316 THE EVOLUTION OF ANIMAL BEHAVIOUR 
avoidance of pain. This is the purpose of behaviour as 
viewed from the psychological aspect. The biological end of 
animal conation is racial survival; its psychological end is 
individual satisfaction. And the two ends are, in the main 
and broadly speaking, consonant—a result which would un- 
questionably be secured by natural selection, but is on any 
other naturalistic hypothesis difficult of explanation. 
But the two ends are not only consonant ; they are supple- 
mentary one to the other. During much of the life of the 
higher animals there is no need, immediately present and 
pressing, for the output of action to meet biological ends. 
There are periods of life and intervals of time when the sharp 
incidence of the struggle for existence does not call for the 
serious business of behaviour. But at these periods and in 
these intervals the animal is not inactive ; indeed, it is restless 
in its activity. Unless it be weary with unwonted exertion, 
or basking in the psychical sunshine of content, due to the 
unsought advent of pleasant stimulation or the after-effects 
of previous behaviour (for example, when hunger. has been 
relieved), the healthy animal must be up and doing. This 
familiar fact no doubt affords the basis in observation of the 
surplus-energy theory of play. But is it necessarily surplus 
energy? Is it not,rather normal energy which expends itself 
in this way when there is no immediate and serious biological 
business on hand? And, as Professor Groos has pointed out, 
play is seen when we have every reason to suppose there is no 
surplus energy, nay, even when the normal energy is at a low 
ebb. There is no more pathetic sight than a sick kitten, with 
energy obviously much below par, utilizing its little remaining 
strength in feeble attempts to play. 
It is unnecessary to do more than remind the reader of the 
theory elaborated with so much skill and care by Professor 
Groos, that the forms assumed by play—in which, it will be 
remembered, he includes a very wide range of behaviour— 
have a very important indirect biological end in practice and 
experimentation. Our present point is, that its direct psycho- 
logical end is the satisfaction it affords. Without this the 
