192 



Part III. — Tiventy-second Annual Report 



Thus, ill the haul of 14th November (1) there were eight specimens 

 between 13 and 16cm. (viz.:— 132, UO, 144, 146, 148, 152, 156, 161mm.), 

 and the next size was 242ram. In that of 28th December the group 

 was represented by ten specimens, measuring from 137mm. to 168mm., 

 the next largest being 215mm.; in the haul of 23rd January it was 

 represented by nine specimens, from 138 to 186mm., the next size 

 being 222mm., and in the haul of 1st April by fourty-four, from 144 to 

 191mm., the next largest being 217mm. 



A consideration of the first and second series or generations in these 

 cases throws light on the rate of growth of the fish, and the measure- 

 ments may be grouped as follows, showing the smallest and largest 

 specimens represented in each case, and the mean size: — 



The arithmetic average, it may be explained, is obtained by adding 

 up the sizes of the fishes represented in each group and dividing by the 

 number of fishes ; it will deviate from the true average size in one 

 direction or the other if the larger or the smaller fishes of the group 

 predominate in numbers. The geometric mean is the middle figure 

 between the extreme sizes, viz. :— the largest fish and the smallest; its 

 accuracy depends upon the limits of the group being truly indicated. 



Considering first the difference in size between the first series of 

 witches and the second series, which are one year older, it is evident that 

 the size of the single specimens of the first series obtained on 14th 

 November and 23rd January respectively, are not representative, the 

 former (56mm.) being too large and the latter (47mm.) too small. This 

 is shown by the townet collections in October and November, as 

 represented in Table A. and Plate XI, in which specimens measuring 



