What is innate in this process-the end of concealment— seems to remain substantially unchanged, but a great variety of means to it, for use on different occasions, comes to be acquired. We cannot call this inherited impulse to pursue the end of concealment an instinct of the child, because it lacks the instinctive behaviour for the attainment of this end.

It is more or less the same with the instincts of anger. The end of destruction appears to be innately given, the end also of breaking down opposition,1 and these ends remain substantially unchanged, but not being able to walk or to dispose his body in the requisite positions, the child can only hit out vaguely or push things away. Here at least is a rudiment of behaviour which will become definite and efficient later, as the child slowly acquires the modes of attack characteristic of man.

To sum up the child inherits only the capacity to perform a few very simple types of instinctive behaviour connected with his appetites and primary emotions; but he inherits all the biological ends of these systems. Most of these instincts are useless fragments until combined with other means which the mother sometimes supplies and the child gradually acquires. These fragments, where sufficiently definite to be called instincts, appear to remain substantially invariable throughout life, though becoming expanded, through their acquired factors, into more complex and efficient systems. The acquired parts of these systems retain the variability which they had from the beginning, except as they are checked by the force of habit.

The primary emotions of man may utilise any of these simple human instincts with their acquired and variable factors, where serviceable to their ends. But the variability of their behaviour is, as we have seen, of a different nature; because they are not confined to the end of any one of these instincts, nor to what is instinctive in their behaviour, learning to pursue their more general ends by the means which in the situation appear the most appropriate.

1 See supra, pp. 225, 226.

6. Whether all Instincts have Distinctive Emotions to Support them when in Difficulties

Since impulses are always, and emotions only sometimes, involved in the operation of instincts, we may next enquire (1) under what conditions these emotions are aroused; (2) what these emotions are; (3) whether each one of them is peculiar to the operation of a particular instinct. We shall answer these three questions together. The operation of every instinctive impulse which is sufficiently strong, tends to arouse anger when it is obstructed, joy or satisfaction when its end is attained under difficulties, and sorrow when it is completely frustrated; but no one of these emotions distinguishes the operation of one instinct from that of another. Now if such impulses were desires-which they frequently become in human beings when, being repeatedly performed, they acquire foresight of their endsthe operation of such desires, in the face of difficulties and change of situations, would tend to give rise to a fresh group of emotions, those we have called the 'prospective,' which are peculiar to desire. They are hope, despondency, anxiety, disappointment, confidence and despair. But again, no one of these emotions would serve to distinguish one desire from any other, but would remain liable to be excited in all.

We should then be disposed to think that when the impulse of some instinct, through the obstruction to which it was subjected, or the shock of surprise, gave rise to an emotion, it would not be one distinctive of this instinct, but the property of others also. How comes it then that any instinct can arouse an emotion more or less peculiar to itself? That it can do so in certain cases is clear. For we have only to think of any of the instincts by which different species of animals conceal themselves, to think also of fear; or of any of their fighting instincts, to think also of anger. And although a concealment-instinct may be connected with the fighting instincts of some animals

(as with certain spiders that watch in a hidden place for the approach of their prey, and with certain crabs that hide in the crevices of rocks, or cover themselves with seaweed so as to remain unseen), and a fighting instinct may be connected with fear (as when an animal fights at bay), yet, as a general rule, the fighting-instincts are connected with anger, and concealment-instincts with fear. Now with regard to the primary emotions of fear, anger, and disgust, we have found that they give an additional support to the instinctive tendencies operating with them when in difficulties; because, beside the increase of nervous energy that accompanies them, they possess the ends of these tendencies in a more general form, and have alternative methods for their attainment. The instincts that may arouse some emotion more or less peculiar to themselves would seem to be those for which primary emotions exist that pursue their ends in a more general form-as the end of escape by clinging to another, or by concealing oneself, is contained in the more general end of 'escape of the emotion of fear. The instincts that are not capable of arousing any emotion peculiar to themselves would seem to be those for which there is no primary emotion that can be shown to contain their ends in a more general form.

We have tried to show in B. II, Ch. 1, that the various instincts of locomotion are of use because they enter as constituents into the systems of the appetites and primary emotions. There is therefore no one primary emotion peculiar to the locomotory instincts; and there is no one primary emotion whose end is a more generalised form of their end.

We shall reach the same conclusion if we consider some of the most complicated instincts as the web-spinning and nest-building instincts. The instinct of some species of spider to make a web, or that of the solitary wasps to make nests for their future progeny, may take a long time to accomplish their task. Great care and perseverance are needed in the execution of these structures, and all the intelligence of the animal appears to be at the service of its instinct. Not only the first rude begin

nings of the spider's web are often destroyed in a night, but but the half-finished work also, when the spokes are not all in their place, nor the transverse stays that strengthen the web and make it an efficient instrument for catching the prey. In making this great construction many obstacles may have to be overcome, and partial failures to be retrieved. The primary emotions of anger or fear may be sometimes excited in the spider's mind; but these do not serve to distinguish this instinctive process from any other. In place of desire with its prospective emotions, sustaining the process until the end is attained, we must rather assume the presence of an instinctive impulse, persisting through a long period of time and under varying conditions, therefore subject to changes of its affective tone, becoming pleasanter or less pleasant or disagreeable, capable of arousing at times the common emotions of anger and fear; but nowhere should we be able to indicate, or have grounds for assuming, the presence of any emotion distinctive of the process as a whole.1

If we cannot accept Dr. McDougall's theory that the principal instincts, when in operation, elicit an emotion more or less distinctive of them-which seems to be based on a confusion between emotion and impulse-can we any more accept Dr. Drever's modification of it, that the "great instincts of human nature have all their accompanying and typical emotions "2?-although he admits that these emotions are only elicited under "tension."8 If we begin by assuming that fear, anger, disgust, curiosity are among these great instincts, then it is easy to indicate their typical emotions. We have tried to show that these primary emotional systems are not instincts, though having many points in common with instincts.

1 Dr. Drever, who must know that, in such a case as we have supposed, we adopt the hypothesis that best explains the facts, to my former denial that we can indicate any emotion distinctive of the nestbuilding instincts, says, 'The obvious rejoinder is we are in no position to say whether there is a distinctive emotion involved' or not. ('Instinct in Man,' p. 159.) From this point of view we must say nothing about the emotions of animals, not even that they exist! 2 Op. cit., P. 161.

• P. 157.

APPENDIX II

OF TWO OPPOSITE TENDENCIES OF JOY AND SORROW

1. Of the Expansive Tendency of Joy

THERE is an important tendency of joy of which no account was given in the preceding work, and in studying it here we shall find the popular opposition between joy and sorrow clearly exemplified. It is the healthy and expansive tendency of joy that is in marked contrast to the concentrated and narrowing tendency of sorrow.

Joy is an expansive emotion.1 When a man rejoices because he has won some honour, or a fortune, or the woman he loves, his joy, though at first centred and concentrated in its proper object, soon shows a tendency to spread to other things. The streets seem more interesting; the country more beautiful. Things that were dull and repugnant to him seem changed for the better. He meets his acquaintances with a glad smile. He becomes more sociable and more sympathetic; coldness and reserve are thrown off. Hence it is that times of good fortune when there is joy in a man's heart are specially chosen for feasts and entertainments. For a man never wants to shut himself up for long with his joy as he does with his sorrow, but soon goes off to communicate it to others.

1 I have to thank Prof. Boyce Gibson for calling my attention to this tendency of joy in an able and sympathetic article on 'The Foundations of Character' in Mind (N.S. Vol. xxv. No. 97). I had also incidentally noticed it in a chapter on 'The Sources of Tender Emotion' in Prof. Stout's Groundwork of Psychology' (Ch. xvi. p. 207).