rjll] on a Fourth State of Matter 9 



and experiment showed this to be very nearly the case. In finding 

 the absorbing power of oxygen I measured first those of carbon and 

 oxahc acid, and then proceeded by calculation ; for the absorption in 

 a gas is difficult to determine. 



Two interesting points appeared in this experiment. In the first 

 place the ratio between the two quantities of cathode rays, which 

 appear on the two sides of a silver leaf through which the " tin rays " 

 pass, is nearly constant for different thicknesses of leaf, and with the 

 thinnest leaf obtainable each quantity \vas about half its full value. 

 It would have been desirable to have had still thinner leaves ; but it 

 is fairly clear that the ratio would be nearly the same for extreme 

 thinness. The cathode radiation, which appears on the side of the 

 leaf whence the x rays emerge, is 1 • oO times that which appears on 

 the other, and we may take it that this would be the case even if the 

 leaf were but one atom thick. Thus when an x ray plunges into an 

 atom in which its energy is converted into that of a cathode ray, the 

 cathode ray may emerge at any point, but there is a 30 per cent, 

 greater chance that it will more or less continue the line of motion of 

 the X ray tlian that it will not. In previous work on the conversion 

 of y ray into (3 ray energy, I have found that the /3 ray may 

 practically be supposed to continue the line of motion of the y ray, 

 so that there is a great difference in behaviour of the two classes of 

 ray in this respect. It is remarkable that the scattering of the y rays 

 shows also a much greater dissymmetry than is found in the case of 

 the X rays. It looks as if the /? rays that appear when y or a; rays 

 impinge on atoms are related rather to the scattered than to the un- 

 scattered primary rays. Putting it somewhat crudely, no doubt, it 

 might be said that when a y or ;c ray is deflected in passing through 

 an atom, it runs a risk of being converted into a /5 ray in the process, 

 so that ji rays are found distributed about the atom in rough pro- 

 portions to the secondary y or « rays. In the case of y rays this 

 practically amounts to their all going straight on at first : in the case 

 of X rays the distribution is more uniform. 



Another interesting point arises in this way. When the x rays 

 from tin are allowed to pass into the ionisation chamber through 

 increasing thicknesses of silver foil, the cathode rays grow at a rate 

 which is not represented by the exponential curve usually assumed. 

 The amount is for some time more nearly proportioned to the thick- 

 ness of the foil. A second foil adds its own effect without destroying 

 much of the one on which it is laid. This may easily be ascribed to 

 the relation of the ionisation due to the /? particle to the energy it 

 has to spend. The ionisation is nearly all at the end of the path, 

 and tlie second layer does not absorb the rays made in the first 

 because they are still at the beginning of their career. 



These few experiments which I have described may serve to 

 illustrate both the justice and the convenience of placing all these 

 rays, a, j3, y, and x, in one class. We are tempted to consider them 



