154 



Prof. A. LI. Hughes on Characteristic 



The character of the curves used in locating the L-radia- 

 tion for carbon (fig. 6) differs somewhat from that of those 

 obtained for boron. One has to take a point of inflexion 

 rather than a break. This no doubt has to do with the form 

 of the curve for the general radiation in this region, if it 

 could be found apart from any superposed characteristic 

 radiation. The points at which the curvature begins to 

 change to an upward direction, and which are taken to 

 correspond to the L-radiation, are 36'0, 34*0, 34*2, and 

 35'0 volts, giving a mean at about 34' 5 volts. The corre- 

 sponding wave-length is \358. 



Fig. 6. 



Photo E/ectr/c Effect 

 E/ectron Current. 



C^fPBOA/ "l " /?ac//ai/o, 



A cce/erat/ng Po ten i/'a/. 



30 



50 



60 Volts. 



The graphs reproduced in figs. 3 to 6 are chosen partly to 

 illustrate the form of the curves for some distance on either 

 side of the breaks. Several of the values for the breaks 

 were obtained from runs limited to a shorter range of 

 potentials, thus giving points closer together. It is, perhaps, 

 hardly necessary to say that the actual values were obtained 

 from graphs accurately plotted on a large scale. 



On many occasions it was found impossible to get curves 

 with a break for either carbon or boron. It was then- 

 noticed that this was generally the case after the target had 

 been in use for a long time. On taking the target out and 

 exposing a new surface, either by grinding (boron) or filing 

 (carbon), or by exposing a newly-fractured surface (carbon 

 and boron), the breaks almost always reappeared. Also,. 



