Hydrogen Nuclei from Swift a Particles. 895 



which slips into a saw-cut in the case. This is normally 

 concentric with the case, but is capable of being altered in 

 position from outside. Its object is twofold : first, it keeps 

 the plane in which the leaf moves parallel to the sides of the 

 case ; secondly, it permits of the sensitivity being adjusted 

 to any desired value after the instrument is set up. This is 

 achieved by varying the eccentricity of the leaf in the circular 

 arc of the plate. The sides of the case are of brass, in which 

 are cut windows, about 3 mm. by 2 mm., through which the 

 leaf is observed. These are covered with mica, which may 

 be made conducting by painting with calcium chloride. 



The capacity of the whole electrode s} r stem is 4*3 cm. 

 Several determinations of this quantity were made, using 

 two different methods. The final result is probably correct 

 to 5 per cent. The first method consisted in sharing the 

 charge on the electrode with a small lead sphere and 

 observing the change in potential. Precautions were taken 

 as far as possible to eliminate the effect of leads and to 

 reduce the effect of neighbouring conductors, and the 

 results obtained agreed well with those given by the second 

 method. This consisted of observing the rate of leak of the 

 electroscope under standard conditions, first with a small 

 cylindrical condenser connected to the electrode and then 

 with it disconnected. The condenser was so designed that 

 its capacity was calculable. 



3. The Experiments. 



As mentioned in the first section, the effect observed was 

 the ionization due to recoil atoms produced by a nearly 

 homogeneous beam of the swiftest a. particles from the 

 source. The velocity of the slowest was, in fact, about 

 3 per cent, less than that of the fastest. The experiment 

 consisted of making such observations with different thick- 

 nesses of absorbing material between source and ionization 

 chamber. The observed ionizations were then plotted 

 against the absorption. Curves Gr and H (fig. 2) were 

 obtained in this way. The ionization is in arbitrary units, 

 but a knowledge of the cons! ants of the apparatus makes it 

 possible to analyse the curves and actually to obtain from 

 them the number of recoil atoms thrown within selected 

 angles of the incident a ray. 



Curve Gr was made with a nickel plate coated with 

 thorium C as a source of a rays. There was an air-gap 

 of about 2 mm. between source and wax screen, so that 

 the maximum range of the a particles incident on the 



