ee 
ge 
1883, ] 133 (Chase, 
servations of absorption bands (Note 401), approximate the gamuts of light 
and sound and suggest the desirableness of some more sensitive method 
for recording audible waves and interferences than is furnished by the 
phonograph. The radial virials of light and the tangential virials of sound 
(Note 390) furnish a field for research which is almost wholly unexplored. 
In view of the wonderful advance of spectral photography during the last 
decade, we may venture to hope that the record may sometime be extended 
so as to include the interferences of sound-waves. 
408. Investigators of Spectral Harmony. 
The earliest indications of harmony in spectral lines of which I have 
found any record, were given by Prof. Gustavus Hinrichs, in the Ameri- 
can Journal of Science for 1864 (vol. xxxviii, p. 31, seq). In the Comptes 
Rendus of the French Academy, for 1869 and 1870, Lecoq de Boisbau- 
dran published several harmonies of a character analogous to those of 
Hinrichs, his first paper being deemed of so much importance that the 
Academy allowed its insertion without abridgment, although it exceeded 
the statutory length. He referred to a communication of Mascart, on the 
‘same subject, in August, 1868, and also to a pli cacheté of his own which 
was deposited in the archives of the Academy in 1865. G. Johnstone 
Stoney (Rept. Brit. Assoc., 1870 ; Proc. Roy. Irish Acad., 1871; P. Mag., 
1871) and J. L. Soret (Bid. Universelle, Sept. 15, 1871, cited in P. Mag., 
1871, xlii, 464) seem to have been next on the list. My own investiga - 
tions began in 1864, with the study of ‘‘oscillations moving with the ve- 
locity of light’? (Proc. Am Phil. Soc., ix., 408), but my first indications of 
harmonic wave-lengths were not published until 1873 (Jb., xiii, 150). 
Guided by a conviction of the physical necessity that all ethereal undula- 
tions must be harmonic, I have been led into the discovery of a great 
variety of spectral and other coordinated harmonies. 
409. Velocity of Wave Propagation. 
As there has been some misapprehension with regard to my deduction 
of the relation between the mean velocity of oscillating wthereal particles 
and the velocity of wave propagation (Note 884), it may be well to explain 
the ground on which it rests. In considering the ‘‘ uniform wave of oscil- 
lation,’’ in a star which is rotating under the condition that gf, = 0) 
(Note 879), the vés viva of a revolving particle at J (Note 381), is } as great 
as the ois viva of the same particle from the indefinite fall* which has produc- 
ed central condensation. Vis viva varies as distance of possible projection 
against uniform resistance ; therefore 7 and }/ may be taken, respectively, 
as the measures of the virials of indefinite and of virtual fall. Hence arises 
a tendency to the formation of an oscillatory node at $1, together with a 
tendency to the radial projection of the node, in the equatorial plane, by 
* This is rigidly true only when the fall is infinite, butin falling from Neptune 
toSun the deviation from exactness would be less than yy of one per cent. 
