300 DISCOVERY REPORTS 



From these it can be seen that the theoretical weights, using mean monthly figures for K from the 



tables, gave errors of: 



+ ii'i5/276*35 kg., or +4-07% for males, 



+ o-8i/i03-29 kg., or +078% for females, 



+ ii-96/379-64kg., or +3-15% for the total sample, 



but consideration of the ' probable minimum ' weight would have saved us from any over-optimism. 

 Converting to units used among British practical fishermen, our theoretical 'guess' at the catch would 

 have been less than 2 stone too big in a catch of 7I cwt., and our cautious ' probable minimum weight ' 

 would have involved the statement that the catch was at least 708 lb. or 6| cwt., instead of the 7I cwt. 

 that we happen to know that it was. 



Another way in which the ponderal index K can assist us is well shown by the figures for this 

 station: the actual weights of the 414 measured males being known, 'true' mean K for them can be 

 computed for this individual catch thus : 



j^ w 203-116 , 



K = j^ X 100 = -^^ — X 100 = 0-603. 



Now we have the bulk weight of the 529 unmeasured males, and if we calculate their theoretical 

 weight from this figure, assuming the same mean length as the measured subsample, and find good 

 agreement between theoretical and actual weights, we have a good argument that our assumption as 

 to length is justified, thus : 



P 

 ^"^ loo '^ ^ ^ 529 = 486-27 X 0-603 X 529 g- = IS5-II4 kg., 



in fact we know that they weighed 155 kg., and the agreement is so close that there is little doubt the 

 unmeasured subsample really had a mean length almost identical with that of the measured one. 



MIGRATIONS 



We have seen that the seasonal changes in condition of M. hiibbsl correspond to those that occur in 

 European hake. Variation in condition is primarily connected with the sexual cycle, and since the 

 mam inshore movement of European hake is a spawning migration, it was expected that the move- 

 ments of Patagonian hake would also correspond with those of the better-known species. It was by 

 no means easy to demonstrate this from our scattered data. The necessity for investigating the whole 

 area as fully as possible prevented us from repeating observations within more localized portions to 

 the extent desirable when trying to follow the seasonal movements of a single species. 



One of the most valuable features discovered by Hickling in his work on European hake was a 

 direct relation between increase in depth and size of fish, which may be locally reversed at the time 

 when the larger fish are moving inshore. By watching the seasonal variation in size over limited 

 ranges of depth one may thus obtain valuable clues as to seasonal movements. But the peculiar 

 topography of the Patagonian Continental Shelf, with its uniform depths prevailing over vast distances, 

 was found to defeat this method of attack. Although correlations were found between depth and size 

 of the fish, and these showed the change of sign with the seasons that one would expect if they moved 

 in a fashion analogous to that of the European species, the correlations were not large enough to be 

 considered significant. The slight gradient of the Patagonian shelf thus masks a feature that is beauti- 

 fully clear off western Europe, where the shelf slopes more steeply and the edge is less abruptly 

 defined, giving a more continuous depth gradient in most localities. 



