Although it might seem from this as though Aristotle placed the happiness of man in the natural activity of the soul, and regarded this as self-sufficient, still he is not blind to the fact that perfect happiness is dependent on other kinds of good whose possession is not absolutely within our power. It is true he expresses an opinion, that outward things in moderation are sufficient, and that only great success or signal reverses materially influence the happens of life; still he holds that wealth, the possession of friends and children, noble birth, beauty of body, etc., are more or less necessary conditions of happiness, though these are partly dependent on accidental circumstances. These wavering and inconsistent views of Aristotle respecting the nature of happiness, naturally rise from his empirical method of investigation. Careful in noting every thing which our limited experience seems to utter, he expressly avoids making either virtue or pleasure his principle, because actual experience shows the separation of the two. Although therefore he gives directions in general to strive after that pleasure in which the good man delights, or which is connected with a virtuous activity, yet is pleasure with him an end for its own sake, and not merely an accident of virtue, an empiricist, Aristotle is here also a dualist, while the Stoics and Epicureans have respectively taken and held fast to each of the two sides. 3. CONCEPTION OF VIRTUE.—As has already been seen in the Aristotelian Polemic against Socrates, virtue is the product of an oft-repeated moral action, a condition acquired through practice, a moral dexterity of the soul. The nature of this dexterity is seen in the following way: every action completes something as its work; but now if a work is imperfect when it has either a want or a superfluity, so also is every action imperfect in so far as there is in it either too little or too much; its perfection, therefore, is only found as it contains the right degree, the true mean between the too much and too little. Accordingly, virtue in general may be explained as the observation of the right mean in action, by which is meant not the arithmetical or absolute mean, but the one relative to ourselves. For what is enough for one individual is insufficient for another. The virtue of a man, of a woman, of a child, and of a slave is respectively different. Thus, virtue depends upon time, circumstance, and relation. The determination of this correct mean will always waver. In the impossibility of an active and exhaustive formula, we can only say respecting it that it is the correct mean as determined by a correct practical insight which is seen to be such by the intelligent man.