16 



This Lea has made the attempt to explain by assuming that 

 in a given year class the herrings are resolved into components 

 according to the state of development, and that the change in the 

 rate of growth is brought about by the herring forsaking one com- 

 j)onent and attaching themselves to another. In another place * 

 I have tried to demonstrate that the products of each spaw?iing 

 ground are congregated into a shoal or shoals, and that they remain 

 and grow and migrate together. It is evident that the large 

 herring in each shoal will have opportunities of leaving to attach 

 themselves to the group of the previous year, and similarly the 

 smaller will be tempted to leave to join the group below. It is 

 clear also that as the immature become mature they join in suc- 

 cessive groups the spawning shoal until all the survivors at the end 

 of this period have been absorbed. During this period the rate 

 of growth gradually declines, but especialty after the third year. 

 As the same series of events follov/ one another every year it is 

 difficult to follow Lea when he says that the immigration of the 

 immature into the mature will cause a change in the growth 

 dimensions. The feature of apparent contraction is equally well 

 observed when immature herring are investigated as they are in the 

 tables above. 



Fortunately we are able to approach the subject in another 

 way. The annual advent of the Northumberland school takes 

 place eo.ch summer, and spawning occurs about the end of August. 

 The first winter is therefore not recorded on the scales. The 

 actual size of the herring ought to ])e then approximately that 

 of the calculated size at each mnter ring + 1 (w.r. + 1). We 

 are thus able to contrast the actual with the calculated size. 



The results of the examination in this manner are detailed 

 in Table IV. The samples A — L are those aheady given in Table 

 I., and samples 7 and 8 have also been considered (Tables II. and 

 III.). The samj^les O — X are trawl caught herring of the season 

 of 1913. The last are especially noteworthy for they are caught 

 in September and October each year, at a period therefore close 

 to that of the formation of the winter ring. 



The above samples represent large enough numbers of herring 

 to warrant general conclusions. Other analyses of a similar nature t 



* 1916. " The Migrations of Fish." 



t One is given on page 7fi, "Migrations of Fish." 



