ARTICLES 213 



already been the subject of an article in Science Progress 

 of July 191 2 ; it consists briefly in subjecting the rays to electric 

 and magnetic fields applied simultaneously and in such a 

 manner that their deflections are at right angles to each other. 

 Under these conditions particles having the same mass but 

 variable velocity will appear on the photographic record 

 obtained as a parabolic streak. The atoms of the ordinary 

 elements tried all gave single, or at least apparently single, 

 parabolas, but whenever neon was present in the discharge 

 tube, two parabolas appeared instead of the one expected. 



The position of these two parabolas showed that the atomic 

 weights of the atoms causing them were, roughly, 20 and 22 

 respectively. The latter was always about one-tenth the 

 brightness of the former, but otherwise similar in all respects. 

 The details of this similarity and other lines of reasoning made 

 it appear highly probable that the parabolas were caused by 

 two elements of exceedingly similar properties, and that it was 

 very unlikely that the 22 parabola could be explained away 

 by a hypothetical NeHg or, e.g., CO2 with a double charge. The 

 idea that neon was a mixture of two isotopes, and that the 

 parabolas were due to these, naturally arose, and the author 

 undertook a systematic investigation to throw light on this 

 important point. 



Attempts to Separate the Isotopes of Neon. — At the time 

 (191 3) when this was started, one had no idea of the amazing 

 degree of similarity between the isotopes of an element. The 

 constituents of neon were regarded as two distinct monatomic 

 elements occurring in the atmosphere, and for convenience 

 the hypothetical constituent was called " metaneon." The 

 accuracy of positive-ray analysis then available was not suffi- 

 cient to distinguish between the atomic weight of the element 

 causing the lighter parabola and the accepted atomic weight 

 (20*20) of atmospheric neon, so that the most hopeful line of 

 attack seemed that of separation. The first method tried was 

 fractional distillation. The vapour pressure of neon over char- 

 coal cooled in liquid air is of a convenient order, and a pure 

 specimen of the gas was carefully fractionated over this sub- 

 stance in a specially designed apparatus. At the same time 

 a quartz microbalance was devised to determine the densities 

 of the fractions with rapidity and with an accuracy of -^ per 

 cent. The result of a laborious series of several thousand 

 operations was completely negative within the error of experi- 

 ment ; it is, however, of some satisfaction to the operator to 

 know now that this result was probably inevitable, as has 

 been recently shown by Prof. Lindemann from theoretical con- 

 siderations. 



The second attempt at separation was made by means of 



