TOWER, PRODUCTION OF SOUXD IX CERTAIN FISHES 169 



is especially well demonstrated in the tracings where the register of the 

 muscular contractions in an animal with the bladder intact is placed directly 

 over these from an animal in which the bladder is collapsed (PI. VIII, fig. 4). 



Experiment XVIII. — The viscera, including the swim-bladder, were removed 

 from a squeteague after an incision had been made in the mid-ventral line from the 

 pectoral fins to the anus. The wire hook of the registering apparatus was inserted 

 into the middle of the central tendon. No noise was produced. The number of 

 contractions was 24 per second. The amplitude of vibration was less than some 

 registered by the muscle and more than others. The experiment revealed no new 

 factor. 



In the experiments just described each contraction of the muscle, repre- 

 sented in the tracing by the apex of the curves, is simultaneous with the 

 sound produced, and thus the rapid series of contractions institute the roll 

 or "drumming." 



Experiments to determine the pressure of the gas in the swim- 

 bladder. 



Experiments were made to discover, if possible, the pressure exerted on 

 the air-bladder by the contraction of the drumming muscles. 



Experiment XIX. — The pressure of gas in the air-bladder of a female squeteague 

 (which has no drumming muscles and can not drum) was determined by making an 

 incision one inch long in the mid-ventral line two inches anterior to the anal fin. 

 The posterior end of the swim-bladder was ligatured and then amputated just back 

 of the ligature. The open end of a small mercurial manometer was inserted and tied 

 by another ligature. The first ligature was then removed and the mercury rose to a 

 height of 4 mm., which was produced by the normal pressure of the gas in the air- 

 bladder. The animal was kept alive by artificial respiration. 



a. The same experiment was then tried on the swim bladder of a male sque- 

 teague, both while it was quiet and while it was drmnming. In the quiet animal, the 

 pressure rose to 4 nmi. and remained there until drumming occurred, when it rose to 

 6 mm. In other words, the increased pressure brought about by the contraction of 

 the drumming muscles equalled 2 mm. of mercury. During the di'umming the 

 meniscus of the mercury could be seen to oscillate between 4 mm. and 6 mm., as the 

 muscles successively contracted and then relaxed. 



One interesting feature is that in all the animals examined the normal 

 pressure in the bladder was 4 mm. in the male and female — the large and 

 small animals alike. The gas pressure within the swim-bladder maintains 

 a tension on the elastic walls, while the increased density of the gas due to 

 the pressure tends to produce a louder sound than would othersvise occur. 

 These experiments show that 



