81 



three winter rings. Fish of this age were shown by the investiga- 

 tions of 1919 to form the greater part of the matties caught, and 

 from Wick to Scarborough they constituted from fifty to nearly 

 seventy per cent, of the samples examined. Further, the con- 

 tinuous investigations which have been made on the Northum- 

 berland shoals point to the importance of herrings with three 

 winter rings in the summer fishery. A shortage of fish of this 

 age would account for the small landings made, the small 

 extent of the shoals as evidenced by the varied landing made by 

 vessels fishing in practically the same waters, and the comparatively 

 high number of fish with two winter rings found in the North 

 Shields catches. 



It is not thought that any hydrographical condition during 

 1920 would account for the shortage of fish with three winter 

 rings as fish a year younger were present in comparatively high 

 numbers. It is more probable that the conditions which obtained 

 when the young fish were hatched, and shortly afterwards, were 

 the cause of the scarcity. If this be so, our summer fishery for 

 1920 was determined by the conditions obtaining in the autumn 

 of 1916 and the spring of 1917. 



The age composition and the state of maturity of the samples 

 examined will be found in Tables I. and II. 



No sample of autumn spawners in a spawning condition 

 was received from any of the ports. The samples which were 

 obtained consisted of mixed herrings, a large number of which 

 were spents. 



The Lerwick samples of 30th June and 25th August were 

 old fish. The majority of them had six or more winter rings, 

 and they were mostly spents or recovering spents. The sample 

 of 30th June is especially interesting in that it points to a very 

 early spawning of some shoals in Shetland waters in 1920. As 

 no evidence is to hand to show that so early a spawning season 

 is a common occurrence in Shetland waters, it is impossible to 

 say whether spawning in northern waters as early as June would 

 have any effect on the catches of autumn spawners made further 

 south along the east coast in August and the beginning of September. 

 This sample contained few fish with less than six winter rings. 

 The August sample contained more of the younger herrings, and 

 the percentage with five winter rings gives indication of the short- 



