¹ Stein, Geschichte des spätrömischen Reiches, 1928, vol. 1 p.3.Google Scholar

² Seeck, Geschichte des Untergangs der antiken Welt, book 1, ch. 5. The evidence of a decline in population there quoted is voluminous and impressive, but one must beware of generalizing from instances of alleged depopulation which, if typical, would forbid us to think that there can have been any population whatever left in the Empire after a few generations.Google Scholar Rostovtseff, Social and Economic History of the Roman Empire, 1926, p.328, has argued that the decline in question was probably confined to Greece and Italy, and adds that many cities in Africa and elsewhere increased steadily in size up to the fourth century at least. He does not apply this dictum to Britain, nor, I think, would it be true if so applied.Google Scholar

³ I use the figures given in Carr-Saunders, Population, 1925, p.7. and take the opportunity of expressing my debt to that admirable book, and to the same author’s larger work on The Population Problem, Oxford, 1922.Google Scholar

⁴ Population, pp. 10–11.

⁵ Agricola, 7, 4.Google Scholar

⁶ There seems to be a certain difference of opinion among geographers and botanists as to how far these chalk and oolite uplands were originally free of timber. Thus, in the recent regional study of Great Britain edited by Dr Ogilvie, Cambridge, 1928, Professor Tansley says that ‘with pasturing excluded chalk grassland would be occupied by scrub and beech forest’ (p. 25), and Professor Rishbeth (p. 75) takes the same view but quite fails to show why, if that is so, these lands were so emphatically chosen as a residence by primitive agricultural man. On the other hand, in the same volume (p. 137) Mr Beckit says ‘it is virtually certain that the heavy undrained clays of the lowlands were covered primevally with forest or swamp, while the uplands, the home of the earliest inhabitants, were relatively clear’. For my purpose it does not much matter which of these two views is correct, so long as it is admitted that the uplands in question (a) were less intractable to primitive man than the lowlands because anyhow they did not need draining, and (b) had already been largely cleared, if they needed clearing, before the Roman period began.

⁷ Air Survey and Archaeology, 1928, second edition.

⁸ cf. Leeds, Archaeology of the Anglo-Saxon Settlements , map on p.19.Google Scholar

⁹ The following observations are not applied to Britain as a whole, but to the upper Thames basin and the region adjacent to it on the north and north-east. But they hold good of most of the more densely-inhabited Romano-British districts. ‘On the clays, i.e. in the lowland, deep and particularly fertile soils occur, but are usually heavy to work and hard to drain; the lighter soils of the uplands, usually sandy or calcareous, are lighter, drier, and easier to work, if less productive and shallower. It was on these soils that the earliest efforts towards tillage seem to have been made’. (Beckit, in Great Britain, cit., p. 137).

¹⁰ Romanization of Roman Britain, ed. 4, 1923, pp. 14–15.Google Scholar

¹¹ Tac. Agricola, xxi.

¹² Roman London (Royal Commission on Historical Monuments), 1928, pp. 35, 74.Google Scholar

^12a J.R.S. XIV, 226–8; XV, 228–9; XVI, 224–5; XVII, 197–199.

¹³ Loc. cit.:… laudando promptos, castigando segnis; ita honoris aemulatio pro necessitate erat.

¹⁴ These were the civitates to which in 410 Honorius sent word that they must provide for their own defence, i.e. take over the work of raising and employing military forces. In 429 St. Germanus found them doing their best (which was not very good) to carry out these orders, and lent them a hand (Bede, I, xvii–xx).

¹⁵ Marcellinus, Ammianus 27 8.Google Scholar I may add that I have called attention to the crucial importance of this invasion before (ANTIQUITY, I, 117–19), but only en passant in a review of Dill’s Roman Society in Gaul in the Merovingian Age; so that there is perhaps no harm in going into it here again and a little more fully.

¹⁶ With regard to coin-finds in villas, the reader can consult the materials collected in the Victoria County Histories, those of Hampshire and Somerset being especially full and illuminating. It may be added that the recent attack on the credibility of numismatic evidence in cases like this, based on a theory of so-called ‘coin-drift’ by MrFoord, Edward (The Last Age of Roman Britain, p. 27),Google Scholar has been completely answered more than once : the latest and most conclusive answer being that of MrSalisbury, F.S., Antiquaries Journal, 7, 268–81.Google Scholar

¹⁷ St. Patrick’s Confessions show that. In spite of Irish raids and kidnappings, the life of the country-side goes steadily on.

¹⁸ Our authorities are explicit on this point. The Romano-Britons were, obviously, adopting on their own account the policy of bringing in friendly barbarians which had long been practised in other parts of the Empire, and is amply explained by the density and distribution of the population. I would add that the destructive wars of aggression waged by the Saxons against the Britons, as recorded by Gildas, belong to a much later period. The conflicts belonging to the period of settlement (the middle of the fifth century), so far as we can recover their history, seem to have been mere cases of local friction between the old inhabitants and the new settlers.

¹⁹ In Gaul, fifth-century writers tell us that the Romano-Gaulish peasants were going over to the ‘barbarians’ of their own accord, because they found they were better off than under the thumb of the Roman tax-collector and large landowner (Salvian, book v). Of course conditions were different in Britain, but not so different as to make a spontaneous movement of this kind unthinkable.