134 MIND IN EVOLUTION CHAP. 



many branches of which evidence, more or less abundant, 

 has been accumulated by careful experimenters in recent 

 years. To summarise the evidence, however imperfectly, 

 will perhaps help us a little further towards understanding 

 the general course of Mental Evolution. 1 



We have already quoted instances showing that among 

 certain fish, at any rate in some individuals, the modifying 

 influence of experience is at a minimum. Even the 

 simplest sort of modification, the inhibition of an impulse, 

 fails in certain cases, as in those quoted by Mr. Bateson, 

 where after twelve months' constant experience, fish per- 

 sisted in knocking their heads against the glass wall of their 

 tank. On the other hand, there is no doubt that certain 

 fish learn to come to the surface for food on the approach 

 of human beings, and perhaps at the sound of a voice or of 

 a bell. This is a simple case of an acquired reaction of 

 the sort described in this chapter. Somewhat more difficult 

 is the avoidance of baits. Romanes states generally that there 

 is a marked increase of wariness in waters which are much 

 fished, and attributes the change to " observation," on the 

 ground that "young trout under such circumstances are less 

 wary than old ones." If the trout really draw an inference 

 from their observation of others being caught and dragged 

 to the surface, they are capable of a much more complex 



1 It must be remembered that in the matter of intelligence there is 

 perhaps even more difference between the individuals of a species, 

 the species of a genus, and even between the Orders of a Class, than 

 in the matter of physical structure. A classification of the animal' 

 kingdom based on intelligence would probably cut right across the 

 classifications based on structure. It would probably associate the nest- 

 building stickleback with the lower birds ; it would class the Cephalopoda 

 with Fish and Reptiles rather than among the Molluscs ; and, whatever 

 it did with ants and bees, it would draw a well-marked line between them 

 and most of the insects. Intelligent acts are, above all others, "adaptive" 

 characters characters taken up by the organism in immediate response 

 to its surroundings, and such characters as Darwin showed are the least 

 to be depended on as tests of genealogical affinity. Hence any general 

 statement as to the intelligence of, say, Fish or Birds, can only mean at 

 the utmost that the kind of intelligence in question is very widely diffused 

 among Fish or Birds so widely, perhaps, that it may be regarded as 

 representing the mean level in the Class. Some species, and perhaps 

 whole Orders, may rise above or fall below it, while within each species 

 the individual differences will often be much greater than they are in 

 matters of visible structure. This qualification must be understood as 

 applying to any general statement in the text. 



