10 



to be " full of herrings," and about the same time they were 

 being caught off the coast of Cumberland. Some small samples 

 were obtained, but not regularly, for there was no drift-net 

 fishery. (Quite big catches were, however, taken in the 

 " baulks " at Heysham.) The fish were exceedingly lean and 

 were very poor eating. Most of them were spent though so 

 many were found to be " full " that it seems probable that 

 these fish were shoaling and spawning. Apparently they were 

 present all along the coast from the Solway down to Great 

 Orme's Head. 



Thus we have to consider (1) the regular summer herrings 

 that spawn in the region between Ardglas, in Ireland, and the 

 Isle of Man, and (2) the equally regular winter spawning in 

 Cardigan and Carnarvon Bays : these two fisheries never appear 

 to fail. Then there are irregular fisheries which occur off the 

 Cumberland, Lancashire and Cheshire coasts after long intervals 

 of time. 



When the biometric investigations were commenced it was 

 thought that each of these regular or irregular fisheries was 

 that for a distinct " race " of fish. In order to establish the 

 characters of these races a large number of fish had, presumably, 

 to be measured and studied. There is so much individual 

 variability between fish and fish that many hundreds would 

 have to be measured in order to get reliable average values 

 for each of the characters taken as diagnostic of the various 

 races. Also the measurements were rather delicate ones, 

 subject to some considerable, unavoidable errors ; the fish 

 were not always in good condition when they were received ; 

 even the examination of so small a sample as 50 was quite a 

 long job ; different measurers did not always get precisely 

 the same results; we were never quite sure what were the 

 best " characters " to measure — in short, the methods were 

 not perfectly satisfactory ones and it was thought 

 desirable to suspend the routine collection of data for a 



