This page needs to be proofread.

TESTIMONY AND AUTHORITY. 67 not occurred or have not been noticed in the requisite relation. The facts capable of introducing order and uniformity into the irregularities and caprices of our own individual experience are then so often not facts of that experience, or are, if there, so often overlooked, that the emphasis and selection constituting the perception of uniformity must to a large extent be performed by deputy ; a substitute for direct experience in the interpretation of other experience must be found in the use of Testimony. To this claim a seemingly strong objection, which has oftentimes been pitted against the empirical basis of Induc- tion, might be urged thus : How can criteria of testimony be included among the principles of induction when they have themselves to be established by induction ? Only under misconception, perhaps generated by false analogy to the alternate deduction of premiss and conclusion each from the other, could this be thought to be illegitimate. The apparent difficulty disappears as soon as we distinguish the successive stages in the growth of the power of reason- ing inductively. First we find the inductive processes im- plicitly occurring in mental operations long before they become explicit. On a higher plane there are the explicit and fully developed processes. Lastly, the organised logical theory in which the principles are enunciated, their inter- dependence exhibited, their legitimate extent and necessary limitations defined ; by which they are reduced to mutual consistency and precise accordance with experience. The perfecting of the principles is the proper business of Logic, not the origination of the processes. Exact principles of induction result from the working of the primitive processes, as, in the grinding of lenses, a truly spherical form results from the mutual attrition of surfaces initially imperfect. If the relation between inductive processes and logical prin- ciples were like that between conclusion and premiss the objection would be fatal. But it is not so : rather does Logic take the imperfect processes, grind them, so to speak, one against the other, and hand them back as nearly as may be in the form of perfected principles. II. CRITERIA OF TRUSTWORTHINESS. Excessive credulity and excessive incredulity have each been fixed upon as the marks of ignorance and simplicity. Within the narrow circle of personal experience the unin- formed person exhibits obstinate prejudice ; outside that circle childlike dependence upon others. A theory of testi-