Physics 539 



of this refraction in 1852 are described. There is a clear 

 presentation of Stokes's theory of sound-refraction due to 

 wind, abstracted for the most part from his paper in the re- 

 port of the British Association for 1857. He was the first to 

 show that on the side of the source of sound toward the point 

 from which the wind was blowing, the sound-waves would be 

 flattened and the sound "ray" reflected upward, so that 

 sounds would not be heard at distances as great as on the 

 other side, where the effect is to make the wave more con- 

 vex and to deflect the " ray " downward. This explanation is 

 in harmony with many observed facts, and especially one 

 noticed by Henry in 1865 namely, that a sound moving 

 against the wind, and inaudible to the ear on the deck of a 

 schooner, became audible on ascending to the mast-head. 

 This observation had suggested the idea that sound was more 

 readily conveyed by the upper current of air than by the 

 lower. Some very important practical conclusions came from 

 this principle of wind refraction, one being that a continuous 

 sound, as from a horn or whistle, would be less likely to be 

 lost by refraction due to adverse winds than sounds of prac- 

 tically a single impulse, as from a bell or gun ; also that it is 

 more probable that sounds of a high pitch will be more inter- 

 fered with by refraction than those of medium tones or lower 

 pitch. 



Refraction from inequality of the temperature of horizontal 

 layers is also explained at length. This may be of a nature 

 to deflect sound rays up or down, according as the lower or 

 upper stratum of air involved is at the highest temperature. 

 Variations of temperature in the atmosphere along vertical 

 lines are shown to be sufficient to account for many acoustic 

 phenomena with which all observing people are more or less 

 familiar. Among them may be mentioned the extraordinary 

 distance at which sounds may be sometimes heard, especially 



