Sound by Radiant Energy, 527 



tained in all parts of the visible spectrum, but no sounds were 

 observed in the ultra-red. 



The maximum effect seemed to me to be in the blue. The 

 sounds were well-marked in all parts of the violet ; and I even 

 fancied that the audible effect extended a little way into the 

 ultra-violet; but of this I cannot be certain. Upon examining 

 the absorption-spectrum of peroxide of nitrogen, it was at once 

 observed that the maximum sound was produced in that part 

 of the spectrum where the greatest number of absorption-lines 

 made their appearance. 



(8) The spectrum was now explored by a selenium-cell ; 

 and the audible effects were observed by means of a telephone 

 in the same galvanic circuit with the cell. The maximum 

 effect was produced in the red. The audible effect extended 

 a little way into the ultra-red, on the one hand, and up as high 

 as the middle of the violet, on the other. 



Although the experiments so far made can only be considered 

 as preliminary to others of a more refined nature, I think we 

 are warranted in concluding that the nature of the rays that 

 produce sonorous effects in different substances depends upon 

 the nature of the substances that are exposed to the beam, and 

 that the sounds are in every case due to those rays of the spec- 

 trum that are absorbed by the body. 



The Spectrophone. 



Our experiments upon the range of audibility of differen 

 substances in the spectrum have led us to the construction of a 

 new instrument for use in spectrum analysis, which was de- 

 scribed and exhibited to the Philosophical Society of Washing- 

 ton last Saturday *. The eye-piece of a spectroscope is re- 

 moved ; and sensitive substances are placed in the focal point 

 of the instrument, behind an opaque diaphragm containing a 

 slit. These substances are put in communication with the ear 

 by means of a hearing-tube ; and thus the instrument is con- 

 verted into a veritable "spectrophone," like that shown in 

 fig. 13. 



Suppose we smoke the interior of our spectrophonic receiver, 

 and fill the cavity with peroxide of nitrogen gas. We have 

 ihen a combination that gives us good sounds in all parts of 

 the spectrum (visible and invisible), except the ultra-violet. 

 Now pass a rapidly-interrupted beam of light through some 

 substances whose absorption-spectrum is to be investigated, and 

 bands of sound and silence are observed upon exploring the 

 spectrum, the silent positions corresponding to the absorption- 



* Proc. of Phil. Soc. of Washington, April 16, 1881. 

 2P2 



