217 
1921-22.] The Magnetic Character of the Quantum. 
with opposed currents, described above, would be equivalent to a magnetic 
shell forming the curved surface of a cylinder having its edges coincident 
with the two rings. For suppose the first ring electron is replaced by a 
magnetic shell in the form of a flat circular disk. The second electron may 
be regarded as equivalent to a magnetic shell of any chosen form provided 
only the edge of the shell coincide with the contour of the ring. Let 
this shell he constructed so as to cover the first shell and also the cylin- 
drical surface between the two rings. Then, as the currents are opposed, 
the two flat portions of the sliells neutralise one another, and we are 
left with a magnetic shell covering the cylindrical surface between the 
two rings. 
The same result may be obtained very simply by imagining the two 
rings to be connected by a conductor carrying equal and opposite currents 
of the same strength as the current in each of the ring electrons.] 
If it were found on detailed examination that such a model would 
absorb energy from an electron in the same fashion as Professor Whittaker’s 
magneton, it would possess certain advantages as compared with the latter. 
There would be no need to introduce bar magnets, and no difficulty would 
arise with regard to the passage of the electron through the matter of the 
bars — there is nothing to prevent free movement through the apertures of 
the rings. Further, such an arrangement would possess both capacity and 
inductance ; and instead of the charging and discharging of a condenser 
which is employed as the type of an oscillating system, we could deal with 
the oscillations of the two rings as they moved nearer together and further 
apart. Professor Whittaker recognises the difficulty of picturing the in- 
ductance of his assumed Hertzian oscillator, but he does not say how the 
difierential equation for such an oscillator can be applied if we retain the 
conception of an electron as an indivisible unit of electricity, and it is to 
be noticed that the electric separation in his condenser “ is precisely a 
separation of two electronic charges e and — e.” If, however, the plates 
of the condenser are moving with reference to one another, there is no 
need to consider variations in the value of the charge. This move- 
ment of the electrons as such seems indeed to be indicated in his paper 
in several places. 
§ 5. At first sight Professor Whittaker’s choice of hhr as the value of 
the natural constant ^(L/C) seemed somewhat arbitrary, and difficult to 
justify apart from the fact that it leads to the desired result Ai^ = U. I find, 
however, on further examination that it is in complete agreement with 
the generalised form of the quantum theory proposed by Wilson and 
Sommerfeld and now commonly accepted. In the present case we have 
