s and that _stratum_. For
in all historical inquiries we are dealing with facts which themselves
come within the control of human will and human caprice, and the
evidence for which depends on the trustworthiness of human informants,
who may either purposely deceive or unwittingly mislead. A man may lie;
he may err. The triangles and the rocks can neither lie nor err. I may
with my own eyes see a certain man do a certain act; he may tell me
himself, or some one else may tell me, that he is the same man who did
some other act; but as to his statement I cannot have absolute
certainty, and no one but myself can have absolute certainty as to the
statement which I make as to the facts which I saw with my own eyes.
Historical evidence may range through every degree, from the barest
likelihood to that undoubted moral certainty on which every man acts
without hesitation in practical affairs. But it cannot get beyond this
last standard. If, then, we are ever to use words like race, family, or
even nation, to denote groups of mankind marked off by any kind of
historical, as distinguished from physical, characteristics, we must be
content to use those words, as we use many other words, without being
able to prove that our use of them is accurate, as mathematicians judge
of accuracy. I cannot be quite sure that William the Conqueror landed at
Pevensey, though I have strong reasons for believing that he did so. And
I have strong reasons for believing many facts about race and language
about which I am much further from being quite sure than I am about
William's landing at Pevensey. In short, in all these matters, we must
be satisfied to let presumption very largely take the place of actual
proof; and, if we only let presumption in, most of our difficulties at
once fly away. Language is no certain test of race; but it is a
presumption of race. Community of race, as we commonly understand race,
is no certain proof of original community of blood; but it is a
presumption of original community of blood. The presumption amounts to
moral proof, if only we do not insist on proving such physical community
of blood as would satisfy a genealogist. It amounts to moral proof, if
all that we seek is to establish a relation in which the community of
blood is the leading idea, and in which, where natural community of
blood does not exist, its place is supplied by something which by a
legal fiction is looked upon as its equivalent.

If, then, we do n