266 Mr. W. Sutherland on the 



atoms have frequencies of the same order as those known in 

 luminous spectra ; and as we shall show in the next section 

 that the frequencies of the free vibration of the aether of an 

 atom are far higher, we must consider the electromagnetic 

 radiation from atoms to be due to a vibratory electrical move- 

 ment in the atom which is forced and maintained by the 

 mechanical vibrations. Thus the atom which has of late been 

 relegated to the background through the prominence of the 

 electron must be rehabilitated. In Section 8 an attempt will 

 be made to assign to atom and electron their respective shares 

 in the production of spectra. 



7. Free Vibrations of the JEther of Atoms. 



By the aether of an atom is meant the aether which takes 

 part in the transmission of light through the atom. We may 

 regard it as a portion of the aether bounded by the least sphere 

 which would inclose the atom, and as having its properties 

 altered by the contiguity of the matter of the atom. Now 

 the phenomena of the refraction of light make it clear that 

 the order of magnitude of the velocity of light through atoms 

 is of the same order as that of its velocity through free aether, 

 namely, 3 x 10 10 . But this is about 10° times the velocity of 

 a mechanical disturbance calculated for the atom of Li in 

 the last section. Therefore the frequency of the vibrations 

 of the aether of the Li atom would be about 10 5 . times greater 

 than that of the mechanical vibrations of the atom, and there- 

 fore the free vibrations of the aether of atoms are of a much 

 higher frequency than those belonging to the vibrations in- 

 vestigated in ordinary spectroscopy; they may have relations 

 to the Bontgen and Becquerel ray-vibrations. Still it is 

 interesting to compare the relative frequencies of the free 

 vibrations of the atoms of metals belonging to the same and 

 different natural families. For this purpose we require the 

 velocity of light through the atom and the liuear dimensions 

 of the atom. In " Molecular Refraction" (Phil. Mag. [5] 

 xxvii.) T have shown, as others had done before, that the 

 refraction-equivalent of a substance (n — l)m/p, where n is 

 the index of refraction, is the same as the refraction-equivalent 

 of a single molecule of it, which may be denoted by {N — l)m/p, 

 where N is the index of refraction of the molecule and mjp 

 its volume. This use of the symbol N must not be confounded 

 with that of the last section, where it stands for rigidity at 

 absolute zero. Here N" stands for the ratio of the velocity 

 of light in matter-free aether to the velocity through the 

 molecule. The refraction-equivalents found for the atoms by 



