Of the British Isles at the time of the Angle invasion we have effected a sketch, rather than a picture; a sketch indistinct in outline, and with several of its details almost invisible. Nevertheless, it is a sketch in which some of the points are pretty clear. Germans of one or more varieties, Kelts either Gaelic or British, Picts who may be anything, Romans and Roman Legionaries are the chief elements. These we have had to distribute in Time and Space as we best could. We have also had, as we best could, to investigate their relations to each other. Let us look back upon what has been attempted in this respect. And first in respect to our _data_. The statements of the early authors, and the value which is due to them, have formed the subject of a separate chapter;[23] and it is hoped, that, without any undue disparagement, they have been shewn to be valid only when they are opposed to a very small amount of either conflicting facts or _a priori_ improbabilities. I also lay but little stress upon them when they assert a negative, and equally little when their apparent testimony may be reduced to an inference. Their absolute testimony, however, must be taken as we find it. Partly for the sake of recapitulation, and partly with the view to give a further investigation to certain questions which could not well be considered until certain preliminary facts had been laid before the reader, the more important inferences are put in form of the following propositions, to some of which a commentary is attached. I. _The British Isles were peopled from the Keltic portion of the continent originally and exclusively._ This implies an objection to the doctrine of any _pre_-Keltic population, and to the inferences deduced from certain real or supposed peculiarities in the shape of the skulls from the tumuli of the Stone period. (_See_ pp. 26-27.) II. _The Gaels cannot be derived from the Britons, nor the Britons from the Gaels; on the contrary, each branch must have been developed from some common stock._ This rests upon the differences between the British and Gaelic languages. (_See_ Chapter V.) III. _Of this common stock the British branch, at least, must have been developed on the continent._ (_See_ Chapter VI.) This, of course, assumes that the Galli of Gaul were not derived from Britain; a view which has never been adopted, and which probably has so little to recommend it as to make its investigation superfluous. The British language of Britain and the Gaelic of Gaul would not have been so much alike as they were had they developed themselves separately, each after their own fashion. This last proposition depends, however, to a great extent, upon the following, viz., that-- IV. _The similarity between the ancient language of Gaul and the ancient language of Britain is measured by that between the present Welsh and the Armorican of Brittany._