08 BULLETIN OF THE 



was first exhibited by Sondhauss, in 1852, by means of a lens of 

 carbonic acid gas confined in an envelope of gold-beater's skin, 

 or preferably, of collodion film. The wave-fronts of sound (con- 

 sidered as practically plane surfaces) being centrally retarded in 

 passing the convex surface of the lens, thus move through and 

 emerge from the lens w^ith concave surfaces, whose normals con- 

 verge to a point. A convex lens of hydrogen would cause the 

 wave-fronts of sound passing through it to emerge (by accelera- 

 tion) with a convex form — that is, would cause the rays or nor- 

 mals to diverge — the focus being negative. And to obtain a 

 positive focus of convergence it would be necessary to make a 

 hydrogen lens concave. The same effect would be produced on 

 light by a lens of hydrogen. 



2. The refraction of sound resulting from differences of motion 

 in the air vv^as first suggested by Prof Stokes, in 1851. As the 

 advance of sound is always in directions normal to the expand- 

 ing spheroidal surfaces of instantaneous compression, any defor- 

 mation of these spheroidal surfaces must correspondingly deflect 

 the line of successive impacts from the original radial direction. 

 Winds being usually considerably retarded near the surface of the 

 earth by frictional resistance, the wave-front of sound moving in 

 the direction of the wind, is more advanced above than below, 

 causing the sensible rays to bend downward ; on the other hand, 

 the wave-front is more retarded above than below when opposed 

 to the wind, causing the rays to bend upward. Prof. Henry was 

 the first to observe (in 1865) that sound moving against the wind 

 could be heard aloft after it had ceased to be audible below, 

 though it was not till some time afterward that he detected the 

 true cause. Prof Reynolds subsequently (in 1874) independently 

 verified by experiment the theory of Prof Stokes. 



3. The refraction of sound, resulting from differences of tem- 

 perature, was first pointed out by Prof. Reynolds in 1874, who 

 has shown that various recorded observations on sound very 

 strikingly establish the indications of theory in this direction. 



These last two conditions of acoustic refraction — inequality of 

 motion in the air and inequality of its temperature — are both 

 susceptible of very simple quantitative determination, and are 

 thus shown to be real and efficient causes of many observed re- 

 sults, and to furnish satisfactory explanations of many hitherto 

 puzzling phenomena of sound. Yarious illustrations were given. 



Mr. Button spoke of the rumbling sound of trains on the Alex- 

 andria Railroad, as heard at the U. S. Arsenal in this city; some- 

 times loud, sometimes moderate, sometimes inaudible; much 

 louder with favorable than with adverse winds ; sometimes, espe- 

 cially if the air was still, the sound, while inaudible at the ground, 

 was distinct, or even loud, at an elevation of 40 feet. 



