Rontgen Radiation in Air. 



611 



But when an Al plate '265 cm. thick was placed at A, to cut 

 off the softer portions o£ the main beam, the third reading 

 above became 



265 



•88 



35 



Thus there is a marked increase in the penetrative power for 

 thick plates, but when the harder rays only are operating, 

 as in the fourth reading, there is a decided drop in P-265- 

 This seems to be direct evidence of a transformation by air 

 into a softer radiation of those rays which were able to get 

 through the '265 cm. plate at A. 



The results given in this table were so remarkable that 

 two new curves were constructed giving compound exponential 

 curves similar to those in fig. 4, analysing the radiations from 

 the hard and soft bulbs. For each electroscope the deflexions 

 per unit time were plotted against the thickness of plates at 

 P in Curve I. (fig. 5) and at S in Curve II. 



Fiff. 5. 



The curves show that the secondary deflexions fall less 

 rapidly than the primary, and that they become homogeneous 

 more quickly. Now if the secondary beam becomes homo- 

 geneous before the primary, then either the hard primary 

 rays must fail to be scattered, or else they are transformed 

 into a more absorbable radiation. 



Now let us plot a new curve in which each ordinate is the 

 corresponding ordinate in the secondary curve divided by that 

 in the primary. According to this curve (No. III. in fig. 5) 



