Ill 



THE PONDERAL INDEX 



195 



progressive changes : delicate — because our measurements of length 

 are very accurate on the whole, and weighing is a still more dehcate 

 method of comparison. 



Thus, in the case of plaice, when we deal with mean values for 

 large numbers and with samples so far "homogeneous" that they 

 are taken at one place and time, we find that k is by no means 

 constant, but varies, and varies in an orderly way, with increasing 

 size of the fish. The phenomenon is unexpectedly complex, much 

 more so than I was aware of when I first wrote this book. Fig. 50 



120 — 



.2 



o 110 



§ 100 



^ U 



90 



22 



25 



40 



30 35 



Length (cm.) 



Fig. 50. Changes in the weight-length coefficient of plaice with 

 increasing size; from March and December samples. 



shews the weight-length coefficient, or ponderal index, in two 

 large samples, one taken in the month of March, the other in 

 December. In the latter sample k increases steadily as the plaice 

 grow from about 25 to 40 cm. long; weight, that is to say, increases 

 more rapidly than the cube of the length, and it follows that length 

 itself is increasing less rapidly than some other linear dimension. 

 In other words, the plaice grow thicker, or bulkier, with length and 

 age. The other sample, taken in the month of March, is curiously 

 different; for now k rises to a maximum when the fish are some- 

 where about 30 cm. long, and then decHnes slowly with further 

 increase in size of the fish; and k itself is less in March than in 

 December, the discrepancy being slight in the small fish and great 

 in the large. , The "point of inflection" at 30 cm. or thereby marks 



