224 Mr Townsend, Secondary Rontgen Rays. 



The curves show that this secondary radiation is rapidly 

 absorbed by the air. The ionization produced in the first milli- 

 metre by the secondary rays is more than fifty times as great as 

 that produced in the sixth. At a distance of 6 centimetres we 

 should expect that the intensity of these rays would be reduced 

 to such a small fraction of its original value that it would be quite 

 impossible to detect the ionization they would produce. 



It appears therefore that at least two kinds of rays are emitted 

 by the radiating body ; one kind, which is rapidly absorbed by the 

 air, gives rise to the surface effect noticed by Perrin, and the other 

 kind which is more penetrating produces the ionization at a 

 distance. 



Defining the surface effect 8 as the ratio of the total effect 

 produced by the rays which are easily absorbed to the effect of the 

 direct rays on a layer of air 1 centimetre thick, we obtain the 

 following values of S 



ForCu,/Sf=2-5 

 „ Zn, £=1-84 

 „ AL S= "4. . 



In this examination of the experimental results we have 

 attributed 3"2 x n ions per cubic centimetre to the primary rays. 

 This is only approximately correct since the secondary radiation, 

 which extends to a distance, contributes to the formation of these 

 ions. The number 3'2 xwis therefore too big an estimate of the 

 direct effect, but is probably not more than a few per cent, in 

 excess of its true value. 



The above values of S are larger than those found by Perrin, 

 thus in the case of zinc Perrin found the surface effect to be '7. An 

 experiment was therefore made to test the accuracy of the above 

 results, the method employed being similar to that used by Perrin. 

 A beam of direct rays from the bulb was passed through the air 

 between two parallel plates without falling on either plate. The 

 ionization produced per centimetre of the beam was found to be 

 31 in arbitrary units. The same beam passing normally through 

 a very thin sheet of aluminium and falling on a copper plate, 

 parallel to the aluminium at a distance 1'5 centimetres from it, 

 produced a number of ions proportional to 140 in the space 

 between the copper and aluminium. Of these 140, 46'5 are due 

 to the direct rays, so that the effect of the surface radiation from 

 the copper and aluminium must be 93'o, or three times the effect 

 of the direct rays per centimetre. This is practically what we 

 obtained by varying the distance between the plates (2'5 + '4). 

 It appears therefore that the discrepancy between these results 

 and those obtained by Perrin must be due to differences in the 

 nature of the direct rays. 



