Data were transformed from percentages 

 (table 2) to arc sin Vpercentage, and all 

 analyses were run on such trainsformed ob- 

 servations. Preliminary analyses showed, 

 for both unshocked and once-shocked fish, 

 negligible average difference in effects 

 between the two times the test conditions 

 were repeated. It was impossible to test 

 both unshocked and once-shocked fish at the 

 same time with each combination of variable 

 factors, because of the limited supply of 

 fish for testing. Therefore, we analyzed 

 separately the data from the tests with 

 unshocked and once-shocked fish as well as 

 the data on the two groups combined. The 

 completed analyses are shown in table 4 

 for unshocked fish and in table 5 for once- 

 shocked fish, and the analysis for the two 

 groups combined in table 6. 



cant differences between "blocks" for both 

 unshocked and once-shocked fish. 



Differences between blocks Ccinnot be 

 attributed to a single cause (see page 4); 

 therefore, we tested for the significance 

 of the relationship between temperature 

 difference and percentage difference in 

 numbers of squawfish that entered the "posi- 

 tive" end zone. Separate tests were made 

 for the north and south blocks, and for the 

 unshocked and once-shocked fish in each 

 block. The four "t" values obtained were 



(1) t = 0.598 for unshocked fish in north block 



(2) t = 0.466 for once-shocked fish in north block 



(3) t = 0.219 for unshocked fish in south block 



(4) t = 1.188 for once-shocked fish in south block 



Shocking condition . — As may be seen 

 from table 6, there was no difference 

 between shocking conditions (unshocked vs 

 once-shocked), nor was the interaction of 

 shocking condition with any other variable 

 significant . 



Effect of blocks . — All three analyses 

 (tables 4, 5, and 6) showed highly signifi- 



The 5-percent significant level for 

 "t" with 25 degrees of freedom is 2.06. 

 There is, therefore, no evidence of a rela- 

 tionship between temperature and percentage 

 of squawfish led. For illustration we 

 include a graph (fig. 5) showing the differ- 

 ence in percentage of fish that entered the 



Table L,— Analysis of Tarlanee of the effects of potential, puis 

 frequency, and pulse duration on the movement of 

 unshocked aquairfish . 



Table 5, Analysis of variance of the effects of potential, pulse 



frequency, and pulse duration, on the movement of once- 

 shocked squawfish . 



11 



