THE SENSE-ORGANS AND PERCEPTION OF FISHES. 243 



moving on tlie other side. Any small oscillating substance may 

 attract themj such as a button dangling to a thread ; and pollack (6r. 

 pollacMus) often snap at even a curl of smoke from a pipe. After 

 repeated attempts to take food on the other side of the glass they 

 will desist ; but some of the oldest inhabitants (plaice, pollack, and 

 bream), which have been living in the aquarium for about a year, 

 will perseveringly try again the next time. Fishes brought newly 

 to the aquarium injure themselves by trying to escape through the 

 glass, and I have seen gurnard fretting themselves for hours against 

 it when the water of the tank has been made turbid by pouring in 

 sand, being evidently of opinion that it is a way into clearer waters. 

 It may here be suggested that perhaps the result of the famous ex- 

 periment of Mobius has been wrongly interpreted. The story runs 

 that pike, having lived for some time in a tank separated by a glass 

 plate from another in which small fish were living, desisted from 

 trying to catch them, and on the glass plate being removed never 

 attempted to do so. The suggestion is that the pike had come to 

 believe these particular fish to be under special protection. While 

 this may be so, it is nevertheless a fact that fish, like other animals, 

 having grown accustomed to the presence of forms which they 

 would naturally eat, do not molest them. On one occasion several 

 pollack were put into the congers' tank at Plymouth, and in the 

 morning two only remained, but these two continued undisturbed 

 for a long period ; and other similar cases have been observed. The 

 explanation should perhaps be referred to that paradoxical instinct 

 which is widely developed among animals of many kinds, in obedience 

 to which they occasionally do not eat or molest those with whom 

 they are constantly associated. It is, of course, this unexplained 

 instinct upon which the " happy family " of the travelling showman 

 is constructed. Probably it is closely akin to many feelings and 

 superstitions of which we are ourselves conscious, and which have 

 received inadequate but rational explanations. 



Many of the actions of fishes are of this paradoxical character. 

 It is a common thing, when two fish swim up to the same worm, for 

 the foremost to retire in a nervous way, leaving the worm for the 

 other ; and this quite independently of the relative sizes of the in- 

 dividuals. A small cod whose gills were injured lived for some 

 time in a tank with bream and bass. This fish rarely if ever ate 

 anything, but always swam up for a moment to each piece of food 

 as it was put in, and then left it. When the cod approached, the bass, 

 though many times his size, used to fall back, and return to eat the 

 food when the cod retired. This process would be repeated again 

 and again, and happens so often in the case of bream and bass that 

 it appeared almost the rule for a fish to refuse at food if another 



