26 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



6 



of atomic numbers. Again, the values of e and - used in the 



calculations are still subject to some uncertainty (an alteration in 

 the accepted value of e from 1.13 to 1.55 X io~ 20 has changed the 

 number of electrons calculated for the average molecule of air from 

 25 to 14). And lastly, it has been shown by Crowther that a large 

 part of the radiation scattered at the smaller angles is not accounted 

 for by Thomson's formula, and Webster has pointed out that this is 

 due to Thomson's neglect of a mutual reinforcement of the scattered 

 radiations from the separate electrons. With so many uncertainties, 

 the extant calculations from Barkla's results cannot have much exact 

 significance for any theory of atomic structure. 



The hypothesis of atomic numbers fails to accord with the chemical 

 properties of the elements. J. W. Nicholson, in a recent criticism 

 (Phil. Mag., 27, 541-564, 1914), has shown that Bohr's arguments 

 about the behavior of electrons in his model atom are open to objec- 

 tion, and that the system with three electrons, for example, which 

 Bohr assigned to Lithium, would actually behave like an inert atom. 

 If applied to the present theory, the hypothesis would cause the same 

 confusion. Nitrogen, with seven magnetons then in the atom, would 

 be the most electronegative element known ; Oxygen, with a single 

 group of eight, would be inert; and Fluorine (then y+i) would be 

 expected to behave like Lithium. 



A circumstance frequently made mention of on behalf of this 

 hypothesis is that the a'-particle, which is a charged atom of Helium, 

 always possesses exactly two units of charge. Now the sudden dis- 

 appearance of the a-particle (as such) when its velocity falls below 

 .82 x io 9 cm. per second can hardly be due to anything but its neu- 

 tralization at this point. If at that still enormous velocity it can 

 become neutral, one would not expect it to lose all its electrons at 

 velocities that are not very much higher, and a theory like Ruther- 

 ford's, or Bohr's modification of it, shows no reason why the two 

 electrons, if there are only two, should not be lost one by one; 

 whereas, if the neutral Helium atom is stable up to a velocity of 

 .82X10 9 cm. per second, the present theory would actually predict 

 that for a considerable range of velocity above that point the atom 

 would be stable with a deficit of two magnetons, partly because each 

 succeeding magneton is harder to extract than the previous one, but 

 mostly because the group of six which would remain comes much 

 nearer to the group of eight in its magnetic stability than does the 

 group of seven or any other small group (§6). 



Lastly, the lack of any very definite evidence of the existence of 

 atoms intermediate in mass between those of Hydrogen and Helium 



