TOWER, PRODUCTION OF SOUND IN CERTAIN FISHES 161 



ments. In the first, he inserted his finger into the stomach of a lyre capa- 

 ble of producing an intense noise. During the production of the noise, 

 he noticed an intense vibration which coincided exactly in duration -with 

 the sounds heard by his ear. He then punctured the wall of the air-blad- 

 der and drew out all the gas. The sound ceased, but the vibrations could 

 still be felt. He then removed the entire swim-bladder, and applied his 

 finger successively to the muscles and aponeuroses which lie alongside the 

 vertebral column, and he found that all the organs were in repose except 

 the intra-costal muscle, which vibrated and gave to his finger the same sensa- 

 tion as when the air-bladder was in its natural position. 



In the second experiment, Dufosse opened the abdomen of a lyre just 

 in front of the anus, and extirpated the swim-bladder entirely ("j'extirpe 

 la vessie pneumatique tout entiere"). He then inserted an artificial blad- 

 der ("poche membraneuse") and inflated it. The fish commenced again 

 to produce a noise similar to that made before the operation. In another 

 he cut the nerve supplying the intra-costal muscles, first the right and then 

 the left. After both were cut, the noise ceased and could not be again 

 renewed. 



From these data, Dufosse argues that there are tv\'o factors in the pro- 

 ducing of the noise, \t1z: the contraction of the intra-costal muscle, which 

 is the primary cause of the sound, and secondarily, the reenforcement of 

 these vibrations of the swim-bladder. The producing of the noise is volun- 

 tary. Dufosse recognizes many difficulties in this explanation, however, 

 because the facts do not agree with those of theoretical physics, as can be 

 seen from the following quotation: — 



" Le mecanisme de la production des sons chez ces poissons a pour complement 

 la transmission des \abrations sonores des muscles a la vessie qui est en contract avec 

 eux. Les parois de est organe commvmiquent ces vibrations au gaz qu'elle renferme, 

 et ceux-ci ^'ibrent de telle fagon comme le prouvent surabondamment mes deux 

 premieres experiences, que I'intensite de ces vibrations est incomparablement aug- 

 mentee. D' apres ce resultat et en considerant que la vessie est une cavite close a 

 parois membraneuses et souples se raoulant si exactement sur la surface des organes 

 qui les enviromnent qu'elles ne peuvent vibrer que comme elles le feraient si elles 

 ^taient reellement adherentes par tous les points de leur superficie a la masse de ces 

 organes on ne pent expliquer, conformement aux principes dc la physique, le ren- 

 forcement des vibrations sonores qu'en admettant que le volume des gaz contenus 

 dans vessie, ou, ce qui est le meme chose, que la capacite de cet organe a naturelle- 

 ment des rapports exacts de grandeur avec cells des nombres de vibrations sonores 

 que lui sont transmises. L'exactitude des rappoits que suppose cette explication 

 ne s'accordant pas avec plusieurs faits ichthyologiques, entre autres avec les inces- 

 sants changements de volume que submit necessairement la vessie pneumatique 

 quand le poisson vient du fond de I'eau a la surface ou s'enfonce dans la profandeur 

 des mers, cette explication n'est acceptable qu'en admettant que si ces rapports 



