The remaining surface area between the ad- 

 vancing net wall and the stationary wall at the 

 time of escape was used as a criterion of the 

 animals' readiness to esca])e. This index was 

 the inverse of latency, since the smaller the area 

 that remained when the animal escaped the 

 longer would be the latency. Although we 

 measured latency in seconds, we felt the net po- 

 sition was the preferable measurement because 

 the rate of net movement was imi)recise, where- 

 as the actual stimulus for escaj^e, the reduction 

 in the swimming area, could be measured rel- 

 atively accurately. In presentation of the data, 

 the logarithm (to base 10) of the remaining area 

 at the time of escape is plotted on trial number. 



Because procedures and plans were modified 

 during the course of the experiments, results 

 and interpretation are combined in the presen- 

 tation of the results. 



PRELIMINARY EXPERIMENTS WITH 

 TRAINED PORPOISE 



We anticipated that the behavior of the por- 

 poise would change rapidly during the course of 

 the experiments; thus, to avoid wastage of the 

 naivete of the limited and expensive supply of 

 untrained animals, we conducted a series of pre- 

 liminary exi:)eriments with three trained por- 

 poise (Table 2, Figures 2-4). 



WIDTH OF OPENING 



The effect of the width of the opening on the 

 escape l)ehavior of the three trained porpoise 

 was first tested by presenting on alternate trials 

 an escape route of standard width, 5.5 m, and one 

 either 3.1 or 3.8 m wide. In Waimea (Figures 

 2-1 and II) and Nohea (Figure 4-1 V) there was 

 some evidence that the porpoise escai)ed sooner 

 when the wider e-scajie route was used but not in 

 Nani (Figure 3-II). We felt this was i)robably 

 an artifact of experimental design as described 

 above and consequently we considered only the 

 data from the block experiments for evaluating 

 effects of width on the trained pori)oise. To de- 

 termine the width of opening that would influ- 

 ence performance, Nohea was tested over six 



FISHERY BULLETIN: VOL. 70, NO. I 

 WAIMEA 



FAILED TO ESCAPE 



092m DEEP, 

 WIDTH VARIED 



092m DEEP. WIDTH VARIED 



TRIAL NUMBER 



Figure 2. — Results of experiments with trained porpoise 

 Waimea. Each plot summarizes one day's continuous 

 experimentation, as follows: I. May 21, II. May 22, 

 III. May 23, IV. May 25, 1970. 



NANI 



55m WIDE, DEPTH VARIED 



/ 



I Im DEEP, 

 WIDTH VARIED 



/ 



/ 



o 



55m WIDE, „ 

 DEPTH VARIED 



OI5m DEEP 



O O O FAILED TO ESCAPE 



55m WIDE. 

 DEPTH VARIED 



TRIAL NUMBER 



Figure 3. — Results of experiments with trained porpoise 

 Nani. I. May 26, II. May 27, III. May 28, 1970. 



52 



