Published online by Cambridge University Press: 02 January 2015
It is with great diffidence that I venture to make a few criticisms upon certain aspects of Mr R. G. Collingwood's most suggestive paper in the September number of ANTIQUITY on 'Town and Country in Roman Britain'. He is an expert, I am an amateur; he has probably forgotten more about Roman Britain than I ever knew; but the points that I desire to discuss are matters of general inference rather than of expert knowledge.
1 A little more information about the basis of the computation would be useful. The quotient of ‘under 9 to the square mile’ is apparently based on the area of England and Wales (58,340 square miles) including the portions of Northumberland and Durham lying beyond the Roman Wall. It is compared with an estimate of 26 to the square mile in 1066. No information is given as to how the figure was obtained, but presumably it is based upon the Domesday Survey of 20 years later. Wales was left out of Domesday for obvious reasons, and we have no means of obtaining even an estimate of the population of Wales in the eleventh century.
2 Oman, England before the Norman Conquest, p.220.Google Scholar
3 Quoted by Haverfield, Roman Occupation, p. 220.Google Scholar
4 Hominum est infinita multitudo creberrimaque aedificia fere Gallicis consimilia, pecorum magnus numerus. B.G. II, 12. The translation is by Dr T. Rice Holmes. Caesar, of course, is speaking only of the portion of southeast Britain that came under his direct observation.
5 It is interesting to remember that a corn-symbol appears on the reverse of the coins of Cunobelin.
6 Mr Collingwood now apparently accepts Hengist and Horsa as historical personages; a few years ago he thought that their names ‘put a strain on our credulity’. Roman Britain, p. 100.Google Scholar
7 ‘Britain : Rome : England’. Edinburgh Review, October 1925.Google Scholar
8 As a general statement I believe this to be correct, but there may quite well be local exceptions. Fleure and Whitehouse have made us familiar with the ‘valley-ward movement’ of upland peoples and this very slow movement may have been accelerated in places where the soils are propitious. Sinodun Hill on the Berkshire bank of the Thames opposite Dorchester appears to be a typical hill-top camp. But it stands on a valley-ward extension of the chalk, a sort of prehistoric Windsor Castle.
9 The clearing of the country was far less thorough in the later stages of the conquest than in the earlier. Freeman pointed out, on the evidence available in his day, that the character of the conquest of Somerset and Dorset was different to that of Kent and Sussex. Wiltshire and Hampshire provide evidence of the survival of some Celtic population. The survey of the English Place–name Society is throwing welcome further light on the problem.