142 Pari III. — Tiventij-secooid Annual Report 



I have likewise carried on a number of experiments in order to ascertain 

 the relation which exists between the growth of fishes and the temperature 

 of the water in which they live. It is well known from previous observa- 

 tions that in the winter season the growth of fishes, at least in the inshore 

 waters, is slower than it is in summer ; in the case of those living 

 in shallow water, subjected to the changes in the temperature of the air, 

 and where the extremes of heat and cold are at their maximum, growth 

 may be entirely arrested in winter. In the Annual Report of the Board 

 above referred to I gave particulars on this point with regard to the young 

 plaice living on the beaches, and exhibited a curve in which the relation 

 between the temperature of the water and the degree of growth of the 

 plaice was established. 



The experiments, which are described in detail below, consisted in 

 keeping fishes of various species in tanks in which the water was arti- 

 ficially heated, and the result on the growth of the fish was very marked, 

 those in the water of a high temperature growing much faster than 

 those in the water at lower temperature. It was, moreover, shown, as 

 might have been anticipated, that the fishes in the warmer water ate much 

 more food than those in the colder water, the digestive ferments being 

 more active at the higher temperatures, and the fish being thus able to 

 digest a larger quantity of food in a given time. It was found that the 

 appetite of the fishes was in relation to the power of digestion, that is to 

 the temperature of the water, those in very cold water scarcely eating 

 at all, although abundantly supplied with food. In the same way, the 

 metabolism in the tissues was more rapid, and nutrition and growth much 

 accelerated. 



Certain differences were found to exist in different species, which are 

 referred to below. 



2. The Relation of Length to "Weight. 



In dealing with the rate of growth of fishes it is customary to take one 

 of the dimensions of the fish and compare the variations of this 

 dimension at different periods or in different collections. In some cases, 

 as with the rays, it is more convenient to take the breadth across the 

 pectorals than the length. The selection of one dimension for com- 

 parative measurement is very convenient, and it is accurate on the 

 assumption that the fish grows equally in all directions, increasing in 

 breadth and thickness in the same ratio as it does in length. It is 

 obvious, however, that the true criterion of growth is the increase in the 

 mass of the fish, and this can be determined either by the \ariation in 

 the volume or in the weight. 



The determination of the variation in volume is a somewhat slow 

 process, and the methods are subject to difficulties in practice. With 

 small fishes a burette may be used with accurate results ; with those 

 of large size the quantity of water displaced by the fish was measured 

 separately in a burette, the fish itself being placed in a convenient 

 vessel. In the case of fishes of moderate dimensions the method used 

 was to place them in a vessel provided with a syphon to draw off the 

 amount of water displaced, which was then measured in a burette ; the 

 bore of the syphon being so adapted as to always remain full of fluid. 

 As a rule this mode of determining the increase in bulk was found to be 

 less satisfactory than the method of weighing the fish, and this was the 

 method chiefly employed. 



According to the well-known law, that the volume of similarly-shaped 

 bodies of the same specific gravity vary directly as the cube of corre- 



