inute ethnology; especially when the area
under notice is an island. The _date_ of its occupancy, although
impossible as an absolute epoch, can still be brought within certain
limits. Whether, however, such limits would not be too wide for any one
but a geologist, is another question.

Now, if I have succeeded in shewing that criticism and disquisition must
necessarily form a large part of such an ethnology as the one before us,
I have given a reason for what may, perhaps, seem an apparent
irregularity in the arrangement of the different parts of the subject.
With the civil historian, the earliest events come first; for, in
following causes to their consequences, he begins with the oldest. The
ethnologist, on the other hand, whenever--as is rarely the case--he can
lay before the reader the whole process and all the steps of his
investigations, reverses this method, and begins with the times in which
he lives; so that by a long series of inferences from effect to cause,
he concludes--so to say--at the beginning; inasmuch as it is his special
business to argue backwards or upwards. Yet the facts of the present
volume will follow neither of these arrangements exactly; though, of
course, the order of them will be, in the main, chronological. They will
be taken, in many cases, as they are wanted for the purposes of the
argument; so that if a fact of the tenth century be necessary for the
full understanding of one of the fifth, it will be taken out of its due
order. Occasional transpositions of this kind are to be found in all
works wherein the investigation of doubtful points preponderates over
the illustration of admitted facts, or in all works where discussion
outweighs exposition.

The period when the British Isles were occupied by Kelts only (or, at
least, supposed to have been so) will form the subject of the earlier
chapters. The facts will, of course, be given as I have been able to
find them; but it may be not unnecessary to state beforehand the nature
of the principal questions upon which they will bear.

The date of the first occupancy of the British Isles by man is one of
them. It can (as already stated) only be brought within certain
wide--very wide--limits; and that hypothetically, or subject to the
accuracy of several preliminary facts.

The division of mankind to which the earliest occupants belonged is the
next; and it is closely connected with the first. If the Kelts were the
earliest occupants of Britai