Ill 



THE GROWTH OF FISHES 177 



back a year, a growth-curve extending over two seasons was obtained ; 

 when extrapolated, it seemed to start from zero about May or June, 

 and this date, at the beginning of the hot season, was in all proba- 

 bihty the actual spawning time. Growth stopped in winter, a 

 common thing in our northern climate but surprising at Madras, 

 where the sea-temperature seldom falls below 24° C. Shells over 

 40 mm. long were rare, and over 50 mm. hardly to be found — an 

 indication that Paphia seldom lives over a third season. Here then, 

 though the numbers studied were all too few, the method tells us 

 with httle doubt or ambiguity the age of a sample and the growth- 

 rate of the species to which it belongs*. 



Dr C. J. G. Petersen of Copenhagen brought this method into use 

 for the study of fishes, and up to a certain point it is safe and 

 trustworthy though seldom easy. For one thing, it is hard to get 

 a "random sample" of fish, for one net catches the big and another 

 the small. The trawl-net takes all the big, but lets more and more 

 of the small ones through. The drift-net catches herring by their 

 heads; if too big, the head fails to catch and the fish goes free, if 

 too small the fish shps through; so the net selects a certain modal 

 size according to its mesh, and with no great spread or scatter. 

 When we use Petersen's method and plot the sizes of our catch of 

 fish, the younger age-groups are easily recognised, even though they 

 tend to overlap; but the older fish are few, each size-group has a 

 wider spread, and soon the groups merge together and the modal 

 cusps cease to be recognisable. There is no way, save a rough 

 conjectural one, of analysing the composite curve into the several 

 groups of which it is composed; in short, this method works well 

 for the younger, but fails for the older fish. 



Fig. 44 is drawn from a catch of some 500 small cod, or codhng, 

 caught one November in the Firth of Forth, in a small-meshed 

 experimental trawl-net. They are too few for the law of large 

 numbers to take full effect; but after smoothing the curve, three 

 peaks are clearly seen, with some sign of a fourth, indicating about 



* R. Winckworth, Growth of Paphia undulata, Proc. Malacolog. Soc. xix, 

 pp. 171-174, 1931. Cf. {int. al.) Weymouth, on Mactra stultorum. Bull. Calif. 

 Fish Comm. vii, 1923; Orton, on Cardium, Journ. Mar. Biol. Assoc, xiv, 1927, 

 on Ostrea, and on Patella, ibid, xv, 1928; Ikuso Hamai, on Limpets, Sci. Rep. 

 Tohoku Imp. Univ. (4), xn, 1937. 



