5i8 FISHES. 



mate, witness the gemmeous dragonet {Callionymus lyra) 

 and the stickleback ( Gasterosteus), and this is especially true 

 at the breeding season. The colours of many fishes change 

 with their surroundings. In the plaice and some others the 

 change is rapid. Surrounding colour affects the eye, the 

 influence passes from eye to brain, and from the brain down 

 the sympathetic nervous system, thence by peripheral nerves 

 to the skin, where the distribution of the pigment granules 

 in the cells is altered. In shallow and clear water this power 

 of colour change may be of much protective value, but it 

 seems likely that this has been exaggerated. An appre- 

 ciation of the protective value of colouring demands careful 

 attention to the habits and habitat of the animals, to the 

 nature of the light in which they live, and to the enemies 

 which are likely to attack them. 



Food. 



The food of Fishes is very diverse — from Protozoa to 

 Cetaceans. Sharks and many others are voraciously carni- 

 vorous, many engulf worms, crustaceans, insects, molluscs, 

 or other fishes ; others browse on sea-weeds, or swallow mud 

 for the sake of the living and dead organisms which it con- 

 tains. Their appetite is often enormous, and cases are 

 known {e.g., Chiasmodon fiiger), where a fish has swallowed 

 another larger than its own normal size. Many fishes follow 

 their food by sight; many by a diffuse sensitiveness, to 

 which it is dilTicult to give a name ; and others, it would 

 seem, by a localised sense of smell. 



Some Points of Stnicttcrc— Fins.— Along the median line of the 

 dorsal and ventral surfaces of some fishes, e.g., flounder, there is a con- 

 tinuous fin — a fold of skin with fin rays and underlying skeletal supports. 



In the embryos of many fishes, the same continuous fringe is seen, 

 while the adults have only isolated median fins. There is no doubt that 

 these isolated median fins — of which there may be two dorsals, a caudal, 

 and an anal or ventral — arise or have arisen from a modification of a 

 once continuous fin, which is suppressed at one part and increased at 

 another. 



Now, the paired fins, which correspond to limbs, often resemble 

 unpaired fins in their general structure, and in their mode of origin. 

 In some Elasmobranch embr}'0s, Balfour showed that the pectoral and 

 pelvic fins were connected by transitory lateral ridges. It is therefore 

 likely that the paired fins have arisen b)- a localisation of two once con- 

 tinuous latcr.al folds. Why there should be only two pairs we do not 

 know. 



