IN FISH. 97 



appearance, and also to prevent the Turbots from burrowing in any 

 mud which might accumulate, supplied the tanks with a partial 

 covering of shingle, which he obtained from the shore near Brighton ; 

 this shingle is screened, to pieces from one inch to half an inch in 

 diameter, and is made up of bright brown pebbles interspersed with 

 white ones of similar dimensions. He found that all the Turbots 

 had taken up the colouring of the pebbles, and became coloured 

 brown with white spots, agreeing exactly in shade with the pebbles 

 on which they lie. It is difficult in this case to accept the theory 

 which has appeared possible in the case of other fishes, as it is 

 almost certain that the Turbots could never either for themselves 

 or through hereditary transmission have had experience of living 

 on pebbles assorted to a certain size and of two colours ; and it is, 

 moreover, believed that they are never found on shingly shores at 

 all, but only w^here they can easily hide in sand or mud. It is 

 difficult to imagine that the position in which the eyes are placed 

 is favourable for the constant observation of the ground over which 

 the fish is passing. A group of nerves also, which, however they 

 might become habituated to the simple changes required in con- 

 tracting or dilating a set of pigment cells, would require considerable 

 tuition before they would learn automatically to accommodate their 

 changes to the size of the pebbles which happened to be present 

 at the bottom of the tank. Our experiments with frogs seem to 

 point in the same direction as those with fishes ; a few minutes' 

 exposure in a white vessel is sufficient to contract the dark pigment 

 cells and disclose the underlying yellow and green skin-colour. 

 The same action is traceable in many other animals, especially in 

 those which live in water. Professor Miall suggests that as cones 

 in the retina indicate perception of colour, the eyes of fishes and 

 frogs should be carefully examined. It is found that the cones are 

 comparatively scarce in mammalia, but plentiful in birds ; he says, 

 " there is reason to suppose that the cones are specially connected 

 with the colour-sense. These are abundant in and about the yellow 

 spot in man and other primates, but usually few in other mam- 

 malia." Thin slices are easily cut from a retina which has been 

 hardened mth. potassimn bichromate, and subsequently in alcohol. 

 Teased out osmic acid preparations are also good. 



We have spoken of imitation of colour only, but the same 



I 



