﻿852 Mr. J. H. Van VJeck on the normal Helium Atom 



and to retain the latter it would appear necessary to make 

 some artificial and improbable assumption concerning the 

 character o£ the mystery force, such as having it depend on 

 the velocity in such a way that it became negligible at the 

 high velocities found in the electrons of the K and L rings. 



» 



Conclusion. 



As yet it appears possible to devise a satisfactory symmetrical 

 model of the normal helium atom only with the aid of some 

 such radical innovation as reformulation of the quantum 

 conditions or modification of the law of force bat ween 

 negative electrons. The probability of the latter alternative 

 is discounted by the success of the ordinary quantum theory 

 of X-rays. It is to be hoped that with one such bold 

 hypothesis we can simultaneously obtain the proper energy 

 values for models both of the helium atom and the hydrogen 

 molecule. The models of normal helium which are physically 

 most plausible seem to be the Langmuir semicircular one or 

 that with axial symmetry. The zero resultant angular 

 momentum of the Langmuir model is perhaps in best accord 

 with the observed diamagnetism of helium, and if the semi- 

 circular model of the hydrogen model is correct, one would 

 expect the normal helium atom to be of a similar type. 

 However, the type with axial symmetry has the advantage 

 of requiring smaller departures from the conventional 

 quantum conditions or less readjustment of the law of force 

 in order to obtain the correct ionization potential* . Some 

 very interesting experimental evidence on the structure of 

 the helium atom is given in a recent article by Millikan j- 

 Observations taken with his oil-drop apparatus indicate that 

 when an alpha particle collides with a helium atom it hurls 

 out both electrons about 16 per cent, of the time, while 

 during the remaining 84 per cent, of the collisions it ejects 

 only one electron. This, Millikan concludes, eliminates the 

 possibility of the Bohr model of normal helium, as the chances 

 of the incident alpha particle having just the right direction 

 to collide with the second electron after already striking 

 the first one would be exceedingly small in a model of this 

 character. Models such as those of Lande or Franck and 

 Reiche, in which the two electron orbits differ considerably 

 in size, are also rejected, as the innermost orbit exposes 



* The modifications would have to produce a change of 30 volts in 

 the ionization potential for the semicircular model as compared with 47 

 volts for the type with axial symmetry. 



t Physical Review, vol. xviii. p. 456, Dec. 1921. 



