The evidence of facts proceeds in the following manner: First of all, there is a fact which has come down to our own time,—which may be seen with our eyes, heard with our ears, and felt with our hands. This can be understood only on the supposition of an earlier fact no longer perceptible to us. Hence such an earlier fact is admitted as having been once perceptible. This rule, that we can accept as proved only so much of the earlier fact as is absolutely necessary for the comprehension of the now-existing fact, is to be taken strictly; for it is only to the Understanding, and by no means to the Imagination, that we can concede any value in historical evidence. Why then should we attempt to educe and define the earlier fact further than is absolutely requisite for the explanation of the present? In all Sciences, and particularly in History, it is of greater importance to understand distinctly how much we do not know than to fill up the void with fiction and conjecture. For example, I read a work which is said to be Cicero’s, and till now has been universally acknowledged to be his:—this is the fact of the Present. The earlier fact to be detected herein is this:—Whether the particular Cicero who is distinctly known to us by means of other history did actually write this work. I go through the whole series of evidence lying in the interval of time between me and Cicero; but I know that herein error and illusion are possible, and this external proof of authenticity is not in itself decisive. I turn therefore to the internal characteristics: Is it the style, the mode of thinking of a man who lived at that time, who filled such a station in society, and was surrounded by such circumstances? Suppose I find these things so, then the evidence is complete:—it is not possible to conceive that this book, as it now exists, could have existed if Cicero had not written it: he was the only man who could write it thus; therefore he has written it.
Another instance:—I read the first chapters of the so-called first book of Moses, and, as must be presupposed, I
