I will begin by taking an illustration of this from the impulses commonly placed lowest in the scale. The appetite of hunger, so far as I can observe, is a direct impulse to the eating of food. Such eating is no doubt commonly attended with an agreeable feeling of more or less intensity; but it cannot, I think, be strictly said that this agreeable feeling is the object of hunger, and that it is the representation of this pleasure which stimulates the will of the hungry man as such. Of course, hunger is frequently and naturally accompanied with anticipation of the pleasure of eating: but careful introspection seems to show that the two are by no means inseparable. And even when they occur together the pleasure seems properly the object not of the primary appetite, but of a secondary desire which can be distinguished from the former; since the _gourmand_, in whom this secondary desire is strong, is often prompted by it to actions designed to stimulate hunger, and often, again, is led to control the primary impulse, in order to prolong and vary the process of satisfying it. Indeed it is so obvious that hunger is something different from the desire for anticipated pleasure, that some writers have regarded its volitional stimulus (and that of desire generally) as a case of aversion from present pain. This, however, seems to me a distinct mistake in psychological classification. No doubt desire is a state of consciousness so far similar to pain, that in both we feel a stimulus prompting us to pass from the present state into a different one. But aversion from pain is an impulse to get out of the present state and pass into some other state which is only negatively represented as different from the present: whereas in desire as such, the primary impulse is towards the realisation of some positive future result. It is true that when a strong desire is, for any reason, baulked of its effect in causing action, it is generally painful in some degree: and so a secondary aversion to the state of desire is generated, which blends itself with the desire and may easily be confounded with it. But here, again, we may distinguish the two impulses by observing the different kinds of conduct to which they occasionally prompt: for the aversion to the pain of ungratified desire, though it may act as an additional stimulus towards the gratification of the desire, may also (and often does) prompt us to get rid of the pain by suppressing the desire.