CA\ADIA\ risHERIE.^ E\I'EniTIU\, HH'i-to 



137 



On the whole, however, the samples are as nearly as possible equal in rejrard to 

 growth of the I'.HH year-class, as they were also found to be nearly equal in respect of 

 age composition. 



The question now arises, whether the picture of growth presented by the 1904 

 year-class can be taken as representative for the remaining year-classes, and espe- 

 cially for that of 1903. In order to investigate this point, table IT has been drawn 

 up. showing the growth of the 1904 year class compared with that of the year classes 

 1903, 1905,, and 1906. as also with all older fish taken together and all younger ones 

 together. 



Table 17. — Averag's for year class 1904 compared with those for other year-classes 

 in the samples from Newfoundland. Averages based upon all nine samples. 



ear class. 



19ll-li)06. . 



1905 .. 



1904. . 



190.3. . 

 1902-1891... 



'... 



61 

 6-61 



It will be seen from the table that the important features are common to the older 

 fish. The younger ones, and also, to some extent, the 190t) year-class, exhibit greater 

 deviation, a phenomenon also noted in the great majority of other herring samples 

 (ride p. 144.). It is impossible to say whether this dissimilarity in the younger tish, in 

 the present instance, is due to selective effect of the implements used, to the dissocia- 

 tion of year-classes, or to intermingling with fish of different growth. Probably each 

 of these features is to some extent responsible. It is at any rate evident that the 1904 

 year-class does not differ essentially, in point of growth, from the older tish. Also 

 that the small f,, at least, is probably characteristic for the herring taken on the 

 coast of Newfoundland as shown by table 18,. the growth of quite young fish from 

 western bay of Port au Port, August 16-17, 1915. The 1914 year-class, which is well 

 represented, and that of 1913, which furnished eleven specimens, show exactly similar 

 values for t^ and t„ to those of the old fish; the 1910 year-class is represented by only 

 two specimens, so that the figures for this are practically valueless, and are only 

 included as a matter of form. 



T.^BLE 18. — Growth of small, immature herring caught by drift nets in West Bay of 



Port au Port, Aug. 1915. 



2. Samples from the Magdalen islands. — The 3903 year-class is the best common 

 year-class among the older fish in the Magdalen islands two samples, taken in spring 

 1914 and 1915. In the one sample, there were seventy-four of this year class, in the 

 other unfortunately but seventeen. 



Table 19 shows the average values for the different increments (t) of this year- 

 class in the two samples. In the 1914 sample, t^ and t^ are less than in that from 1915, 

 whereas t. — f^ are greater. The remaining growth dimensions are practically alike in 

 both samples. 



6551—13 



