no DEPARTMENT OF THE XATAL SERVICE 



6 GEORGE V, A. 191& 



line, AC, was drawn from the centre of the scale making an acute angle with the 

 first, and of a length which would represent the length (on some convenient scale, 

 usually 1 :10) of the fish from which the scale was taken. A line was then drawn 

 connecting ^B and C and a series of lines were drawn parallel to BC from the 

 points of intersection, of the line AB with the rings on the scale. The growth of the 

 fish which would correspond to any of the winter rings would then be read oif on the 

 line AC. According to this, the fish, which it is demonstrated was lY cm. long at the 

 end of the first year, 35 at the end of the second, 53 at the end of the third, 68 at the- 

 end of the fourth, and 80 at the end of the fifth, while it was 86 cm. long when 

 caught. 



The ages of all the fish whose scales were taken were calculated in this way and 

 set down in table I. In every case two scales were used and unless the determinations 

 from the two agreed, or nearly so, as in the figure, other pairs were taken until two 

 were found that did agree. "When all these were averaged up, it was found that the 

 average sizes for a codfish were : — 



First year, length, 14-5 cm. 



Second year, length, 35-9 cm. 



Third year, length, 49-8 cm. 



Fourth year, length, 64-9 cm. 



Fifth year, length, 82 cm. 



Sixth year, length, 90-5 cm. 



Seventh year, length, 99-3 cm. 



Eight year, length, 115-0 cm. 

 Of the older ones we had too few samples to yield strictly correct results. 

 The length frequencies of the age classes calculated from the scales of all the 

 fish in which these could be done satisfactorily were also plotted. The older ones 

 and a good many of the younger ones were omitted owing to the difiiculty of applying 

 this method to any but the very clearest scales. For the sake of comparison another 

 curve was made in exactly the same way including only Nos. 1 to 112, i.e., only fish 

 caught between June 11 and 24. Since the curves, calculated on the assumption that 

 the growth of the scale of the fish is proportional to the growth of the fish, tell the 

 same story as the curves based on actual measurement of fish, the growth of the scale 

 must be proportional to that of the fish. 



The most casual study of the tables and graphs prepared showed that 

 the rates of growth for the individual fishes vary widely, so that scarcely any two 

 fish have the same life history in this respect. Nevertheless it is possible to obtain 

 an average rj^te of growth for the given locality. This was done, using the averages 

 obtained from the calculated lengths. Separate curves for males and females were 

 plotted, and though not much importance must be attached to "them, they seemed 

 to show that the females grow faster than the males during the first two years, and 

 then suddenly their rate of growth falls ofi so that the curves cross between second 

 and third years. If the relative proportion of males to females varied in the different 

 year-classes, it is quite possible this might account for the irregular features in tlie 

 graphs studied. 



