368 
PACIFIC SCIENCE, Vol. XXI, July 1967 
Fig. 2. Anatomy of the sound-producing mech- 
anism and surrounding bones in Myripristis berndti. 
1 , Main body of swim bladder; 2, sound-producing 
muscle; 3, dorsal portion of clei thrum bone with 
tendon attached to sound-producing muscle; 4, ante- 
rior lobe of swim bladder; 5, preoperculum; 6, 
scapula. 
the air bladder. A small tendon connected the 
muscle to the cleithrum bone laterally. Another 
small, flat muscle (not shown in the figure) at- 
tached to the skull and ran between the sound - 
producing muscle and the anterior lobe of the 
air bladder, to the operculum. This muscle was 
routinely cut during ablation experiments, with 
no apparent effect on sound production. The 
sound-producing muscle was highly vascularized 
and appeared to be composed of three distinct 
myomeres. 
Removal of one sound-producing muscle 
resulted in a relative decrease of 2-7 db in 
operated fish, when compared with their own 
normal sounds (Table 1). Sound pressures 
were reduced in all octave bands but were 
greatest in the 75-150 cps band. Normal 
sounds contained frequences below 75 cps to 
under 4,800 cps, with most energy between 
300-600 cps. Oscillographs of these sounds 
are shown in Figure 3. 
The temporal patterns of the pulses within 
these sounds are shown in Table 2. Normal 
fish produced sounds composed of 7-10 pulses 
(mean, 8.2). Operated fish showed more vari- 
ability in pulse range (6-11), and a mean 
value of 9.04 pulses per sound. Increases in 
number of pulses were correlated with increases 
in sound duration. Interpulse intervals were 
variable, but in most sounds the intervals be- 
tween the penultimate and the last pulse were 
greater than between other pulses. A few mus- 
cle potentials recorded from two fish were 
composed of 6-8 spikes (Fig. 3). The inter- 
spike intervals and total duration for a series 
of spikes were comparable to values for inter- 
pulse and total-duration measurements of sounds 
with the same number of pulses. 
The effect of removing the superficial mus- 
cles and bones near the sound-producing mus- 
cle is shown in Table 3 and Figure 3. The peak 
pressure of sounds produced by operated fish, 
when compared with pressures of their own 
sounds before the operation, increased in one 
fish, decreased in two, and remained the same 
in two fish. 
Five fish in which both sound-producing 
muscles were removed produced no audible 
sounds. 
The role of the air bladder in sound pro- 
duction was determined by replacing the gas 
in the bladder with sea water. Five fish, in 
which a small hole had been punctured in the 
lobe of the air bladder with a syringe, con- 
tinued to produce sounds at intensities com- 
parable to their own normal sounds (mean 
peak sound pressure =1.1 db above normal 
fish). Only a few bubbles of gas escaped 
through the puncture. When the puncture was 
held open the intensity of the sounds decreased 
as gas escaped until finally, when the air blad- 
der was completely filled with water, no au- 
dible sounds were produced although the 
muscles could still be felt to contract. The 
presence of only a small bubble of gas in the 
bladder resulted in production of sounds of 
very low intensity. 
Field Observations and Diel Recordings 
At least 20 different schools of menpachi 
were found in the Pokai Bay area. In all cases, 
these were mixed assemblages of M. berndti 
and M. argyromus, from 13 to 23 cm in total 
length. In shallower waters, M. argyromus pre- 
dominated. Both species were found to pro- 
duce the same types of sounds and to have 
similar nocturnal-diurnal activity patterns in 
