414 Donald R. Griffin 



It is pertinent to recall in this connection that the critical evidence that bats (and 

 at least one species of bird) employ echolocation is not alone their production of 

 sound but, more important, their ability to fly without seeing and the disorientation 

 caused by impairment of hearing or sound production (Griffin and Galambos, 1941 ; 

 Griffin, 1953 b). Comparable experiments are obviously more difficult with cetaceans, 

 but observations of Winded porpoises, or even the use of intense interfering noises, 

 should permit a resolution of this question. It is even possible that porpoises use 

 fainter sounds for echolocation than those studied to date, and that more sensitive 

 apparatus, or improved signal-to-noise ratios, would disclose a more continuous 

 emission under conditions where vision is restricted. The most intense high-frequency 

 sounds of bats are not the pulses used for echolocation, but the noisy cries of much 

 longer duration (which are audible owing to minor low frequency components even 

 though most of their energy is at frequencies above 20 kc). Prolonged listening to 

 bats with relatively insensitive equipment in the presence of moderately high noise 

 levels at ultrasonic frequencies might well have led to a picture of their sounds not 

 unlike our present information that porpoise sounds are numerous, loud, and varied, 

 but often not detectible when needed for acoustic orientation. This comparison is 

 especially relevant for those species of neotropical bats that feed on fruit and orient 

 themselves adequately by means of faint pulses which can easily be overlooked even 

 with reasonably adequate apparatus (Griffin and Novick, 1955). All of these 

 considerations warrant an open mind regarding the role of acoustic orientation in 

 marine mammals, and further careful investigation of the matter is clearly called for. 



Discussion 



In most considerations of the possibility that marine animals orient themselves 

 by echolocation it has been assumed that they would use an analogue of pulse sonar, 

 emitting sounds of very short duration and hearing echoes arriving in the silent 

 intervals between the pulses. Pulsed sounds are employed in the echolocation of 

 all the bats studied to date, although in bats of the family Rhinolophidae the pulses 

 may last about 100 msec, so that there must be considerable overlap between out- 

 going sound and returning echo (Mohres, 1953). While some of the sounds emitted 

 by fish and cetaceans consist of rapidly repeated clicks, others are continuous tones 

 or noises ; and since sound travels faster in water than in air most of the known fish 

 and porpoise sounds seem poorly suited for echolocation owing to the inevitable 

 mixing of echoes with the outgoing sound. In this regard " echo fish " discussed 

 above, and certain ticking sounds of uncertain origin recorded during Woods Hole 

 Oceanographic Institution cruises, seem to be exceptions to the general rule. It is 

 quite possible, however, for echolocation to be based upon continuous sound, but 

 since the echoes will almost always be fainter than the outgoing sound some special 

 means must be employed to discriminate between the two — frequency as in the case of 

 frequency modulated radar, or the high degree of isolation of the ear from the source 

 of the emitted sound that Mohres postulates for the bats of the family Rhinolo- 

 phidae, which use 100 msec pulses to detect objects at such close range that echoes 

 begin to return in 5-10 msec. 



There is another aspect of underwater sounds which might be related to echo- 

 location, and which has not been thoroughly considered. The wave lengths of the 

 sounds emitted by most fish, and those to which they are most sensitive, range from 



