For purposes of condensation, the classification of response 

 types "excellent" and "good" are combined in Table 1, since both re- 

 sponses were satisfactory. In addition, incapacitation as used in the 

 Table includes both "over -control" and "immediate electronarcosis". An 

 arbitrary niimerical value was assigned to each type of response listed 

 in the table and the average rank of response was computed for each size 

 class. An inspection of the average rank of response for each group in- 

 dicates the trend of the relationships among size, voltage gradient, and 

 response. The information is svimraarized in Table 2 which presents the 

 average rank of response for each 1-inch size class at the various volt- 

 age gradients. The true relationship is no doubt obsciired because of 

 the paucity of representatives in the majority of the size classes. 



For each size class tested, the desired galvanotaxic response 

 could be elicited over a definite range in voltage gradient c Similarly, 

 a given voltage gradient stimulated the desired movement in several 

 1-inch classes. The exact range of effective voltage gradients was not 

 clearly determined for each size class because of a lack of specimens. 

 It appears that a large portion of the smaller fish would respond favor- 

 ably over a rather wide range in voltage gradient. For the 13- and 

 lU-inch length classes the effective range extended over approximately 

 0.22 volt. The data tend to indicate also that the effective range de- 

 creases with an increase in length of the test animals. 



The limits of size in which a given voltage gradient is effec- 

 tive are not readily discernible o For instance, an applied voltage gra- 

 dient of O.liS volt per inch may affect the majority of rainbow trout 

 18-22 inches long, with its maximum effectiveness probably applicable to 

 20-inch fish. That same voltage gradient would be too high for the 

 majority of trout above 22 inches in length and too low for trout falling 

 below 18 inches. However, as indicated in Table 1, a good response was 

 obtained from one trout only 1$ inches long which was subjected to 0,U8 

 volt per inch. Another l5-inch trout was electronarcotized at a still 

 lower gradient, 0.37 volt per inch. Similar examples of variability or 

 individuality displayed by the test animals were noted throughout the ex- 

 periment. They contributed considerably to the variability of the data. 



Field tests on rainbow trout .— Only 3U adult rainbow trout 

 were available for the field tests which were conducted in the Little 

 Ocqueoc River. The stream was approximately l5 feet wide, 9 to l5 inches 

 deep, and had a msLximum water velocity of slightly under 2 feet per second 

 at the experimental site. Water temperature remained at 62° to 63° F. 

 throughout the testing period. The stream bottom consisted of a mixture 

 of fine and coarse gravel with some scattered small rubble. The trap 

 (anode) was installed in lU inches of water, and the cathode was placed 

 along the opposite bank in 9 inches of water. The small size of the stream 

 limited the distance between the centers of the electrodes to 10 feet. 

 Below the device, the stream was blocked by a seine to confine the fish 

 to the experimental area. 



13 



