of Young' 1 s Modulus for Glass. 415 



endwise, notes so nearly identical in pitch that beats were 

 scarcely perceptible. This trial was carried out in the manner 

 described farther on, and every precaution taken to secure 

 uniformity of temperature. In the course of subsequent 

 experiments many opportunities presented themselves of com- 

 paring the notes given by pieces of these bars of approxi- 

 mately equal length, and the differences not accounted for 

 were always so small as to be unimportant, the only serious 

 difference noted amounting to about six vibrations per second 

 in the case of two pieces of exactly equal length, about -40 cms. 

 To determine the velocity of sound in the steel rods, the 

 length of one of them was cautiously altered by cutting and 

 tiling until its note nearly coincided with that of a Konig fork 

 giving 1024 double vibrations per second ; and, subsequently, 

 taking advantage of a spell of cold weather, when room 

 temperature had been long constant at 0°, rod and fork were 

 compared together, beats being counted for 20 seconds. 

 According to Konig the rate of his forks, standardized at 

 20° C, increases by "000118 per cent, for each degree below 

 standard temperature. With these data, the product nx2l 

 for the steel bar was found to be 



51G040 cms. per second at 0°. 



Similarly favourable opportunities did not occur for testing 

 the other bars at 0° ; but repeated comparisons of the bars 

 with each other, and with 768 and 896 Konig forks at other 

 temperatures, left no doubt of the substantial accuracy of 

 this determination. 



Some very careful determinations of the velocity of sound 

 in rods of various metals and of St. Gobain glass, at tem- 

 peratures ranging from 0° to 100° C, have been made by 

 Mayer*, who found for a rod of Bessemer steel at 0° 

 V = 515090 cms. per sec. In view of possible variations in 

 the composition of the metal, the agreement must be con- 

 sidered satisfactory. 



According to Mayer also the velocity of sound iu such a 

 rod diminishes by "2 per cent, when the temperature rises 

 from s t-j 20°, or by nearly 50 cms. per sec. for each degree. 



AVith these data the rates of vibration of the steel rods 

 here employed were calculated, measurements of length being 

 throughout made by a steel metre-scale, the very trifling 

 errors of which had been determined at Kew. 



In the experiments with glass, a steel rod was first cut 

 giving a note higher in pitch than any one of the bundle of 



* " Researches in Acoustics," by Alfred M. Mayer, Phil. Mag. [5] 

 vol. xli. p. 168. 



