The lower or downstream end of the experimental area was divided by a 

 plywood vane into two channels with entrances 3 feet and 7 feet wide respectively. 

 Besides directing the flow of water around the island, the plywood vanes served to 

 separate the fish that were diverted by the electrical field into the narrow channel 

 from those that passed through the electrical field into the wide channel. Baffle-type 

 traps constructed from 1/4-inch-mesh galvanized hardware cloth collected the fish 

 in the two channels. 



The electrical barrier was created by two parallel rows of electrodes suspended 

 from wires stretched across the experimental area. The parallel wires were adjust- 

 able at angles at 40°, 60°, and 90° in relation to the long axis of the experimental are4, 

 the wires could be spaced 2 or 3 feet apart (fig. 2). The distance between these wires 

 is designated as the width of the electrical field. Electrodes of hollow aluminum tubes 

 were fastened to each wire by slim -nosed alligator clips at 6, 12, 24, and 36-inch 

 spacings: they were suspended in the water to within 3/4 inch of the bottom of the tank. 

 For a comparison, electrodes of 1/2- and 2 -inch outside diameters were used. A 

 pulsating direct current with a square wave form was supplied to the two rows of 

 electrodes with the positive row upstream. To eliminate the possibility of a visual 

 leading effect, two parallel rows of control electrodes suspended from nonconductive 

 material were placed opposite the rows of electrodes which were electrified; both 

 sets of electrodes were in the water at the same time. Figure 2 illustrates a typical 

 arrangement of bor'i test and control electrodes used in the experiments. 



Light was supplied by four 500-watt lamps spaced uniformly over the tank. A 

 variable auto-transform.er controlled the light intensity between 3.4 foot -candles and 

 less than 1 foot-candle. During the tests the light intensity was reduced in order to 

 stimulate a downstream movement of the salmon finger lings. Under maximum light 

 intensity the fish tended to school in the experimental area and any attempt to force 

 them downstream resulted in startled swimming movements. All changes in intensity 

 were made very giadually to avoid startling the fish. 



A pulsating square -wave direct current was supplied to the barrier with the 

 following characteristics: pulse frequency 8 pulses per second, pulse duration 40 

 milliseconds, and voltage gradient 1 volt per centimeter. These electrical character- 

 istics and energy levels were found to be effective in the preliminary experiments of 

 Collins, Volz, and Lander (unpublished manuscript) . 



The total voltage was measured with a standard RCA WO-56-A oscilloscope; 

 the voltage gradient was calculated from the total voltage and the distance between 

 the parallel rows of electrodes. The voltage gradient represents an average value 

 since the electrical field resulting from the tubular electrodes was not uniform. 

 The actual voltage gradients were measured with a probe (fig. 3). The lines of 

 equal potential resulting from one arrangement of electrodes are shown in figure 4. 



