82 Professor G. F. Fitz Gerald [March 21, 



the hypothesis that the atomic charge oscillates across the diameter of 

 the atom, we find that it may be millions of millions of times as great 

 as Prof. Wiedemann has found to be the radiating j)ower of a sodium 

 atom in a Bunsen burner, so that if there is reason to think that any 

 greater oscillation might disintegrate the atom, it is evident that we are 

 still a long way from doing so. It is to be observed that ordinary light 

 waves are very much longer than the period of the vibration above re- 

 ferred to. Dr. Lodge has pointed out that quite large oscillators in 

 comparison to molecules, namely, about the size of the rods and cones in 

 the retina, are of the size to resound to light waves of the length we see, 

 and so might be used to generate such waves. This seems to show 

 that the electro-magnetic structure of an atom must be more compli- 

 cated than a small sphere or other simple shape with an oscillating 

 charge on it, for the period of vibration of a small system can be 

 made long by making the system complex, e. g. a small Leyden jar 

 of large capacity with a long wire wound many times round connect- 

 ing its coats, could easily be constructed to produce electromagnetic 

 waves whose length would bear the same proportion to the size of the 

 jar as ordinary light waves do to an atom. The rate at which the 

 energy of a Hertzian vibrator is transferred to the ether is so great 

 that we should expect an atom to possess the great radiating power it 

 has. This shows, on the other hand, how completely the vibrations 

 of an atom must be forced by the vibrations of the ether in its 

 neighbourhood, so that atoms, being close compared with a wave 

 length, are, in any given small space, probably in similar phases of 

 vibration. It is interesting to consider this in connection with the 

 action of molecules in collision as to how far the forces between 

 molecules after collision is the same as before. In the same connec- 

 tion the existence of intra-atomic electromagnetic oscillations is 

 interesting in the theories of anomalous dispersion. An electro- 

 magnetic model of a prism with anomalous dispersion might be con- 

 structed out of pitch, through which conductors, each with the same 

 rate of electromagnetic oscillation, were dispersed. In theories of 

 dispersion a dissipation of energy is assumed, and it may be the 

 radiation of the induced electromagnetic vibrations. These can 

 evidently never be greater than the incident electromagnetic vibra- 

 tion, on account of this radiation of their own energy. In some 

 theories a vibration of something much less than the whole molecule 

 is assumed, and the possibility of intra-atomic electromagnetic oscil- 

 lations would account for this. Some such assumption seems also 

 required, in order to explain such secondary, if not tertiary, actions 

 as the Hall effect and the rotation of the plane of jDolarisation of 

 light, which are, apparently at least, secondary actions due to a reac- 

 tion of the matter set in motion by the radiation on this radiation. 



Some further diagrams were exhibited, plotted from Hertz's 

 theory by Mr. Trouton, to whom much of the matter in this paper 

 is due. They are here reproduced, and show eight simultaneous 

 positions of the electric and magnetic waves during a semi-oscillation 



