it has learnt, or invents on the occasion, after having reflected. on the possible alternatives, and judged that best. Nor, again, does it persist blindly in the conduct which it has adopted, but, watchful of its effects, is ready to change it, when unsuccessful, for some other. This selectiveness, control and judgment, fear only attains to through the influence of the sentiments in which it comes to be organised; but once attained, the emotion is placed in supreme control over the instinctive and other forces of the system; and its abstract and general end to direct their particular and proximate ends.

We shall, therefore, enunciate the following law: (35) The instincts which at first dominate the system of fear tend, in the mental development of man, to fall under the control of the emotional side of the system.

Thus, having learnt subordination to the sentiment, the emotion brings all its own tendencies into subordination to itself. Through the course of this development, fear gains in flexibility, and power to adapt itself to the new forms of danger that arise; its end, too, being so abstract, and merely to escape from some threatening danger, admits of a variety of interpretations according to the situation. For what is not, or what may not be, a danger to such a creature as man, who loves and hates so many things!

6. Of Fear as one of the Primary Emotions and Root-Forces of Character

There are several tests we may apply to determine whether or not a given emotion is primary, and whether the force which it has belongs originally to its own system, or is derived from any other. (1) There is the time at which the expression and behaviour of the emotion is first observed in child-life; for although some emotions are primary which arise at a late period, as the appetite of sex; yet when an emotion is manifested in the first months of child-life, that is evidence of its primary nature. (2) The second test is the diffusion of an emotion in the animal-world; the more widely diffused, the more is the probability increased that it is a primary emotion. (3) The third test is whether we can, by analysis, reduce it to other emotions. (4) The fourth is, whether the emotion in its

earliest forms is manifested in genuinely instinctive behaviour, and is at first instinctively aroused by sensory stimuli. From all these points of view we must conclude that fear is one of the primary emotions and root-forces of character. It is the first, or one of the first emotions to manifest itself in child-life. It is very widely diffused in the animal-world. It is unlikely that we shall ever succeed in reducing it to more rudimentary emotions. And, lastly, it is both instinctively aroused, and manifests itself in a variety of instinctive forms of behaviour in different animals.

This opinion dates only from modern times; and one of the most remarkable differences between ancient and modern writers on the emotions is the denial by the former that fear and anger are primary emotions. They are found neither in the lists of the primary emotions of Descartes and of Spinoza, nor, to come to more modern times, in those of Hutcheson and Hume. But these writers except the last, make no careful attempt to trace fear to the primitive emotions at its source. A curious error runs through all of them. Too much influenced by introspection, they take into account only the later or ideational fears which spring from desire, and overlook the primitive forms aroused by sensations. Thus Spinoza defines Fear as an "inconstant pain arising from the idea of something past or future." That he overlooks the earlier forms is clear from his conclusion that there is "no hope unmingled with fear and no fear unmingled with hope."2 Descartes also treats these two emotions as if they were inseparable, and likewise connects them with desire. He defines Hope as "a disposition of the soul to be persuaded that what it desires will come to pass," and Fear as "another disposition of the soul" to be persuaded that what it desires "will not come to pass."3 But while he offers an ingenious derivation of Hope as due to a mixture of "joy and desire," he makes no complementary attempt to resolve Fear. For Hope is a later emotion, dependent both on ideas and on desire, and therefore offers

1 'The Ethics,' part iii. 'The Definitions of the Emotions,' xiii.

2 Ibid.

3 'Les Passions de l'Ame,' 'Troisième Partie,' Art. 165.

some prospect of resolution. Still the inference is suggested by Descartes' account of Hope, that Fear must be some mixture of sorrow and desire. And this conclusion is expressly drawn by Hutcheson, and adopted with certain modifications by Hume. The former substitutes the term " aversion for " desire" and defines Fear as a mixture of sorrow and aversion." But Hume alone attempts to furnish a conclusive proof of the derivative character of Fear, and one that might be regarded as scientific and experimental. Like his predecessors, he overlooks the primitive manifestations of fear, and couples it with hope, regarding both as arising "from the probability of any good or evil." 2 "Suppose then," he says, "that the object, concerning which we are doubtful, produces either desire or aversion; it is evident that, according as the mind turns itself to one side or the other, it must feel a momentary impression of joy or sorrow." And he proceeds with a fine psychological observation : "The imagination is extremely quick and agile; but the passions, in comparison, are slow and restive: For which reason, when any object is presented, which affords a variety of views to the one and emotions to the other, though the fancy may change its views with great celerity, each stroke will not produce a clear and distinct note of passion, but the one passion will always be mixed and confounded with the other. According as the probability inclines to good or evil, the passion of grief or joy predominates in the composition; and these passions being intermingled by means of the contrary views of the imagination produce by the union the passions of hope or fear."

But this proof is not cogent enough for him; he endeavours to complete it by making joy and sorrow pass and re-pass into hope and fear in the laboratory of his imagination. "Throw in," he says, "a superior degree of probability to the side of

1 'The Nature and the Conduct of the Passions and Affections,' sect. iii

2 'Essays,' A Dissertation on the Passions,' sect. I, 3. Compare also 'A Treatise of Human Nature,' book ii., 'Of the Passions,' part iii. sect. ix.

3 Ibid.

grief, you immediately see that passion diffuse itself over the complexion, and tincture it into fear. Encrease the probability, and by that means the grief; the fear prevails still more and more till at last it runs insensibly, as the joy continually diminishes, into pure grief. After you have brought it to this situation, diminish the grief, by a contrary operation to that, which increased it, to wit, by diminishing the probability on the melancholy side; and you will see the passion clear every moment, till it changes insensibly into hope; which again runs, by slow degrees, into joy, as you encrease that part of the composition, by the encrease of the probability. Are not these as plain proofs, that the passions of fear and hope are mixtures of grief and joy, as in optics it is a proof, that a coloured ray of the sun, passing through a prism, is a composition of two others,

Striking and original as is this attempt of Hume to apply an experimental method in psychology, it is not conclusive. According to this theory there is no possibility of fear arising until the imagination represents probable views of the future. Now hope and despondency are subject to some such conditions; fear is not. And paraphrasing his argument we may use it to enforce a different conclusion. Throw in,' we may say, 'a superior degree of probability on the side of grief, you immediately see that passion diffuse itself over the complexion, and tincture it into despondency. Encrease the probability, and by that means the grief, and the despondency prevails still more and more, till at last it runs insensibly, as the joy continually diminishes, into pure grief.'

Experiments in the imagination, however, necessary to the novelist and dramatist, are not conclusive evidence in psychology, nor would they carry any degree of conviction except in the hands of men of imaginative genius. Whether joy and sorrow when blended in the process of desire-if they ever be produce even the contrary emotions of hope and despondency is doubtful, that they would originate fear is in the highest degree improbable. But these abortive attempts to account for an emotion that has been held to be the first to manifest itself in child-life, suggest one conclusion. Sorrow has been one of the emotions consistently employed

in these attempts, and sorrow seems to bear a greater resemblance and affinity to fear than to the other painful and primary emotion of anger. In melancholia, sorrowful dejection and fear are frequently found together, as if the pathological conditions of the one involved extreme susceptibility to the other. The prospect of death is apt to afflict us with both. But we are not here concerned with the relations between them. For us the conclusion of modern writers that fear is one of the primary emotions and rootforces of our nature must be held to be substantiated.