1881.] ' 5Jt) f Chase. 



forward to a greater geueralizatiou which shall include elasticity as a form 

 of motion." — J. T., in Am. Jour, of Science, Nov. 1881. 



My first physical paper (Proc. Am. Phil. Soc, ix, 283-8), deduced ap- 

 proximate values of solar mass and distance from the combined action of 

 daily rotation, yeai'lj' revolution and atmospheric elasticity. All my sub- 

 sequent radiodynamic investigations have been based upon the considera- 

 tion of the various forms of harmonic relation which ought to folloAV from 

 the undulations of an all-pervading elastic medium, such as the luminifer- 

 ous aether Is generally supposed to be. 



141. Harmonic Spectra. 



Schuster {Proc. Roy. Soc, xxxi, 337-47), discusses the probability of acci- 

 dental harmonic coincidences in spectral wave-lengths, giving the follow- 

 ing summary of his results for the iron spectrum : — 



"1. There is a real cause acting in a direction opposed to the law of har- 

 monic ratios, so far as fractions formed by numbers smaller than seventy are 

 concerned. 



"2. After elimination of the first cause a tendency appears for fractions 

 formed by two lines to cluster round harmonic ratios. 



"3. Most probably some law hitherto undiscovered exists, tchich in special 

 cases resolves itself into the law of harmonic ratios." 



The comparison between the planetary harmonic roots and the spectral 

 harmonic quotients (Note 37), suggests the probability that the opposition 

 to strict harmonic ratios may be due to diiferences of inertia in the w^ave- 

 systems, which would be more rigidly harmonic were it not for such dif- 

 ferences. The simple tendency of all elastic media to harmonic vibrations 

 would then be the general law, instead of a law which becomes harmonic 

 "in special cases." 



In waves which are propagated with such rapidity as those of light, it 

 seems reasonable that there maybe large factors of harmonic length, which 

 are modified by smaller disturbing elements. Schuster's analysis does not 

 reach such cases as are given in Notes 36, and 39-42. Note 36 gives 11 

 harmonic divisors, which deviate from the observed divisors by a mean 

 amount of less than -^^ of one per cent. The greatest difference is in the 

 C line, where the harmonic divisor is 1.1530, the observed divisor being 

 1.1592, giving a deviation of -||^ of one per cent. In Note 39, the greatest 

 dift'erence between either of the hax'monic lines and the corresponding basic 

 line is \ of one per cent. In Note 41 the greatest difference is -^^ of one 

 per cent. In Note 42, -^^ of one per cent. 



142. The Magnesium Spectra. 



Liveing and Dewar (Proc. Roy. Soc, xxxii, 189-203), give some re- 

 sults of their investigations on the spectrum of magnesium, which seem to 

 strengthen the probability of large harmonic factors, modified by small 

 disturbances. The only two single lines which are found in the flame- 

 spectrum, the arc-spectrum and the spark-spectrum, have wave-lengths- 



PROC. AMEK. PHILOS. SOC. XIX. 109. 3w. PKINTED DEC. 31, 1881. 



