408 Donald R. Griffin 



frequency, but the fish responded just as well after the inner ear had been exposed 

 and the semi-circular canals plucked out with forceps — a procedure which seemed 

 likely to damage the organs of hearing. Hence Kreidl concluded that goldfish 

 detect underwater sound not by means of the inner ear but through mechanical 

 receptors of the lateral fine or skin. 



BiGELOW repeated Kreidl's experiments, working under the guidance of the late 

 G. H. Parker, who had just demonstrated the auditory sensitivity of Fundulus, and he 

 extended them with a simple elegance of experimental design which merits the 

 respectful consideration of current investigators. Having found that his goldfish 

 responded more than 80% of the time when stimulated by a 100 c.p.s. tuning fork, 

 he cut virtually the entire sensory nerve supply to the skin and lateral line, and still 

 observed almost the same percentage of responses. Yet when both eighth cranial 

 nerves were cut there were no responses in 73 trials with seven different goldfish. 

 As a control against side effects of exposing and cutting the auditory nerve, both 

 eighth nerves were exposed but only one was cut; this fish still responded to sound, 

 but ceased to do so when the remaining auditory nerve was severed. In another 

 control the spinal cord, lateral fine nerves, and the cutaneous branches of cranial 

 nerves V and VII were cut without effect on the response to sound. Bigelow went 

 on to repeat Kreidl's operation, and he obtained the same result. But he also found 

 from histological sections that plucking out the semi-circular canals left much of the 

 sacculus and lagena intact, these being the portions of the ear that would be expected 

 to play the major role in responses to sound. Q.E.D.: goldfish could hear. 



Like many biological problems the question of hearing in fish underwent a long 

 subsequent history of complications and controversies, and a generation later an 

 active debate still centred around the simple question, " can fish hear, and if so, 

 what ? " (For complete reviews see von Frisch, 1936, and Kleerekoper and 

 Chagnon, 1954). In many of the earlier experiments spontaneous responses were 

 obtained from fish when vibrations of various frequencies and intensities were 

 imparted to the water by a wide variety of methods. In other cases fish were con- 

 ditioned or trained to give responses when the water around them was set into oscilla- 

 tion, and considerable evidence pointed to the importance of the lateral line and the 

 skin, especially for low frequencies (Parker and van Heusen, 1917). The methods 

 used to generate underwater sound in these earUer experiments were (1) to generate 

 the sound in air near the tank containing the fish, (2) to place a vibrating object such 

 as a tuning fork in contact with the tank, or (3) to immerse a buzzer or telephone 

 receiver in the water itself. In all cases the stimulus undoubtedly set the water into 

 vibrations with a complex frequency spectrum, but the component most obvious to 

 the experimenter may not have been the one which was most effective in stimulating 

 the fish. This situation was greatly clarified by the careful work of von Frisch and 

 his associates from 1929 to 1941; and although most of the experiments employed 

 freshwater fish, important general conclusions were reached which can be applied 

 directly to marine species, as has recently been demonstrated by Dijkgraaf (1952). 



Stetter reported in 1929 that two common freshwater fish, (the European minnow, 

 Phoxinus laevis, and the common catfish, Ameiurus nebulosus) could be trained to 

 come for food to a particular part of the tank (or to give other characteristic feeding 

 movements) when a sound was produced in the air at some distance from the aquar- 

 ium. The fish were blinded to prevent them from responding to visual rather than 



