(■i ) 



3. If the expression (3) for the potential energy of an electron 

 is replaced by 



the isotropy has disappeared. 



One of the components of the electric doublet now coincides with 

 the original line. 



We will do well therefore, not to attach too much importance to 

 the simple ratio of the displacements of the components of the doublet, 

 which follow^s from the considerations in § 2. In the following 

 discussion we consider a doublet, which can have yet very different 

 [)Ositions relatively to the original line. 



4. Ten years have passed already since the appearance of Voigt's 

 first paper concerning an electric analogue of the magnetic spectral 

 effect, but till now physicists have not succeeded in verifying its 

 existence. Two reasons can at once be given for this negative result. 



Some idea of the probable order of magnitude of the electric effect 

 can be inferred from observations concerning the influence of an 

 electric field on the refractive index. This estimate gives extremely 

 small values for the electric change of frequency. 



According to Voigt's estimate the change of frequency in a field 

 of 30.000 Volts per cm. would hardly amount to the \/iooo^'' part of 

 the distance of the sodium lines. A field of 3000 Volts per cm. 

 would again diminish it 100 times. 



Even if a source of light giving very narrow spectral lines could 

 be placed in the mentioned intense electric fields, the observation of 

 the electric effect would not be without difficulties. 



A greater difficulty than the smallness of the effect is, however, 

 due to the impossibility of subjecting metaliic vapours to intense 

 electric fields. A sodium flame almost immediately equalizes a large 

 potential difference between the plates of a condenser. 



Circumstances are perhaps somewhat more favourable with rapid 

 electric oscillations. During part of the period of discharge of the 

 spark of a condensor the luminous vapour between the electrodes may 

 be subjected to intense electric forces. 



Of course a mere displacement of the spark lines relatively to the 

 flame or arc lines is not to be explained by the influence of electric 

 forces, now under consideration. Yet a displacement of spectral lines 

 is the first thing one may expect to observe. It will depend upon 

 circumstances whether a polarization at the borders of the displaced 

 line shall be visible. Finally this also involves the establishing of an 



