THE SENSE OF HEARING IN FISHES. 91 



importance in the determination of an experimental result. Hence 

 it is not surprising that Haempel's outcome on Cyprimis, Scardinms, 

 Gobio, and Trutta should have been negative, for he states (1911, p. 

 320) at the outset that movements of the pectoral fins, of the caudal 

 fin, and of the respiratory apparatus, however called forth, are not 

 accepted by him as evidences of sound stimulation. To any one 

 familiar with the responses of fish such a declaration must seem to 

 say the least, arbitrary and condemns without further ado any nega- 

 tive results that its author might claim. Such movements are often 

 most characteristic and significant and they call for close scrutiny 

 and careful observations. Although they can be seen clearly and 

 beyond question when the fishes are in aquaria, they would very 

 probably escape attention when these creatures are at some distance 

 in open water. In consequence it seems doubtful if negative results 

 recorded under these conditions (Bernoulli, 1910, p. 640) can be 

 said to be well grounded. 



From the dbservations of Parker and Van Heusen (1917, p. 

 477), it is clear that Amiurus is by no means equally responsive to 

 tones of dififerent pitches. It responded with greatest certainty to 

 tones of 43 complete vibrations per second, and with less and less 

 certainty to succeeding octaves up to 688. It failed entirely to re- 

 spond to the two tones above 688, namely 1,376 and 2,752. It is, 

 therefore, clear that Amiurus is much more receptive to tones of a 

 low pitch than to those of a high pitch. Since most of the sounds 

 produced by fishes are of low pitch, being described usually as 

 croaking, grunting, or drumming sounds, it is pro'bable that fishes 

 are adapted chiefly to this class of tones. It is, therefore, not im- 

 possible that many tests that have yielded negative results may have 

 done so because the tones employed were too high in pitch for the 

 fishes. This may have been the case in the sound from the " cri- 

 cri" employed by Korner (1905, 1916) and with that from the elec- 

 tric bells used by Maier (1909), by Bernoulli (1910), and by Haem- 

 pel (1911'). If the sounds thus produced were out of range for the 

 fishes, it is not to be expected that they would react. All such 

 tests, therefore, that have yielded negative results are open to this 

 objection until doubt on this point has been removed. Thus the 

 negative evidence of practically all the recent workers on this sub- 



