﻿252 Dr. J. W. Low on the Velocity of Sound in 



swimmer now from the end 0, increased by Rayleigh's cor- 

 rection, is an odd multiple of a quarter wave-length. The 

 difference between two successive maxima or minima is thus 

 half a wave-length. After careful trial I found that the 

 maxima, at all events for my ear, could be fixed with much 

 greater precision than the minima. 



From the observed half wave-length, v , the velocity of 

 sound in dry air at 0° temperature and 760 millim. pressure 

 was calculated by means of the formula 



where 



9 X/ at 3 S\ 



n = the vibration-frequency of the fork, 

 X = the observed wave-length, 

 a = the coefficient of expansion of air, 

 t = the temperature of the air in degrees Cent., 

 S = the vapour-tension of water, 

 and B= the barometric height. 



Any influence of the intensity or vis viva of vibration upon 

 the velocity of sound, which Regnault* thought he had dis- 

 covered, but which Binkf has with justice disputed, I could 

 not observe. Kundt* also and Kayser* have found the 

 velocity of sound invariable for different degrees of vibration- 

 energy of the sounding body. 



In making my observations I proceeded in the following 

 manner : — With the water-bottle in the hand I raised and 

 lowered the swimmer a few times until I had with tolerable 

 certainty ascertained the positions of the maxima. More than 

 one reading of the same maximum at the same time I never 

 took ; I rather returned to it four or five times in order to 

 get my observations as independent of one another as possible. 



The temperature was observed above and below at the 

 beginning and the end of each experiment. 



The tuning-forks c t , e t , g n c tl > c m were all made by Konig, 

 of Paris. The vibration-frequencies of the first four were 

 found to be in the ratio of 4 : 5 : 6 : 8, but c lH made 1023*25 

 vibrations when c n made 512. 



I used three different tubes, which were of the same length, 

 1250 millim., and of which the diameters were 28 millim., 

 17'1 millim., and 9*35 millim. 



In order to convey a clear idea of the degree of precision 

 of which the method is capable, I shall quote the details of a 



* Regnault, Kundt, Kayser (see footnotes, ante, p. 249;. 

 t l'ogg. Ann. cxlix. p. 54G (1873;. 



