124 DISCOVERY REPORTS 



'Discovery II ' has been scarce at this time of the year owing to the difficuhy of fishing nets in the 

 prevaihng weather and ice conditions, but adolescents have been obtained throughout the year, and 

 it is the half-grown specimens belonging to stages 3, 4 and 5, which are predominant during the winter 

 months. Young adults do not appear until September and full adults not before October. On the 

 evidence available, therefore, it is reasonable to conclude that adults do not exist in the catch after the 

 pairing and spawning season is over, because they have died out. 



It would be interesting if the time of duration of each stage could be established with some cer- 

 tainty, but this is not so easy. Growth in the female is slower than in the male, and it seems likely that 

 the earlier stages anyhow take longer to pass through. There are indications in the table of maximum 

 frequency of occurrence of the stages (Table 13, p. 122) that each stage lasts 2 months in the male 

 and 2i months in the female, and on this basis it is possible to work out the months in which each 

 stage should be theoretically at a maximum. Supposing that: (i) the spawning period extends from 

 November to April, i.e. 5^ months, (2) the adolescents first appear in August, i.e. 9 months later, 

 (3) the maximum spawning period occurs in February, March and April, and (4) each stage lasts 

 2 months in males and 2^ months in females, then, in theory, the maximum occurrence of each stage 

 should be as in Table 16: 



Table 16 



The months, in which the stages have actually been found to be at a maximum in practice, are 

 already known (Table 13), and have been italicized in, or (where necessary) added to, the above table. 

 There is on the whole a very fair degree of correspondence between expectation and reality, especially 

 when it is remembered that the material used came from very many seasons and localities. It is 

 interesting to note that, in practice, growth is slower during the winter months than it is in theory, 

 the actual maxima occurring later in the season than expected. 



Of course the stages overlap one another in occurrence; this is the outcome of the protracted 

 spawning season. Some idea of the extent to which this takes place can be obtained by tracing the 

 growth of the generations arising from eggs spawned in different months. Suppose that the months 

 considered are those in which the eggs are known to be at a maximum, namely February, March and 

 April. We know that these eggs can, under optimum conditions, reach the first adolescent stage in 

 7 months, though normally they take an average of 8-9 months to do so. In each successive batch of 



