27 



The fish with three winter rings in the samples of February 

 and March, 1920, have a considerably larger growth up to the third 

 winter ring than have those found in the sample of July, 1919, 

 and they are very like those found in the sample of May, 1919. 

 The February and March samples are from spring spawning shoals, 

 and it will be seen in Table II. (maturity) that a large number 

 of the fish with three winter rings were either spamiing or spent 

 fish, or were likely to become spring spawiiers. The May sample 

 of 1919 came from a shoal of fish caught on grounds frequented 

 by recovering spents at that time of year. We have therefore 

 in the first three samples given above a large number of fish which 

 can be classed as spring spawning fish. They resemble one another 

 very much with regard to third year growth, and are considerably 

 larger than the fish of similar age found in the feeding shoals in 

 July. The sample of 23rd September contained 55 fish with three 

 winter rings, and of these 37 were spents ; they were autumn 

 spawners. The growth of these fish is greater than the growth 

 of those in the July sample, which consisted chiefly of developing 

 fish, but is not so great as that of the spring spawners, which have 

 come to maturity when some six months younger. 



If the fish with three winter rings in the five samples be traced 

 back in the growth tables the same features are noticeable for 

 the second year growth. The first year growth is of particular 

 interest. In the summer feeding shoal, 22nd July, the fish have 

 a first year growth ranging from 6 to 15 cm., in the autumn spawn- 

 ing shoal, 23rd September, a range of from 6 to 14 cm., and in the 

 sample of 9th March, 1920, a range from 7 to 15 cm. Fish in 

 both spring and autumn spawning shoals have a range for first 

 year growth similar to that found in the summer feeding shoals. 

 It has been pointed out already that the summer feeding shoals 

 contain young fish arising from both spring and autumn spawners. 

 The particulars are as follows for the fii'st year growth of the three 

 samples,: — 



NUMBERS AT CENTIMETRES. 



The sample of 9th March, 1920, has a higher first year growth 

 than the other samples, but in this comiexion it must be remembered 



