yet they find it pregnant with moralising and elevating characteristics.  We are thinking of a very different attitude—­that, namely, of persons who believe a creed to be not more morally useful than it is intellectually sustainable, so far as they themselves are concerned.  To them it is pure and uncompensated error.  Yet from a vague and general idea that what is useless error to them may be useful to others, they insist on doing their best to perpetuate the system which spreads and consecrates the error.  And how do they settle the question?  They reckon up the advantages, and forget the drawbacks.  They detect and dwell on one or two elements of utility in the false belief or the worn-out institution, and leave out of all account the elements that make in the other direction.

Considering how much influence this vague persuasion has in encouraging a well-meaning hypocrisy in individuals, and a profound stagnation in societies, it may be well to examine the matter somewhat generally.  Let us try to measure the force of some of the most usual pleas for error.

I. A false opinion, it may be said, is frequently found to have clustering around it a multitude of excellent associations, which do far more good than the false opinion that supports them, does harm.  In the middle ages, for instance, there was a belief that a holy man had the gift of routing demons, of healing the sick, and of working divers other miracles.  Supposing that this belief was untrue, supposing that it was an error to attribute the sudden death of an incredible multitude of troublesome flies in a church to the fact of Saint Bernard having excommunicated them, what then?  The mistaken opinion was still associated with a deep reverence for virtue and sanctity, and this was more valuable, than the error of the explanation of the death of the flies was noxious or degrading.

The answer to this seems to be as follows.  First, in making false notions the proofs or close associates of true ones, you are exposing the latter to the ruin which awaits the former.  For example, if you have in the minds of children or servants associated honesty, industry, truthfulness, with the fear of hell-fire, then supposing this fear to become extinct in their minds,—­which, being unfounded in truth, it is in constant risk of doing—­the virtues associated with it are likely to be weakened exactly in proportion as that association was strong.

Second, for all good habits in thought or conduct there are good and real reasons in the nature of things.  To leave such habits attached to false opinions is to lessen the weight of these natural or spontaneous reasons, and so to do more harm in the long run than effacement of them seems for a time to do good.  Most excellences in human character have a spontaneous root in our nature.  Moreover if they had not, and where they have not, there is always a valid and real external defence for them.  The unreal defence must be weaker than the real one, and the substitution of a weak for a strong defence, where both are to be had, is not useful but the very opposite.