Anode Rays to the Investigation of Isotopes. 861 



On changing the fine tube the trouble was cured. It is 

 easy to tell when a photograph has been atYected in this 

 way, as every line on the plate is similarly doubled, and 

 the position of the components of such doublets with respect 

 to each other is different from what it would be if they were 

 caused by rays of different e/m. It may happen that this 

 effect only occurs when the stray field from the plates is 

 such as to deflect the beam towards the edge of the tube 

 where the fault is. This results in a marked difference 

 between the appearance of photographs taken with deflex- 

 ions in different directions, which adds considerably to the 

 difficulty in guessing the cause of the fault. 



Lithium. 



A large number of photographs have been taken with this 

 metal, both for its own sake and afterwards as a standard 

 with which to compare other elements (see PL XXVI.). 



The most striking result obtained was that published with 

 Dr. Aston *: namely, that lithium consists of two isotopes of 

 atomic weights 6 and 7. The accepted atomic weight of 

 lithium is 6'94, so that there should be 6 per cent, of the 

 lighter element. It is difficult to compare the relative in- 

 tensity of photographic lines produced by direct impact, 

 as there is no easy way of obtaining a blackening curve 

 for the plate ; but in agreement with a result obtained by 

 Dempster f , it appears that the ratio of the intensities is not 

 constant. I have obtained two photographs, A and B, such 

 that the " 7 " line in A is stronger than that in B, while the 

 " 6" line in A is weaker than that in B ; in neither case 

 was there any approach to solarization. This seems to exclude 

 any possibility of explaining the effect as a peculiarity of the 

 blackening curve, and it is confirmed by a large number of 

 other photographs which show a variation in the relative 

 intensities over a wide range. A variation in proportion of 

 this kind seems more than can be accounted for by any dif- 

 ference in properties caused solely by the difference in mass, 

 as is the case with all the means so far used for the separa- 

 tion of isotopes, and apparently indicates a real difference in 

 "chemical" properties analogous to that between different 

 elements of the same series. 



As far as can be seen from mere inspection, the proportion 

 of the intensities in the plates which show the greatest con- 

 trast is about that to be expected from the atomic weight. 



* ' Nature,' vol. cvi. p. 827. 

 t < Science,' April 15th, L9L>1. 



