The moral feeling, as we have seen, follows immediately and necessarily on the moral judgment. But what considerations guide the moral judgment? Our moral judgments, as we have also seen, are the result of a logical process of reference to a class or of association with similars. This particular action is like certain other actions, or belongs to a class of actions, which we habitually regard as right or wrong, and, consequently, as soon as the reference or association is made, the moral feeling supervenes. Now, in this process, there are two possible sources of error. In the first place, the act of reference or association may be faulty, and the action may not really belong to the class to which we refer or really be like the other actions with which we associate it. This fault is one of classification, and can only be remedied, as all other faulty acts of classification, by learning to discriminate between the essential and the non-essential marks of similarity, and insisting on the presence of the essential marks. In criminal cases, this is one of the functions of the jury, and, unless they exercise great care, they may easily be mistaken as to whether an alleged act of fraud, theft, assault, &c., was really an act of that kind. But, even if the action be referred to its right head, there remains the second question whether we are really justified in regarding the class of actions itself as right or wrong. Failure to prosecute for or punish heresy or witchcraft was at one time regarded at least as wrong as failure to punish or prosecute for theft or murder would now be. To decline to fight a duel was, till quite recently, to place yourself outside the pale of gentlemen. A reluctance to sacrifice herself on the funeral pile of her dead husband was, till the practice of Suttee was abolished by the British government, one of the most immoral traits which a Brahman widow could exhibit. Now, have we any means of discriminating, and, if so, how do we discriminate, between those acts which are really, and those which are only reputed, right or wrong? That there is great need of such a test, if it can be discovered, is plain. The wide divergences of opinion on matters of conduct in different ages, in different countries, in different classes of society, and even amongst men of the same class In the same country and at the same time, shew at once the vast importance of ascertaining some common measure of actions, and that there is no uniform rule of right and wrong to be found in the human mind itself. If there is such a rule, it must be derived from some external considerations, and, if there is no such rule, then morality must be, to a large extent, a matter of prejudice, fancy, and caprice. Now I conceive that there is a simple mode of ascertaining whether there is any test of actions other than the merely subjective determinations of our own minds, or, in other words, whether there are any reasons or