136 
Proceedings of the Eoyal Society of Edinburgh. [Sess. 
§ 4. The Transformation of the Energy into the Radiant Form 
BY THE Mechanism within the Atom. 
So far we have seen that the energy absorbed by the atom from the 
moving electron becomes resident in the atom as the energy of a magnetic 
current: we have next to study the subsequent transformations of 
this energy. 
Now it is well known that a closed electric current is equivalent to a 
magnetic shell which occupies the gap formed by the circuit, and whose 
magnetic moment per unit area is proportional to the current-strength. 
Precisely the same equivalence subsists between a magnetic current and an 
electric shell — an electric shell being to all intents and purposes what is 
called a charged condenser in electrostatics : so that we may, if we please, 
regard our magnetic current as equivalent to a charged condenser in the 
atom. This will be the preferable way of considering it when we are 
studying the transformation of the energy into radiant energy : for there 
is a great body of evidence in favour of the supposition that the element 
in the atom whicli performs the function of radiating is something of the 
nature of a Hertzian oscillator : * and a Hertzian oscillator is essentially a 
condenser in the act of discharging. 
Let us then denote by Q the charge on either plate of the condenser 
which is equivalent to our magnetic current, by d the distance apart of 
the plates, and by K the dielectric constant. We have 
= magnetic current x area of magnetic circuit, 
27T 
X 7ra- 
so 
O^d — 4KM6t"a). 
Introducing the value of co from (11), we have 
2e 
( 12 ) 
But in a condenser whose charge is Q and whose plates are at a distance 
d apart and have the area ird^, the energy is 
U = 
’ 
(13) 
This seems at any rate to be established as regards the fluorescent radiations of 
preparations of the sulphides of the alkaline earths. Lenard, Pauli, and others have found 
that the frequency of the radiations can be altered at will by modifying the dielectric 
constant of the solvent in a way which conforms exactly to the supposition that the 
frequency depends on capacity as in a Hertzian oscillator, and the capacity is affected by 
the dielectric constant just as in a condenser. 
