8 Professor William H. Bragg [Jan. 27^ 



metal the /3 rays will cross it in their movements to and fro ; and if 

 a little air is introduced into the cavity the ionisation produced in it 

 will be a measure of the density of the /? rays, and therefore the 

 average distance each moves in the metal. Experiment shows that 

 we get twice as much ionisation in a cavity in the lead as in a similar 

 cavity in the aluminium, and we conclude that the /? particle really 

 has a longer track in the heavier metal. This experiment gives us 

 the second constant of fB ray absorption — that is to say, the rate at 

 which its energy is taken away from it : the other experiment gave 

 the chance of deflection only. AVu see that the path of a y8 ray in 

 aluminium is more direct, but of less length, than in lead ; in the 

 latter metal it has really a longer path, but it does not get so far 

 away from its starting point because it suffers so many more 

 deflections. 



Finally, let us take a problem from the x rays. Let us see how 

 we may test the idea that x and y rays do not ionise themselves, but 

 leave all the work to be done by the (B rays which they produce. 

 Suppose a pencil of x rays to pass across a vessel and to produce 

 ionisation therein. It is convenient to use, not the original x rays, 

 which are heterogeneous, but the rays which are scattered by a tin- 

 plate on which the primary rays fall. Such " tin-rays," as we often 

 call them briefly, are fairly homogeneous and give cathode-rays of 

 convenient penetration. In some experiments of mine the rays crossed 

 a layer of oxygen 3* 45 cm. wide, having a density -00137, and the 

 ionisation produced was 227 on an arbitrary scale. The result may 

 be put in the following way. Suppose, provisionally, that all this 

 ionisation is done indirectly ; the oxygen has converted so much a-ray 

 energy into cathode-ray energy, and these cathode-rays penetrating 

 their one or two millimetres of oxygen, which is all they can do, have 

 ionised the gas. Then we may say that in crossing a layer of oxygen 

 weighing 3*45 x -00137, or -00473 gr. per sq. cm., enough cathode- 

 rays have been produced to cause an ionisation of 227 units : and 

 therefore that a layer weighing one milligram per sq. cm. would pro- 

 duce 48 units in the same way. We now proceed to compare this 

 production in oxygen with the similar effect in a metal such as silver. 

 Stretching a silver foil across the chamber in the path of the rays we 

 find that under the same intensity of rays the ionisation is largely 

 increased, and the change is due to cathode-rays which the a;-rays 

 have generated in the silver. Not all these rays get out of the silver, 

 but we can overcome this difficulty by taking silver foils of different 

 thickness, drawing a curve connecting tlie effect of the foils with their 

 thicknesses, taking the curve back to the origin, and so finding what 

 would be the effect of a foil so thin that all the cathode-rays did get 

 out. In my case I found that a milligram of silver produced enough 

 cathode-rays to give an ionisation 1580. This is 33 times as much 

 as the oxygen could do. Now according to our theory this should 

 be because silver absorbs tin-rays 33 times more than oxygen does ; 



