taneously at the beginning and end of each trial. Figure 5 shows the 

 gates down. In Figure h they are all raised and held in position as 

 during a sound or control run„ The pond in which this structure was 

 built is U50 feet long and 60 feet wide. The bottom and sides are 

 almost pure marl mud, a carbonate, which absorbed sound very well. 

 The structure was placed in the center of the pond to avoid standing or 

 echo-waves of sound from interfering with the beamed signal. 



In developing the method of investigation and in designing the 

 trough with its gates and sections to measure the reaction of fish to 

 soundj three basic assumptions were made: (1) if the trout were 

 unaffected by or indifferent to sound waves, they would move within 

 the trough and between sections in a pattern similar to that of the 

 control, (2) if they were attracted by some frequency of sovind they 

 would tend to proceed to the end of the trough nearest the sound 

 soxirce or (3) if the fish were frightened by the sound they would 

 travel away from the sound source„ Obviously, it was assumed that 

 other stimuli had no effect upon the movement of the trout. Insofar 

 as possible precautions against such extraneous stimulation were taken. 



"Water ham^I'^R" - electrO"Iuc]I\ietic transducer 



For the first tests with low audible frequencies, the 600 pound 

 audio speaker (Figure 6) was suspended in the water so that the round 

 aluminum piston (lower center) was approximately 1-1,''2 feet below the 

 water surface, and 2=l/2 feet above the bottom. The heavy framework 

 to support it was erected at the extreme end of the trough outside 

 section No, 1. The greatest intensity of sound, therefore, was in 

 section No. 1. The least intensity was found in section No. 10, as 

 determined by hydrophones. The speaker was never moved from its posi- 

 tion adjacent to section No. 1, predominantly because of the enveloping, 

 non-directional nature of its sound pattern. The unit, designed as a 

 sonic mine sweep, had a frequency range of 6? cycles per second to 

 3,500 cycles per second utilizing 110 volts D. C, at 3 amps. In order 

 to use the full potential of the speaker, it was necessary to connect 

 two 125 vjatt amplifiers in tandem to obtain 200 watts power. A signal 

 generatox- or oscillator completed the lovj frequency equipment (Figure 7). 

 It might be noted that 6? to 3,500 cycles per second is roughly the 

 range of a piano. At a frequency of 60 cps. to 2_>,000 cps. at 3,000 

 dynes/cm^ , the intensity was 70 decibels above 1 dyne/cm^ at a distance 

 of 3 feet, rising to 12,000 dynes/cm^, or 82 db above 1 dyne/cm^, at 

 2,500 to 3,500 cps. 



The complex interference patterns caused by reflection from the 

 surface and bottom prevented a uniform fall-off of the sound along the 

 length of the trap structure and also prevented accurate measurements 

 of the soun<i field. 



