustment in 

 this case than in the previous one ; and we should expect 

 the number of particles with approximately half the maximum 

 kinetic energy to be large compared with the number having 

 nearly the maximum. As a matter of fact, however, the 

 lines corresponding to the doubly charged atoms are con- 

 spicuous by the uniformity in their intensity. This is incon- 

 sistent with the ionization occurring in two stages ; and we 

 conclude that in some of the collisions which produce the 

 charged ions the conditions are such that the ions receive a 

 double charge. 



Beading in the Molecular Lines. — We have seen that the 

 possibility of double-charged ions gives a beaded appearance to 

 the atomic lines, the beads appearing nearer the vertical line 

 than the main portion. The molecular lines do not show 

 this type of beading, but they show frequently another type 

 in which the beads, instead of being nearer the vertical axis 

 than the head of the parabola, are further away from it. A 

 conspicuous example of this is shown in fig. 21 b, in which the 

 larger quantity of the molecules of hydrogen and oxygen occur 

 after the first beading. Sometimes, as in figs. 16 and 16 a, 

 there are several sets of beads following each other at short 

 intervals. The only atoms in which I have observed clear 

 indications of this type of beading are those of the mon- 

 atomic gases helium and mercury vapour, and for these gases 

 there is no distinction between the atom and the molecule, so 

 that this does not form an exception to the rule. This kind 

 of beading depends very much on the position of the cathode 

 in the discharge-tube and its shape. With some cathodes I 

 have never observed them ; with others I have found them 

 when the cathode was in one position, not when it was in 

 another. I think these are due to the impact of the cathode 

 rays against the walls of the glass giving rise to slow 

 secondary rays of great iouizing power. A slight inclination 

 of the cathode might cause the rays to hit a portion of the 

 discharge-tube which previously escaped, and thus start 

 secondary rays nearer to the cathode, which would produce 

 ions reaching the cathode with smaller energy than those 

 which started from the negative glow. 



It is remarkable that on many of the plates the head of 

 the first beading on the molecular line is just twice the 



