Hearing and acoustic orientation in marine animals 409 



auditory stimuli associated with feeding. Clear responses were obtained from sounds 

 generated by tuning forks, whistles, and pipes, even when the frequency was as 

 high as 6,960 c.p.s. with the minnows, and 13,139 c.p.s. with the catfish. 



Von Frisch and Stetter (1932) later experimented with the cITects of sensory 

 impairments on the abihty of minnows to respond to various frequencies. Damage 

 to the utriculus and semicircular canals caused little reduction in sensitivity to sounds, 

 although equilibrium was severely impaired. But when the sacculus and lagena were 

 put out of action the sensitivity to frequencies above 150 c.p.s. was almost totally 

 destroyed. Even frequencies as low as 32 c.p.s. were less elTective in arousing res- 

 ponses than when the sacculus and lagena were intact. Very low frequencies such as 

 16 c.p.s., however, were almost as easily perceived as they had been before the opera- 

 tion. Nor did impairment of the lateral line receptors reduce the fishes" sensitivity 

 to low frequencies. These investigators therefore concluded that low frequencies 

 must be detected through very sensitive tactile receptors in the skin (see also Rein- 

 HARDT, 1935). 



Absolute thresholds could not be determined in these experiments, because no 

 cahbrated underwater transducers were available for the purpose. In an attempt 

 to obtain approximate thresholds, the source of the stimulating sound was moved 

 to greater and greater distances from the aquarium, and the threshold of response 

 of the fish was compared with the threshold of hearing of men standing beside the 

 aquarium. The fish ceased to respond at levels somewhat above the human threshold; 

 but when a man was held entirely underwater in a large aquarium, he was unable to 

 hear the sound at intensities to which the minnows would still react. 



In later experiments at the same laboratory by Boutteville (1935) and Diessel- 

 HORST (1938) the sound was generated by a loudspeaker close to the aquarium, and 

 its intensity in the air was measured by means of a calibrated microphone. The thres- 

 hold of response of a fish could thus be expressed in terms of sound pressure in the 

 air just outside of the glass-walled aquarium. The most sensitive of the minnows 

 gave consistent responses to 652 c.p.s. when the sound level in the air was 20 decibels 

 above the customary reference level of 0-0002 dynes/cm% or ten times the threshold 

 sound pressure for a typical human Ustener. This is a remarkable sensitivity when one 

 considers that the energy loss was unquestionably great as the sound waves passed 

 from air to water. I once duplicated approximately the acoustic conditions of this 

 type of experiment and measured sound pressures inside the aquarium with a cali- 

 brated hydrophone while those in the air were measured with a calibrated condenser 

 microphone. The sound pressures in air and water were roughly the same (within 

 5 to 10 decibels). But the greater acoustic impedance of water causes a given sound 

 pressure to correspond in water to a 35 db lower energy flux (watts cm') than in air. 

 This means that the minnows studied by von Frisch and his associates had auditory 

 thresholds of the same order of magnitude, in terms of energy flux, as the human 



auditory threshold in air. 



When a number of species were compared with the minnow Phoxinus with respca 

 to their sensitivity and frequency range of hearing, it became clear that lish can be 

 divided into two distinct groups. The minnow Phoxinus. the catfish Ame.urus, and 

 certain other fish, display low thresholds and a wide frequency ^'^^^'-^'^'^^^''^''^^ 

 fishes studied were less sensitive and responded only to frequences below 1000 .000 

 c.p.s. This difference is correlated with the anatomy of the fish : for the more sensitive 



