332 



DISCOVERY REPORTS 



dividuals were obtained at that season. From this it seems safe to infer that schooHng of A'^. ramsayi 

 is most marked during autumn. 



Table 31. Seasonal variation in relative abundance of Notothenia ramsayi at different deptlis 



An investigation of the mean lengths within depth categories at each season showed that there was 

 a significant increase of size with depth. The fish caught between 201 and 250 m., and 251 and 300 m., 

 in winter, do not show this ; but all the other observations, taken in pairs successively (twelve pairs in 

 all) show the deeper sample of the pair to contain significantly longer fish (Table 32). It is therefore 

 clear that A^. ramsayi conforms to the general rule 'larger fish in deeper water'. 



Table 32. Variation in mean length of Notothenia ramsayi at different depths during 



each of the four seasons of the year 



These results also indicate that the vast majority of that part of the N. ramsayi population found 

 beyond the shelf edge (i.e. in depths greater than 200 m.) are more than 20 cm. long. Reference to 

 the actual frequencies shows that only sixty-three out of 1894, or 3-3 % , of the fish captured over the 

 edge were less than 20 cm. long. The winter migration to deeper water is thus almost entirely confined 

 to the larger fish that, as will presently be shown, are almost certainly more than two years old. 



The numerous length-frequency data available for N. ramsayi can only be used to test the probable 

 ages of the younger fish, by Pettersen's method. It was rarely possible to sex the individuals, owing to 

 pressure of work upon the more obviously useful species, and, when sexing was achieved, it was 

 found that a very large majority of the larger fish (of more than 25 cm.) were females. It is therefore 

 almost certain that after maturity is reached the male and female growth rates diverge, the females 

 growing the faster. Consequently one cannot base any conclusions as to age groups or growth rate 

 upon the length frequencies of the larger fish in unsexed data. If the mature males are considerably 

 smaller than the females, the apparent excess of the latter will in the main be accounted for by the 

 selective action of the net. For the younger stages, however, the rich autumn hauls (many of them 

 taken in the ' accessory nets') furnish length-frequency data that seem to show age groups with some 

 certainty, judging by the consistency with which modes recurred at the same lengths. 



