XV. 5] 
THE ZEEMAN EFFECT 
215 
5. Magnetic or Relativity Doublets 
Observations on the anomalous Zeeman effect lead to the 
conclusion that optical doublets must have a magnetic origin. 
It seems justifiable to apply the same reasoning to the case of 
X-rays, although for X-rays Sommerfeld has interpreted terms 
such as L n and Lni as simple relativity doublets and has shown 
that their separations can be calculated accurately on that basis. 
Lande * has pointed out that the relativity and optical doublet 
separations can both be represented by the same general formula. 
The view that the optical and X-ray doublets are essentially 
similar in origin is borne out by the distribution of electrons in 
various levels advocated by Stoner. 
Sommerfeld draws the following important conclusion from 
Larmor’s theorem : 
“ The electron describes in the magnetic field the same path 
as when no magnetic field is acting but doing so with respect 
to a system of reference which is rotated with the velocity o. 
Regarded from the standpoint of this system of reference the 
orbits are traversed as if no field were present. Precession of 
the system of reference and action of the magnetic field are 
interchangeable and equivalent to one another.” 
In view of this statement we may reverse the order and con 
sider precession (as it occurs for instance in relativity theory) as 
replaced by the action of a magnetic field. From a mathematical 
standpoint it may be simpler to replace a magnetic field by a 
motion of precession. From a physical standpoint it may be 
simpler to replace a motion of precession by the action of a 
magnetic field. These remarks are of importance in connection 
with the work of Millikan and Bowen f on “ stripped ” atoms, 
i.e. atoms which have been stripped of practically all their 
electrons. These authors conclude that their results require one 
of two alternatives : (1) discarding the relativity explanation of 
X-ray and optical doublets and with it all relativity effects in 
electronic orbits, or (2) introducing a dissymmetry not hereto 
fore contemplated in atomic models, abandoning Bohr’s inter 
penetration ideas and his assignment of azimuthal quantum 
numbers. 
Here, then, the older quantum theories led to conflicting 
results. There appeared to be two different types of explana 
tion, one in terms of relativity, the other in terms of a mag 
netic field. By the employment of the spinning electron it is 
possible to account for all the results in terms of a single theory. 
* Lande, Zeits. /. Physik, vol. 24, p. 88, and vol. 25, p. 46, 1924. 
t Millikan and Bowen, Phys. Rev., vol. 24, pp. 209, 223, 1924.