20 



five winter rings, nor the herrings of 1916 with five winter rings, 

 and comprising 21 per cent, of the samples being represented in 

 1917 by 4 per cent, with six winter rings. 



The presence of a high percentage of herrings with three winter 

 rings in the samples of 1918 may be accounted for by the arrival 

 on our coast of a rich year class, but again we have no evidence 

 of the predominant year classes of 1916 and 1917 influencing to 

 any great extent the number of older fish amongst the shoals, 

 and the percentages of herrings with five and more winter rings 

 were less than in 1916 and 1917. 



Another point worthj^ of attention is the small percentage 

 of herrings with two winter rings in the samples. In 1914 these 

 comprised 27 per cent, of the samples, and were more abundant 

 than in any other year. In the following year, 1915, the per- 

 centage of herrings, 67 per cent., with three winter rings was the 

 highest of which we have any record. But there is no sign in the 

 samples of 1917, Avhen the fish with two winter rings represented 

 less than one per cent, of the fish examined, that in the following 

 year, 1918, herring with three winter rings would be the pre- 

 dominant year class. It cannot be said that the number of 

 herrings with two winter rings gives an indication of what the 

 age composition of the shoal will be in the following yesir. There 

 is a slight indication in 1914 but not sufficient to give reason 

 for expecting in 1915 that herrings with three winter rings would 

 be present in such a high percentage as 67. Whilst the drift net 

 may be selective as regards the size of the fish caught, yet had 

 henings with two winter rings been present in the shoals fished 

 in 1917 in sufficient numbers to make herrings with three winter 

 rings the predominant year group of 1918, they ought to have been 

 represented in the samples by more than one per cent., or the mesh 

 of the nets used has been much greater than that used in other 

 years. Also, if the small percentage of fish with two winter rings 

 be attributed to the selective power of the method of fishing this 

 is more e\ddent in 1916 and 1917, and to some extent in 1918, 

 than in the other j^ears. It is not thought that the low percent- 

 age of herrings with two winter rings is due to selection arising 

 from the method of fishing, but that these fish are not present in 

 any great number in the shoals fished. In support of this, there 

 is a sample of Northumberland Coast herring caught by trawl 



