CHAP, xviii.] SAVABT'S EXPERIMENTS. 93 



of the excursions of its vibrations. The contraction of the stapedius 

 muscle causes the base of the stapes to compress the membrane of the 

 fenestra ovalis to a greater or less extent, so that the degree of ten- 

 sion of that membrane depends on the condition of this muscle. Com- 

 pression exerted upon the membrane of the fenestra ovalis extends 

 to the perilymph and through it is propagated to the membrane 

 of the fenestra rotunda, and in this way the same apparatus which 

 regulates the tension of the membrane of the fenestra ovalis performs 

 that office for that of the fenestra rotunda, and Savart has devised 

 an apparatus which very prettily illustrates the manner in which 

 this may take place. In a disc of wood (a, b, fig. 148) of sufficient 

 thickness, he hollows out two cavities, 

 o and r, which communicate at their bot- 

 toms with each other by a narrow canal 

 (c) hollowed in the wood, but not open on 

 its surface ; a thin membrane is extended 

 over each of the cavities. Thus, the air con- 

 tained in these cavities may easily pass from 

 one to the other, and may always maintain 

 the same degree of elastic tension in both. If a vibrating glass be 

 brought near the membrane r, covered with a layer of sand, it will 

 be found to enter freely into vibration, as evinced by the active 

 movements of the grains of sand. If, now, pressure be made on o 

 with the finger, r will become convex in proportion as o is rendered 

 concave by the pressure, and when in this convex state, the move- 

 ments of the sand upon it will be much less considerable than 

 before, presenting an effect precisely similar to that produced on the 

 tympanic membrane by an increase of tension. Thus, the extent 

 of the excursions of the vibrations of the membrane r, is limited by 

 the pressure exerted upon o 3 and as the membranes of the two fenes- 

 trse are related to each other in an analogous manner, we may argue 

 that pressure upon the membrane of the fenestra ovalis, will occasion 

 tension of that of the fenestra rotunda, thereby limiting the extent 

 of the excursions of its vibrations.* 



Moreover it appears, upon reference to the anatomy of these 

 parts, that the only muscles that have been satisfactorily demon- 

 strated are tensors of the tympanum ; and that at whichever extre- 

 mity of the chain of ossicles muscular effort be first exerted, a cor- 

 responding effect will be produced at the other ; that when the sta- 

 pedius muscle acts, the malleus is thrown into a position favourable 



* All these experiments have been frequently repeated and exhibited by us, 

 with the same results as those stated in the text. 



