Water temperature may be a factor contributing to negative 

 results obtained with white suckers in the laboratory under theoretically 

 ideal conditions. The favorable results in the field were obtained at 

 water temperatures of 75° to 78° F. whereas the water in the laboratory 

 ranged from 36° to 39° F. No tests were made at intermediate water tem- 

 peratures. It appears probable that at low temperatures the reduced 

 metabolism of the white sucker limits galvanotaxic response. However, 

 no visible difference was observed in the reactions of rainbow trout at 

 various temperatures between 36° and 63° F. 



The experiments show that the device is selective as to size, 

 and this factor alone restricts its use as a collecting device. Voltage 

 gradients effective for leading rainbow trout vary inversely with the 

 length of the fish. For each size class of rainbow trout, the leading 

 response can be obtained only from voltage gradients that fall within a 

 specific range. The determination of the actual limits of this range for 

 each group is complicated by the variability of the response displayed by 

 test animals of similar lengths. A voltage gradient applied to stimulate 

 the desired movement in large fish would not cause a similar movement in 

 small fish nor be harmful to them. On the other hand, a voltage gradient 

 sufficient to control small fish would narcotize, injure, or even kill 

 any large fish in the electrical field. 



Of all the fish used in this study, only 7 brook trout suffered 

 injury or death directly attributable to exposure to pulsated direct cur- 

 rent. Of these, 6 were killed by a severe dislocation of several vertebrae 

 just anterior to the dorsal fin. In some of these fish, the displacement 

 was such that the vertebrae were separated from the ribs. Holmes (19U8) 

 and Hauck (I9li9) described similar injuries to rainbow trout and other 

 species subjected to alternating current. The remaining brook trout s\xf- 

 fered severe internal hemorrhage and had a large quantity of blood present 

 in the air bladder. 



The field tests demonstrated that the leading device could be 

 operated successfully in a stream and indicated the desirability of fur- 

 ther investigation. A continued study should be directed toward the 

 determination and evaluation of factors which prevent the attainment of 

 uniformity of the voltage gradient. The field tests demonstrated that 

 the area of the immersed electrode and the ratio of stream bottom to water 

 resistivity are factors of prime importance. The maximum distance permissi- 

 ble between electrodes is determined by the above factors. The voltage 

 gi'adient, in turn, will vary inversely as the electrode spacing. In theory, 

 the distance between the electrodes in a completely insulated tank can be 

 increased indefinitely, as long as the peak input voltage is increased 

 proportionately, without changing the uniformity or the value of the volt- 

 age gradient. 



22 



