DISTRIBUTION AND GENERAL NOTES ON THE SPECIES 



Table 15. Condition factors of the largest Merluccius hubbsi weighed 

 {Note. All were from deep water in the southern region.) 



297 



A similar phenomenon is suggested among fishes of other species: TSning (1937, pp. 8, 23) has 

 shown that the largest of the west Greenland cod do not take part in the return spawnmg migration 

 to Icelandic waters which is characteristic of their slightly smaller sisters. It is certam that such 

 spawning as occurs in Greenland waters themselves is on quite a minor scale; moreover, the extra 

 large stationary west Greenland cod tend to be found near the extreme northern limit of the range 

 of the species, where spawning does not occur. The analogy with our observations on the largest 

 female M. hnbbsi seems very close, and the suggestion that the latter are past spawning gains strength 

 from the facts that males are scarce or absent from the most southerly waters where the large females 

 are found and that at no time have the latter been met with farther north.i 



If the figures in the left-hand columns of Table 14, showing variation of K with length, are trans- 

 posed to a percentage basis, it is possible to read oS the correction to be applied to seasonal values 

 for K (as plotted in Fig. 24), for fish of given mean length, working to the mid-point of each month. ^ 

 This has been done in Tables 16 (males) and 17 (females) for the length range likely to cover the 

 mean lengths of any sizeable sample of M. hubbsi. 



With the aid of these tables the approximate weight of any sample of fishes of known mean length 



can be calculated thus : W= — xi^xf 



100 

 where / is the number of fishes in the sample, the sexes being dealt with separately and the results 

 summed. Some error is unavoidable, and these mean K values do not apply well to individual fishes. 

 (Hake weights are too erratic to be dealt with individually, unless they are weighed gutted. Food niay 

 be snatched up in the trawl or stomachs may be evacuated, and their gorging habits are such that 

 even ' in nature ' the weights of any two fish of the same length at a given time might be expected to 



"^^ An Zttnl source of error is the distance in time from the mid-point of the month in which the 

 sample is taken, but equally obviously it is impracticable to tabulate K for every day of the year. 

 Such derived values cannot take account of abnormal seasons. By using the tables given it has been 

 found that with samples of ten or more fishes, using the formula for each sex separately, and checking 

 against eighteen samples of which the actual weight was known, the 'theoretical weights mean 

 . gross error was 6-7%, with a range of -16-4 to +8-4%. Further, it could be shown by summing 

 fhe squared deviatiot from known weights that the derived values tabulated gave a better estim^e 

 than other average values for K, such as mean K for the whole year, either umping the sexes or 

 treating them separately. The range of error is great, but from a purely practical P^^^ f ^^ ^ ^ 

 be extremely valuable to know the probable rmnimum weight o catches where it w s ^^^^^^^'^ ^ 

 get actual weights in the field. For this reason ' theoretical ' and probable minimum weights for the 

 unweighed catches of the first and second surveys have been calculated wherever the length data 

 permit and are tabulated along with the numbers of fishes in those catches in Appendices IIa and 

 1 Of course they may be there at depths greater than we were able to fish. 



