seen to fall off while the droppers were being retrieved. It is thought likely 

 that an even greater number of sardines may have pulled loose before the hooks 

 became visible to the observer. There is no evidence to indicate that the rate 

 of loss during hauling increases with soaking time. 



Table 5. --Sardine bait return and total catch by 

 method of baiting (single vs. double), 

 station 6, John R. Manning cruise 12 

 (based on a total of iSo single-hooked 

 baits and 60 double -hooked baits) 



Table 6. — Summary of chl-square analyses of sardine bait 

 return and total catch by method of baiting 

 (single vs. double), John R. Manning cruise 18 

 (based on a total of 1,260 hooks per method of 

 baiting; table 28) 



i/ Stations have been combined to give expected values greater than 5. 



Herring 



On Manning cruise 14 (January-March 1953) a single vs. double hooking 

 experiment was carried out on eight fishing stations using herring. The method 

 of double hooking was similar to that of the sardine experiment on station 1 

 (fig. 3e) of Manninj^ cruise 12, which has already been discussed. The analyses 

 of both bait return and catch are shown in table 7. The higher return of the 

 double-hooked herring is statistically significant as indicated by the total and 

 pooled chi-squares (both P values less than 0.01) . The double-hooked returns 

 ranged from 1.3 to 7.4 times as many as for single-hooked herring. The heavy loss 

 of single-hooked baits appears to be reflected in the size of the catches, for 

 the significant pooled chi-square of 4.76 (P<0.05) indicates a tendency for more 

 fish to be caught on the double-hooked baits (table 7). However, this should not 

 be taken as conclusive evidence in view of the following experiment carried out 

 on the Cavalierl cruise (August-September 1952). 



