- sion of truths of equal, if not superior, authority to those of
either natural or moral science, we shall have performed a task which may not be wholly useless or altogether uninstructive.
Our first business, in such an inquiry as this, should be to determine the method on which it ought to be conducted. In analyzing the religious systems of the world, the question of method is all-important. Indeed, it will be abundantly evident in the course of the ensuing investigations that the conclusions reached by those who have cultivated this field of knowledge have often been unsound, simply because they have failed to pursue the only proper method. Nothing can be easier, for instance, than to construct elaborate systems of religious philosophy, the several parts of which hang so well together that we find it; difficult to urge any solid objection against them, while yet the whole edifice rests upon so insecure a foundation that at the least touch of its lowest stones it will fall in ruins to the ground. This too common mistake arises from the fact that the first principles of the system are assumed without adequate warrant, and will not bear examination. Half, if not many more than half, the common errors of believers in the various current creeds are due to a similar cause. These persons start from some principle which they conceive to be indisputable, and proceed to draw inferences from it with the most complete confidence. An extreme instance of this is mentioned by Dr. Sprenger, who was asked by a Musselman how he could disbelieve the religion of Islam, seeing that Mahomet's name was written on the gates of paradise. In a less palpable form, the same mode of reasoning is constantly adopted among ourselves. Either we do not take the trouble to submit the evidence of the facts upon which we erect our arguments to a sufficiently rigorous scrutiny, or we fail to perceive that the axioms we take for granted are in reality neither self-evident, as our system requires, nor capable of any satisfactory demonstration.
Another and perhaps scarcely a less common kind of error arising from defective method is a failure to distinguish between adequate and inadequate evidence of religious truth. A sound and exhaustive method would not fail to disclose, if not what kind of evidence is sufficient, at least what kind of evidence is
