transfer itself to other things without being divided against itself? Yet joy does possess both tendencies, and we shall now attempt to understand how they are related.

The law of the conservative tendency does not mean that joy in maintaining the present situation tends to hinder all movement and change of its object; for that conclusion would only follow if by 'present' was meant the mathematical present which excludes duration. Whereas what is meant is the psychological present which includes it. The object of joy, then, is something which is both liable to change and in many cases undergoing a change that we notice. If we have joy in the presence of another person we notice that his expression and gestures change, and that the direction of our eyes and our attention change with them. The conservative tendency of joy tries to maintain the present state of change on which its joy is dependent, but is opposed to any other that would diminish it. If a friend in the midst of our enjoyment of his company gets up to go away we say, 'Don't go away'; if he begins to be angry, we say 'Don't be angry.'

The conservative tendency of joy cannot alone initiate progress, but may welcome it when it has come. If, for instance, some thing that is the present object of joy becomes more beautiful, interesting or attractive to us our joy is increased as when some change of the atmosphere makes the scene before us more beautiful. But, were we able to, we could not initiate such change if our joy felt complete.

Our joy in the presence of something is often not complete. We watch a woman dance. We have joy in the grace and beauty of her movements; but some of them are ungraceful, and these displease us and detract from our joy. There is some anger in this displeasure, because joy detains us in presence of the object; and the impulse of this anger is to destroy or put an end to these defective movements: 'Stop them!' we should like to call out. But since on the whole the dancing is beautiful, the conservative impulse of joy tends to maintain the present state; and the natural exclamation of this joy would be, 'Go on,'

'Encore!' The desire for progress therefore implies that we feel the presence of something defective in the existing situation, and that joy, if there is joy, is not complete.

We have seen in what way the conservative tendency of joy acts so as to maintain the present state of the object, let us now notice how it acts so as to maintain the perception or thought of the object and the attention directed to it, and whether this excludes change. We notice when we have joy in looking at some motionless thing, as a beautiful vase, how our eyes and attention are detained by it, yet are not themselves motionless. Attention passes from the whole to the detail, from one detail to another, from the form to the colour till we have surveyed its various beauties. Why all this change? Why after having directed attention to some part of the object and felt joy in looking at that do we not maintain the attention to it unchanged, as long as the joy is felt? It is perhaps that the other parts of the object catch the attention, and that the movement of attention from one to the others maintains our joy longer than would otherwise be the case. It was in making an observation of this kind that La Rochefoucauld reached his curious reflection that constancy in love is a kind of inconstancy, since our thought does not remain fixed in contemplating a single charm of the beloved person, but passes from one to others.

Perhaps the result will be different if we take as simple an object of joy as we can find. Let it be the joy of looking at some luminous colour-as the crimson colour of wine looked at through a glass with the light behind it. Here the movements of the eyes and the attention are more restricted, but there is still movement. The eyes do not remain fixed on the same point for more than a few moments, but wander to contiguous points, getting a fuller knowledge of the whole from a knowledge of the parts. It is by such constant change or movement that joy maintains the attention to its object.

Why, then, is there this movement within the system of joy, and why is joy sustained longer with a complex object and especially with one that changes than with one that

does not appear to change? For people will gaze for long at a waterfall or at the waves breaking on the shore. It is because the attention and thought are naturally mobile, and have an innate impulse to change which we feel whenever we attempt to arrest their movement. It is difficult to hold the attention for long to any point of the visual field, or to maintain the same thought. After a few moments the effort of maintaining it becomes disagreeable and extinguishes the joy. But if the object is complex, and thought can move over its different parts and follow its changes, the joy is prolonged.

Let us then formulate this law: (81 G) The duration of the joy felt in the presence or thought of an object tends to be greater other things equal—in proportion as the complexity of this object facilitates change of attention and thought: the simpler it is and the less it facilitates such change the less the duration of the joy.

4. Why Joy has an Expansive Tendency

We are now coming within sight of the expansive tendency of this emotion. The curious way that our thought has of moving about the different parts of its object and not resting anywhere for more than a few moments, whatever the degree of joy felt in them-this, though not the same thing as the expansive tendency of joy, is clearly akin to it. In defining the law of this tendency we said that "joy though at first confined to its own object tends afterwards to spread to other things." We can now understand why it is at first confined to its own object and why it afterwards spreads to other things. The same universal tendency of thought and attention to change accounts both for the change from one part of the object to others and for the expansion beyond this object afterwards. Joy, with the surprise that so often accompanies it, has an arresting influence on the mobility of attention and thought, and at first confines them to its object. But apart from the change and complexity of this object this arrest would be

of short duration.

Even under these conditions the conservative impulse of joy is partly in conflict with this universal tendency to change. The expansive tendency of joy is one solution of this conflict. It is the effect on the emotion of this tendency to change, sustaining joy longer by obliging it to desert for a time its present object and to attach itself to other things. The simpler the object and the more unchanging the sooner does this expansion take place. We speak of the monotony of such things and how soon they fatigue us. What is the explanation of this monotony and fatigue? It is that the arrest of attention to them, in fatiguing the mind, extinguishes the joy. The law of change is little or but partly capable of being harmonised with the conservative impulse of joy, and least of all where the object of joy is apparently simple and changeless.

If the expansive tendency has been forced upon joy by this most powerful tendency of thought and attention to change, we cannot regard it as part of the proper conation of this emotion. For it is not natural to joy to part from its own object, but to maintain the present union with it. This its conservative impulse tries to do, but it can only prevail for a little time against this universal tendency, or, if for a little longer, only so far as its object facilitates change and re-distribution of attention; and if it persist longer in the contest joy is extinguished. We can then understand why our joys last for such a little time and why even happiness, if it sometimes lasts longer, is so little present to consciousness. For though we try again and again to remind ourselves of this happiness, still we cannot go on repeating the same thought, or succeed in holding it long before the mind. It disappears and carries our conscious happiness along with it. And when it has passed, in reflecting on it we say so often, 'How happy I was then and how little I realised it!'

There are other conditions on which the expansive tendency of joy depends. The tendency of thought and attention to change is manifested by all our emotions and sentiments, but these organise it along their own lines in the

pursuing the means, change David Hume recognised this "The imagination," he says,

pursuit of their own ends.1 The emotions and sentiments are also prone to change, but do not change so rapidly as the thought and attention they employ. For they have to remain constant while these, from one object to another. contrast in our mental life. "is extremely agile; but the passions, in comparison, are slow and restive."2 It is partly through the comparative steadiness of the emotion that joy obtains its expansive tendency. For if our joy is chiefly perceptual-as when we rejoice in being re-united to a beloved person after long absence the joy lasts longer than the perception of him; for he may leave us for a little time, and when present, the eyes and attention wander at times and are caught by contiguous things. And if our joy is also conceptual, and dwells in thought, on the longed for re-union, still thoughts are so fluid that they can only be kept fixed for a few moments. Thus the joy tends to remain for some time after the perception and thought of its object have been displaced.

Still, this does not explain why the expansive tendency, though in some measure attaching to other emotions, yet chiefly characterises joy. Joy is more opposed to the tendency of attention and thought to change than any other primary emotion. Whereas other emotions have new results or ends to attain, and when these are attained themselves subside and come to an end, and set free attention and thought; joy only begins when what seems to be its end is attained, and has thenceforward to maintain the present state as long as its emotion is felt. It therefore can only give play to the movement of attention and thought within the limited complexity of its object, and as soon as this is exhausted they are transferred elsewhere. Thus the expansive tendency is chiefly characteristic of joy, but if other emotions have less of it, why has sorrow an opposite or contracting tendency upon the mind?