EFFECT OF THE CA VITIES UPON THE VOICE. 1235 



interesting to observe that in Greek, when two such consonants as k and 

 g occur together, nasal resonance is added to one of them, e.g. svyyaXu, 

 pronounced sunkalo. If, ch, y, and the burring r are also produced at 

 this position. 



It is only by a proper appreciation of the method of production of the 

 consonants that Grimm's law * of the permutations of these letters can be 

 thoroughly understood. Thus /rater in the Latin is represented by brother 

 in the Low-German, and ^>ruoder in the Old High-German. It is not difficult 

 to see why this should be, since /, 5, and p are produced by very similar 

 muscular movements. So also the Greek (xyv) is represented by the Low- 

 German "goose," and the Old High-German "Kans," all three initial con- 

 sonants being produced by very similar muscular mechanisms. 



The acoustic analysis of the consonant sounds has not yet been 

 undertaken, and it is clear that considerable difficulties must be 

 encountered in such an investigation. Tracings representing the 

 consonant sounds have been obtained by Blake, 2 but for the purposes 

 of analysis these are not satisfactory. 



EFFECT OF THE CAVITIES IN CONNECTION WITH THE LARYNX AND 

 PHARYNX UPON THE VOICE. 



It needs but little consideration to understand that the various 

 cavities, which are in more or less direct connection with the mouth 

 and larynx, must have at least some effect upon the quality of the voice. 

 It is, however, difficult to say exactly what these effects are. 



Moura 3 has attempted to show that the trachea has very consider- 

 able influence upon the pitch of the voice, but it is clear, from physical 

 considerations, that he attributes too much importance to the phonetic 

 role of the trachea. In the first place, its proper tone is far removed 

 from the lower notes of the human voice ; and, secondly, although it is 

 a more or less rigid structure, its elasticity must be small when com- 

 pared with the rigid character of substances used as resonators, especially 

 since its inner surface is covered with mucous secretion. Nevertheless, 

 it is probable that the trachea may resonate to some of the notes of the 

 voice, and thus produce an effect upon their quality ; the pitch, on the 

 other hand, is determined by the vocal cords themselves. 



It is doubtful if the laryngeal ventricles have any effect upon the 

 voice in the human subject, on account of their small size. In some 

 animals, such as the howling monkey, the orang, and the horse, it is possible 

 that they may act as cavities reinforcing the voice. The loud deep roar 

 of the lion, on the other hand, comes from a larynx in which the 

 ventricles are absent. 4 In the human subject the laryngeal ventricles 

 probably perform a function unconnected with the voice. Wyllie 5 has 

 shown, on the dead subject, that if the false cords be partially approxi- 

 mated, and a blast of air be driven up against them from below, the 

 effect upon these structures is to drive them further inwards, and hence 

 close the glottis more completely in a valve-like fashion. The action, 

 indeed, is somewhat similar to that which occurs when the mitral valve 



1 Morris's " Historical English Grammar," 1882, p. 46. 



2 Blake, Am. Journ. OtoL, K Y., July 1879. 



3 Paris Soc. d. Laryngol., RhinoL, etc., March 1893. 



4 Wolff, "Diss. Anat.," p. 9. 5 Edin. Med. Journ., 1866, vol. xii. p. 214. 



