212 Prof. Barkla and Mr. Ay res on the Distribution of 



Calling, the three angles POA, POB, POC, a, ft and y 



respectively, 



sin 2 /3=l-cos*/3 

 and cos 2 j3=l — cos 2 a — cos 2 y , 



and as the average value of cos 2 ft = average of cos 2 7 

 o n 1— cos 2 CL 



.'. average sin 2 /3 = average 



2 

 1-f- cos 2 a 



It follows that the intensity of scattered radiation in any 

 direction is proportional to 1 + cos 2 a_, and may be written 



I tt = T ff / 2 (1+ COS 2 a), 



the suffix denoting the angle between the direction of radia- 

 tion considered and that of propagation of the primary 

 radiation. 



This expression gives also, as was shown by Lord Rayleigb, 

 the distribution of monochromatic light scattered by fine 

 particles, though we are not aware of any experimental 

 verification in the case of light. As previously stated, the 

 relative intensities in the two principal directions — those in 



IT 



which a= — and as nearly ir as practicable — have been 



studied by one of us*. 



In order to test the theory more completely it seemed 

 desirable to make a series of measurements at various angles. 



Carbon was used as the scattering substance, because when 

 it is subject to a beam of moderate penetrating power, the 

 secondary radiation at a distance of several centimetres in 

 air is principally if not entirely scattered radiation. If a 

 true secondary X-radiation — fluorescent X-radiation — is 

 emitted by carbon and other light elements, it is either very 

 easily absorbed and does not penetrate that distance in air, 

 or is so penetrating that it is not excited by a moderately 

 penetrating primary radiation. It seems probable that there 

 are both types of secondary radiation, as with heavier elements. 

 To avoid complications due to this fluorescent radiation the 

 X-ray tube was kept fairly soft throughout the experiments,. 

 The secondary beam studied was thus entirely a scattered 

 X-radiation. 



* Phil. Mag. Feb. 1908, pp. 288-296. 



