Comparing the curves C and F , it is seen that for thin 

 screens the emergent is much greater than the incident 

 scattered radiation. The greatest value of the ratio is about 

 9:1. 



Comparing the similar curves for Au, fig. iv., it is again 

 seen that a considerable lack of symmetry exists between 

 emergence and incidence radiation, though not so marked. 

 In this case the greatest value of the ratio is about 4'5 : 1 . 

 The maximum for the emergence radiation is reached at 

 about -0008 cm. 



The effects of scattering in the case of j8 rays are thus 

 very similar to those observed for y rays, a material of high 

 atomic weight being able to turn back in the process of scat- 

 tering more of the original radiation than a material of 

 smaller atomic weight. 



Comparing ourves A and B, it is observed that A reaches 

 a maximum sooner than B. A more careful examination of 

 A and B for smaller thicknesses of screen has shown that 

 the ratio of .4 to B is practically constant until about one- 

 third of the maximum reading is reached, after which the 

 ratio gradually decreases. It would appear that while the 

 ratio remains constant we are concerned with only a single 

 collision of any ^8 particle, that as the screen is further thick- 

 ened it becomes possible for a ^ particle to suffer more than 

 one collision before emerging, thus making the emergent 

 beam appear to gradually swing round from its original di- 

 rection, a greater thickness of screen being required to pro- 

 dvice the maximum intensity for very obliq^xe rays than for 

 those corresponding more nearly with the direction of the 

 original stream. 



A fuller consideration of the effects of scattering and 

 absorption for very thin films will be reserved for a future 

 paper. 



A theory of scattering similar to that proposed by Sir 

 J. J. Thomson in "Conduction of Electricity through Gases" 

 seems capable of explaining the observed results. The near- 

 ness of appi'oach of a )8 ray to a constituent of an atom will 

 determine the amount and nature of the deflection experi- 

 enced, the speed of the jS ray and the constitution of any par- 

 ticular atom being also necessary factors. 



Until a /3 ray is subject to more than one collision the 

 distribution is approximately constant for a given material, 

 the intensity of the radiation deflected by an angle 6 from 

 the original direction being a function of that angle for any 

 one material and with rays of a given quality. 



We are to consider this function of 6 as being different 

 for the different atoms. 



