16 



At first sight it is natural to conclude that the contraction 

 in size is due to an actual concentration of the structure of the 

 scale, but it is at once evident from the above tables and from 

 fig. 1, which has been constructed from the tables, that no such 

 concentration takes place. The scales are not all of the same 

 size in a given fish, but when sufficient numbers are examined 

 and expressed either in terms of the length of the fish or in actual 

 size it is plain that the scale expands pro]Dortionally to the length 

 of the fish. In fig. 1 the growth of the scale has been plotted out 

 with reference to the length of the fish and it has been found that 

 the measurements resolve themselves into a straight Une originating 

 approximately in " o " as the various external measurements 

 are found to do. The size of the scale from the centre as defined 

 is about 0-0182 of the length of the herring. 



The figure has been completed by the horizontal lines indicat- 

 ing the average size of the scale at the formation of each winter 

 ring, and the points where these lines intersect the hne of growth 

 of the scale ought to express the average size of the herring at 

 the formation of each winter ring. 



If the facts thus indicate that the areas of summer growth 

 deUmited by each winter ring on the scale do not suffer contrac- 

 tion, the only other explanation which appears to be possible is 

 that which has been advanced by Lea.* It is that the figures are 

 accurately stated, that they express the average size at each 

 successive year, that there is a progressive reduction in size due 

 to selection. Many obvious facts could be adduced to support 

 the contention. 



It is evident from the figures submitted in the above tables 

 that the herrings of one winter ring are large herrings of their 

 group, and it is easy to understand that in drift net fishing the 

 few examples caught would be large herring. The fish are not 

 only large but their rapid growth is evidenced by the large size 

 of the summer growth of the scale. Selection due to the mesh 

 of the net takes place with reference to the younger age grouj)s. 

 The selection by the net cannot be said, however, to be the cause 

 of the subsequent reduction in size amongst herring, all of which 

 are equally Uable to be caught by the drift net. 



* 1913. Further Studies concerning the methods of calculating the Growth of the Herring. 

 Publications de Circonstance, No. 66. 



