RECENT ADVANCES IN SCIENCE 539 



In these experiments no additional radiation was emitted 

 as the electrons were speeded up to greater energies by in- 

 creasing the impact voltages until such values were attained as 

 caused an arc to be struck in each case between the terminals, 

 producing the P.D., when the vapour in question emitted its 

 full series arc spectrum. This arcing potential is 10*4 volts for 

 mercury vapour. Further investigation revealed the fact that 

 the above statement had to be qualified. Two investigators, 

 Bergen and Davis, discovered that when the P.D., in rising 

 from the value 4-9 volts, attained the value 67 volts, the 

 mercury vapour was stimulated to emit radiation 1849 A.U. in 

 addition to the original 253672 A.U., but still without arcing, 

 i.e. without ionisation, and without emission of the full line 

 spectrum which makes its appearance on ionisation. A 

 further investigation of this point was made by McLennan 

 and his pupils. For zinc it was found possible to stimulate 

 radiation 2139*33 A.U. as well as 3075-99 A.U., mentioned 

 above, and with cadmium vapour 228879 could be brought out 

 as well as 3260*17 on reaching the suitable impact voltages 

 as calculated by the quantum theory. The paper gives an 

 account of the experimental methods employed, and prac- 

 tical details are also given in the Phil. Mag. papers mentioned 

 above, and in other papers therein referred to. From the 

 theoretical standpoint there are several interesting points. 

 Each of the substances treated possesses definite series of lines, 

 in particular a principal series defined by 



y = (i*5- a )-(n, &) (w = 2, 3, 4 . . .) 



a and /3 being known constants for each element. It appears 

 that the convergence wave-number (1-5, a) of each series corre- 

 sponds for each element to the ionisation potential calculated 

 on the quantum basis {e.g. 10-4 volts for mercury vapour). 

 Now this has a bearing on atomic structure and in particular 

 on Bohr's theory of stationary electronic orbits. When it was 

 discovered that mercury vapour could be stimulated to emit 

 the radiation 2536-72 by an impact voltage of 4-9, it was 

 thought that this corresponded to the ionisation potential of 

 mercury, inasmuch as the vapour acquired conductivity. 

 However, this conclusion was resisted and controverted by 

 other work which pointed out that the conductivity was due 

 not to the electrons bombarding the vapour, but to photo- 



