ADULT HEKRINGS 169 



us, ' the young will have time to form the scale, and the follow- 

 ing winter will be recorded ; but obviously in autumn spawners, 

 the first winter ring will refer to the second winter.' He thus 

 follows the Norse theory that the autumn-spawned herring 

 carries its birth-mark all through hfe in the shape of a scale 

 marked with a first winter ring, which is relatively distant from 

 the centre of the scale. The point is important as Storrow is 

 not convinced that autumn fish can thus he distinguished from 

 the spring herring. 



Between 1911 and 1914 observations were made on two 

 schools of herring — the Northumberland coast herring which 

 are caught in drift-nets, and which spawn from the middle of 

 August to the beginning of September ; and the Dogger Bank 

 herring caught in trawls, which spawn from the end of Septem- 

 ber to the middle of October. It is important to remember here 

 Hjort's warning on p. 1G5, that the seine-net is a much better 

 instrument than either the trawl or the drift-net for the purpose 

 of herring investigations, because it retains fish of all sorts and 

 sizes, whereas the meshes of the trawl and drift-net allow the 

 youngest fish to escape, and do not in fact ever catch more than 

 1 per cent, or 2 per cent, of herrings marked with one winter 

 ring. 



The observations showed that the annual percentage of 

 various broods of herrings in the samples were as follows : 



The tranverse lines indicate the history of each brood through 

 the period of three years, which is, needless to say, much too 

 short a period to form a cycle. The interesting points which 

 emerge are that fish between three and four years old formed 

 a very large proportion of the samples in all three years ; that 

 the ' four ringers ' w^ere common in 1912 and 1913 but not in 

 1914, and the ' two ringers ' in 1914 but not in 1912 or 1913, 

 and that few of the fish are over six years old. 



t 



