or 80 till A.D. 110 or 120; and the fact that the masonry there, so far as I could see, showed practically no sign of repairs or reconstructions, points also to a comparatively brief occupation. And indeed we may well believe that by A.D. 110 the hills of South Wales were quiet enough to allow of reductions of garrison. The conquest of the district, according to our ancient historians, began about A.D. 50, but was actually effected between about A.D. 75 and A.D. 80 forty years later the fort at Gellygaer may have become superfluous. Excavation alone can show whether that was also the case at the Gaer, and, if so, whether the buildings were subsequently squatted in by others than military inhabitants. It is, however, likely enough that some of the outlying little forts were held long after the first period of conquest and pacification. It was found possible, in the second and third centuries, to detach vexillations" of the Second Legion to the Roman wall for temporary purposes, and this suggests that South Wales had then become comparatively peaceful. But, even so, a fort like that near Brecon may still have been kept up. How long it lasted is, however, outside our knowledge. The roads and forts of the south coast, from Cardiff to Carmarthen, seem to have been, at least partially, restored by Constantius Chlorus or Constantine, early in the fourth century; but it is hard to say exactly what this restoration was, and it is as yet impossible to say how far inland it extended. When local research and excavation have gone further forward we shall be able to write more fully, not only the history of this single fort, but of the system of forts and roads to which it belonged. It may still remain a nameless fort, a blockhouse X. But it and its kindred forts will illustrate the methods of an imperial people faced by difficult hills and stubborn men.

[The inscriptions found at the Gaer cannot be dated. Some tiles of the Second Legion may belong to the foundation of the fort (compare Tacitus, Annals, xii, 38. 3). I am told that coins have been found at the Gaer, but 1 cannot learn their dates.]

THE EARLY SETTLERS OF BRECON.

BY PROFESSOR E. ANWYL, M.A.

In spite of the striking modern developments of Anthropology, Archæology, Comparative Philology and Comparative Mythology, the reconstruction of the prehistoric past of Man must always be, at best, of a very tentative character, and especially when the evidence, as in the case of Breconshire, is far from abundant. What evidence there is appears to be more suggestive than conclusive, and the interpretation of it is by no means free from ambiguity. However, it is not impossible that, in course of time, further evidence may be found, especially if, at some future date, this and the neighbouring counties of England and Wales undergo a thorough Archæological and Anthropological Survey. We are fortunate in possessing for Herefordshire an excellent Archæological Survey, in the carrying out of which Mr. Haverfield has taken a prominent part. We have a most valuable Antiquarian Survey of East Gower by Colonel Morgan, and a survey of the archæological remains of Pembrokeshire, with maps indicating the position of ancient monuments, carried out by several learned members of this Association. It would be an excellent thing if a similar survey could be undertaken also for the county of Brecon.

As the evidence relating to the early settlers of Brecon is not abundant, and any clue that may suggest a possible solution of the problem is useful, attention will be called in this Paper to certain considerations derived from the river-names of the district, apparently the most ancient place-names that we have. It is generally admitted that river-names often survive great changes in the ethnology of any country, and Wales is probably no exception to the rule.

For the purpose of the present paper, it will be convenient to treat of the early settlers of Brecon in the order of the great stages of civilisation through which European man has passed: the stage of stone implements, the stage of bronze implements, and the stage characterised by the use of iron. In dealing with these phases of civilisation, it should never be forgotten that they must have largely overlapped; that, for example, the use of stone implements must have continued long after the introduction of bronze,1 and the use of bronze weapons after the introduction of iron. Moreover,

at any rate in the earlier periods of these stages, some parts of a country or district would naturally be in possession of the higher phase of civilisation, while others would still be in the lower: The distribution of early civilisation, like that of more modern times, was very largely determined, not by conquest and colonisation only, but by economic considerations of barter and exchange, and by the direction of the ancient traderoutes along which goods passed by a system of intertribal barter. Hence, a district which was favourably situated from this point of view, could steal a rapid march in civilisation upon another where the conditions were less favourable. Much of the best archæological work to-day-as, for instance, that of Mr. A. J. Evansconsists in a thorough and painstaking investigation into the ancient trade-routes of the world.

Of Paleolithic Man in Breconshire, so far as the writer is aware, there are no traces on record. It would, however, be obviously rash to infer that, even if no Palæolithic flint implements have been found in Breconshire, the men of that period in their hunting expeditions never set foot in the county. Roughlyhewn flint implements, the most common remains of Palæolithic Man, are naturally most abundant in

1 At Clegyr Foia, the Rev. S. Baring-Gould has found indications of the use of stone arrows, even in the "Iron Age."

2 Mr. J. Romilly Allen has called my attention to bronze objects ornamented in imitation of patterns found on implements of iron. 6TH SER., VOL. III.

2

districts like the South of England, where flints abound. Where skulls belonging to this period are found, they are marked by an extreme dolichocephalism. As to the affinities of the Palæolithic men of Britain, several anthropologists have suggested that they were closely related to the Eskimo, and that, as the ice of the Glacial Period or periods melted, they followed the receding fringe of it to the North, in quest of the Arctic animals that accompanied it. If such was the case, could not others, to whom an Arctic climate was not a vital necessity, have remained in Britain, and thus established a link of connection between Palæolithic and Neolithic Man? The investigation of the Hoxne Palæolithic remains by Sir John Evans and others, seems to lead to the conclusion that they are Post-Glacial in character, and so far tends to support the theory of continuity.

The next great phase of civilisation is the "Neolithic," or that of the polished Stone Age. Between this and the former there must have been, in some parts of the world, a transition period, and this has been called by Mr. J. Allen Brown, the "Mesolithic," characterised by flints of a better form than those of the Palæolithic period. Much, however, remains to be done in tracing the continuity of the Stone Age, on the Continent as well as in Britain. The chief facts, as at present known regarding early man in Britain, afford prima facie evidence of a contrast in point of culture between the Paleolithic and the Neolithic periods, and this has not unnaturally been interpreted as indicating a difference of race.

The

Breconshire, so far as the writer can discover, has yielded no skull which can be assigned to the Neolithic Period, vast as that must have been in Britain. skulls of this epoch, mostly found in the Long Barrows, are, like those of Paleolithic Man, remarkably oval and dolichocephalic, the dolichocephalism, however, being more extreme in the case of the older type. Both 1 Journal of the Anthropological Institute for 1893, p. 92. 2 Keane, Ethnology, pp. 110, 111.

types have a lower average cephalic index than any men in modern Europe, except the Corsicans, and the stature of both types was below that of any variety now living in Britain.1 This resemblance of type between Palæolithic and Neolithic Man, in spite of the contrasts in culture, warns us not to assume too hastily a complete difference of race. The spread of culture, even in prehistoric times, was by no means necessarily coincident with the racial extension. It is a remarkable fact that the long-headed or dolichocephalic type of head is characteristic of Northern, Western, and Southern Europe, as well as North Africa; but with this important difference, that in the North it is combined with blonde characteristics, whereas in the other areas the complexion is, in varying degrees, brunette. The prevalent type of head found in Central Europe and its outlying districts, is, on the other hand, the brachycephalic or broad-headed, and this type has now spread into many parts of France, and even as far west as Brittany. There are sufficient indications that Neolithic Man of the polished Stone Age inhabited Breconshire, and he, too, doubtless conformed to the general dolichocephalic type of Britain. In British Neolithic graves this type of skull is generally combined with short stature. Within the large dolichocephalic area above mentioned, where a dark complexion prevails, modern research seems_to establish the existence of well-marked sub-groups. For instance, in the neighbourhood of Périgueux, in France, the ancient Cro-Magnon type of skull, with its marked dolichocephalism, but with an unusually broad face, survives conspicuously in the present population; and, as Ripley points out, this type was at one time much more widely distributed over Europe than it is now. Again, in the case of the Berbers of North Africa, the

1 Ripley, The Races of Europe, p. 306.

Ripley, The Races of Europe; Deniker, The Races of Man; Sergi, The Mediterranean Race; Keane, Ethnology, and Man, Past and Present, contain valuable discussions on these points.