THE LANGUAGE OF ECHOES 



of sound difficult, and so echolocation seems an unlikely 

 explanation. 



As mentioned earlier in connection with the underwa- 

 ter hearing of fish and porpoises, sound waves have great 

 difficulty in passing from air to water or vice versa. When 

 airborne sound impinges on a smooth surface of water, 

 with its direction of travel perpendicular to the water 

 surface, only 0.12 per cent of the energy of the airborne 

 sound continues beneath the surface as underwater 

 sound waves. For a sound wave travehng from water to 

 air, the same small fraction of the acoustic energy strik- 

 ing the surface from below continues outwards into the 

 air. This means that a flying bat's orientation sounds 

 striking the water, penetrating into it, being reflected 

 back from a fish, and passing out into the air again 

 would be reduced to (0.0012)2, or 1.44 x 10"^ of the 

 original sound, during the two trips through the air- 

 water interface. To this great reduction must be added 

 further losses: only a small fraction of the emitted 

 sound would be reflected by a fish, and only a small 

 fraction of what did escape into the air would strike 

 the ears of the listening bat. These figures make it 

 seem almost hopeless for a bat to try to detect fish 

 through the water surface by their echoes, but before 

 dismissing the whole idea as utterly impossible, let us 

 compare what insect-eating bats are known to do in air 

 with the hypothetical location of fish by their echoes. 



Certain of the FM bats are able to detect a pebble or 

 a flying insect 1 centimeter in diameter from at least 200 

 centimeters away. At distances of more than about 10 

 centimeters from the bat's mouth the sound intensity falls 

 off as the square of the distance. Since a 1 -centimeter 

 insect is a small target, sound is scattered from it ap- 

 proximately as it would be from a point source, so that 

 the echo intensity also varies inversely as the square of 



97 



