that account receive a shock at the sight of the calamities of others. It is true that a very keen and absorbing sentiment of benevolence would be in itself sufficient to divert men from any habit that impaired their power of gratifying it. It is true that compassion involves a certain amount of pleasure, and conceivable that that pleasure might be so intensified that we might seek it by a crime. The error in these theories is not that they exaggerate the possible efficacy of the motives, but that they exaggerate their actual intensity in human nature and describe falsely the process by which the results they seek to explain have been arrived at. The function of observation in moral philosophy is not simply to attest the moral sentiments we possess, leaving it to the reason to determine deductively how they may have been formed; it is rather to follow them through all the stages of their formation. And here I may observe that the term inductive, like most others that are employed in moral philosophy, may give rise to serious misconception. It is properly applied to those moralists who, disbelieving the existence of any moral sense or faculty revealing to us what is right and wrong, maintain that the origin of those ideas is simply our experience of the tendency of different lines of conduct to promote or impair true happiness. It appears, however, to be sometimes imagined that inductive moralists alone think that it is by induction or experience that we ought to ascertain what is the origin of our moral ideas. But this I conceive to be a complete mistake. The basis of morals is a distinct question from the basis of theories of morals. Those who maintain the existence of a moral faculty do not, as is sometimes said, assume this proposition as a first principle of their arguments, but they arrive at it by a process of induction quite as severe as any that can be employed by their opponents.(99) They examine, analyse, and classify their existing moral feelings, ascertain in what respects those feelings agree with or differ from others, trace them through their various phases, and only assign them to a special faculty when they think they have shown them to be incapable of resolution, and generically different from all others.(100)