PHILOSOPHICAL SOCIETY OF WASHINGTON. 159 



M. Mercadier, in the course of his researches in radiophony, 

 passed an intermittent beam from an electric lamp through a prism, 

 and then examined the audible effects produced in different parts 

 of the spectrum. (Comptes Rendus, Dec. 6th, 1880.) 



We have repeated this experiment, using the sun as our source 

 of radiation, and have obtaiued results somewhat different from 

 those noted by M. Mercadier. 



(1.) A beam of sunlight was reflected from a heliostat (A, Fig. 

 12) through an achromatic lens, (B,) so as to form an image of the 

 sun upon the slit (C.) 



The beam then passed through another achromatic lens (D) and 

 through a bisulphide of carbou prism, (E,) forming a spectrum of 

 great intensity, which, when focused upon a screen, was found to be 

 sufficiently pure to show the principal absorption lines of the solar 

 spectrum. 



The disk interrupter (F) was then turned with sufficient rapidity 

 to produce from five to six hundred interruptions of the light per 

 second, and the spectrum was explored with the receiver, (G,) 

 which was so arranged that the lamp-black surface exposed was 

 limited by a slit, as shown. 



Under these circumstances sounds were obtained in every part of 

 the visible spectrum, excepting the extreme half of the violet, as 

 well as in the ultra-red. A continuous increase in the loudness of 

 the sound was observed upon moving the receiver (G) gradually 

 from the violet into the ultra-red. The point of maximum sound 

 lay very far out in the ultra-red. Beyond this point the sound 

 began to increase, and then stopped so suddenly that a very slight 

 motion of the receiver (G) made all the difference between almost 

 maximum sound and complete silence.* 



(2.) The lamp-blacked wire gauze was then removed and the 

 interior of the receiver (G) was filled with red worsted. Upon 

 exploring the spectrum as before, entirely different results were ob- 

 tained. The maximum effect was produced in the green at that 

 part where the red worsted appeared to be black. On either side 

 of this point the sound gradually died away, becoming inaudible on 

 the one side iu the middle of the indigo, and on the other at a short 

 distance outside the edge of the red. 



* The results obtained in this and subsequent experiments are shown in a tab- 

 ulated form in Fig. 14. » 



