Mechanics of Luminosity. 393. 



consequently at the time f=24x 10 -18 sec. the initial intensity 

 has sunk to -f^. 



The duration of the vibrations investigated was J 10 _u . 

 Here, then, whilst the light has sunk to ^ - of its initial in- 

 tensity, 2*4 x 10" 13 /i 10 " u vibrations, i. e. only about TO Joo 

 vibration, have been performed, which is quite absurd. 



According to an analogous calculation for sodium, 



I_(10--) 2 l-3xl0 6 x5 2 _ 



b~- 3-2 x 10 3 ~ yX1U * 



The brightness^- would be reached at a time 9xl0 -17 

 seconds. The duration of a vibration for sodium is J 10~ 14 

 seconds ; consequently a decrease of brightness to ^ would 

 have taken place after 9 x 10 _1/ 7J 10~ u vibrations, i. e. 4*5/100 

 of a vibration. 



Thus for sodium as well as for platinum the above assump- 

 tion leads to quite contradictory results. 



The above considerations thus lead to the same result. The 

 store of luminous energy is dependent upon the vibrations of 

 the material molecules, and not upon those of the luminiferous 

 cether. 



The differences in the spectra of different substances thus 

 depends upon the different constitution of their material mole- 

 cules and not upon that of the surrounding aether envelopes. 



With the luminous atoms, moving freely through space, 

 which yield line-spectra, these vibrations can only consist in 

 opposite changes of position of the constituent parts of a 

 material atom, which therefore can have no absolutely fixed 

 form ; they are thus exactly analogous to the vibrations of 

 a deformed sphere. By this proof we obtain a secure basis 

 for those theoretical views which seek to deduce special 

 oscillatory motions for each body from the equations which 

 hold good for elastic bodies. 



For the undecomposed luminous compounds (which even 

 with the greatest dispersion yield bands not resolved into 

 lines, p. 376), it is displacements of the separate entire atoms 

 which compose a molecule with reference to each other which 

 call forth luminosity. Since the amplitudes are so very small 

 in comparison with the diameter of a molecule, such vibra- 

 tions may very well occur without the separate atoms passing 

 beyond their mutual spheres of action, that is without decom- 

 position occurring. 



From what has been already shown it follows at once that 

 with the altogether different nature of the forces upon which 

 the vibrations depend, which are in the case of free atoms of 

 the nature of elastic forces, and in the case of molecules 



