^60 On the Velocities of Sound in different Bodies, 



is then mucli less here than in air; but the slight errors of ex-' 

 periments leave this siibject still in some incertitncle. 



In Older to ap|)l\ this resnlt to finids. I shall take water as the 

 example. According to the experiments of Canton, recorded in 

 thefifty-second and fifty-fourth volumes of the Philosophical Trans- 

 actions, when the heat of the barometer is 0'",76, the centigrade 

 thermometer marking 10\ the pressure of the atmosphere dimi- 

 nishes the volume of water 42-.5 milHonths : the linear diminu- 

 tion is three times less. Thus a column of water of a metre hi 

 length is diminished 14"5 millionths by a pressure equal to that 

 of a vertical column of the same fluid of the height of 1()'"'^,325. 

 The reduction of the first column compressed by a weight equal 

 to its own is then O'"%000001,4044. By dividing 9™%80S8 by 

 that fraction 2642,8, the square root of the quotient will be the 

 number of metres which sound traverses in water during one 

 second ; — its swiftness in this fluid is then nearly eight times 

 greater than in air. 



The experiments of Canton upon sea-water, the specific weight 

 of which is 1,028, give 37,5 millionths as the diminution of its 

 volume by the pressure of the atmosphere; whence it follows 

 tliat sound must traverse salt-water 2807'",4 in a second. These 

 two proportions taken relatively to a temperature of ten degrees 

 vary very sensibly with it. 



Experiments to determine the velocity of sound in different 

 substances, such a^ have been made in the preceding instances, 

 appear to me well worthy of engaging the attention of phi- 

 losophers. 



On (he Velocilies of Sou?id in different Bodies. 



Newton has given in the second book of his ]\IatliematicaI 

 Principles of Natural Philosophy, a theorem of the velocity of 

 sound, which is one of the most remarkaijh; traits of his genius. 

 The velocity deduced from this theorem is about a sixth less 

 than that which results from experiments made with great care 

 in 1738 by the members of this Academy (French Academy). 

 Newton, who had already discerned diat difference from the ex- 

 periments made in his time, has endeavoured to explain it : but 

 the modern discoveries on the nature of atmospheric air have 

 destroyed that o.planation, as well as everv other which geome- 

 tricians have subseciuently |)ropo6e(]. Thee discoveries fortu- 

 nately present us with a phcenomenon which appears to be the 

 true cause of the excess of swiftness observed in sound, above 

 the degree of swiftness calculated, and which the greater [lart of 

 philosophical geometricians have since adopted. This phaeiio- 

 inenon consists in the heat which the air develops by its com- 

 pression. When its temperature is raised, its pressure remaining 



the 



