202 M. Savart on the Mechanism of the Hwnian Voice. 



side by a thin leaf of parchment, the sounds are emitted more 

 easily, and are more grave, more full, and more agreeable. 



When the sides of the orifice are made as in Fig. 3, the 

 sounds are more grave, and less loud. Here the margin of 

 the orifice seems to act like the bevel in organ-pipes. When 

 the margin is rounded as in Fig. 4, the same effect takes 

 place. 



In long organ-pipes, the substance of the pipe has little or 

 no influence on the number of vibrations of the column of 

 air ; but, in short pipes, the substance exerts a powerful in- 

 fluence. If we construct a cubical pipe, with paper or parch- 

 ment stretched over small square frames joined together, to 

 form a cube, the sound which it yields is as acute when the 

 parchment is stretched, as if the sides had been solid ; but if 

 their tension is diminished by humidity, the sound may be 

 made to descend more than two octaves before it ceases to be 

 heard. In the quiet of night there seems to be no limit to 

 this lowering of the sound. By strewing sand on the sides 

 of the cube, they exhibit a nodal, elliptical, or circular line, 

 and the upper and under surfaces vibrate most energetically. 



If we take a prismatic tube nine inches long, and eighteen 

 lines of a side, and form half its length next the embouchure 

 of a membrane, thin, and well stretched, then, though it 

 should give the sound re±, yet it produces much more grave 

 ones, even those between ut 3 and ut±, or even some of those 

 of the octave from ut s to ut 5 . 



The sounds of membranous pipes partake of the quality of 

 those of a flute, and of free mouth-pieces. They have no 

 analogy with those of any musical instrument, and such pipes 

 are in some respects the reverse of stringed-instruments. 

 In the latter, the air in the case is put into vibration by the 

 solid sides which inclose it, while, in membranous pipes, the 

 air is the body which is put directly into motion, and which 

 then communicates its vibrations to the sides which contain it. 



If we fix a portvent at the convex surface of the hunter's 

 whistle, and add a pipe at the other as in Fig. 5, 6, and 7, this 

 combination will give a sound which will be exactly that which 

 corresponds to the column of air contained in the pipe, provided 

 that, among the sounds which the small whistle may give, 





