M. Savart on the Mechanism of the Human Voice. 203 



there is one which is the same as one of those which the 

 column of air may give. This result, deducible from theory, 

 is confirmed by experiment. 



The sounds obtained by this method have a particular 

 character, distinct from those of all ordinary organ-pipes. 

 They may become very intense and very loud, especially 

 when the apparatus is made of metal, and the dimensions of the 

 eolumns of air properly chosen. 



This instrument, like organ-pipes open at both ends, can 

 only give the series of sounds ut x ut 2 sol 2 ut 3 mi 3 $ol 3 la% s 

 •ut± &c. It may, nevertheless, happen, that the small vessel 

 which serves as the mouth-piece may sound independently of 

 the column of air ; but then the sounds are feeble, and want 

 distinctness. From what has been now said, it may be easy 

 to conceive, that a tube like Fig. 7, composed of elastic mate- 

 rials, may give out all possible sounds comprised within cer- 

 tain limits, depending on the tension of the sides and the vo- 

 lume of air. 



When the pipe, in which the air sounds is pierced with late- 

 ral holes, if we blow uniformly by the portvent, the sound 

 may be varied when the holes are shut and vice versa, so that 

 it would not be impossible to construct a musical instrument 

 upon this principle. 



The fundamental sound of a pipe shut up at one end, and 

 of an uniform diameter, is in general more grave by an oc- 

 tave than the sound which the same pipe gives when it is 

 open at both ends. This, however, is not the case with pipes 

 of unequal diameter, such as conical and pyramidal ones, 

 when they are made to_vibrate at their narrowest part. In 

 these the interval, between their sounds when open and shut, 

 becomes as much greater, for an equal length of tube, as the 

 inclination of its sides increases. A conical pipe 4* inches 

 long, and truncated at its summit, and having its larger end 2 

 inches in diameter, and its smaller one six lines, gives, when 

 open, the sound ut 5 , and when shut mi 3 . By widening its 

 large end, or lessening its small one, the other dimensions re- 

 maining the same, the sound may be lowered even more than 

 two octaves. 



In order to determine the exact form of the larynx, I took 



