82 M. Savart on the Improvement of the Pipes of Organs. 



of a very small pipe of the same length as the side of the cube, 

 by a series of sounds graduated as we desire, and which will 

 be given by pipes of the same length, but whose magnitude 

 gradually decreases. I have verified this result with ten pris- 

 matic square pipes, three inches long, whose sizes were such 

 that they described diatonically an octave and a half, com- 

 mencing with the cube which gave the sound sol ? , and terminat- 

 ing with the sounds 5 , which was emitted by the smallest pipe. 



We may also save room in organs by taking advantage of 

 the fact, that prismatic square tubes may be indefinitely di- 

 minished in thickness, by the approximation of their sides. 

 It does not appear to me that the sound undergoes any sensi- 

 ble alteration by diminishing their thickness one-half, and even 

 two-thirds. 



There is some reason for thinking that we do not obtain 

 the best possible result by placing the embouchure at one of 

 the extremities of the pipe, as is usually done. Nay, there 

 might even be some advantage in placing it in the middle of 

 its length. I have constructed several pipes in this way, and 

 they yielded a sound of a very fine quality, whether they 

 were open or shut at both ends. I have observed also, that 

 in placing the opening on the side, and at one end, as in 

 flutes, the sound has a very agreeable timbre, which organ 

 pipes seldom possess. 



We may likewise construct flat pipes (made to vibrate by 

 their tranche) of an infinity of powers, cylindrical, triangular, 

 and elliptical ; and I am convinced, by experiment, that we 

 can obtain from them very fine sounds. The simple law of 

 the number of vibrations for pipes of a similar form, allows 

 us to determine, with facility, the dimensions which each pipe 

 requires to give out any sound of the gamut. 



Very short pipes, in which all the plates of air perpen- 

 dicular to the embouchure are not actuated by the same 

 kind of motion, do not appear to be susceptible of yielding 

 agreeable sounds. Spheres of air, for example, yield sounds 

 very dull, and approaching to a noise, without respect to the 

 impression which they make upon the ear. The same thing 

 takes place with cubes, having their embouchure in the mid- 

 dle of one of their faces. This quality of sound may de- 



