although they may well have been concentrated in the more attractive portions of these natural clearings, while woodland, valley swamp, and bleaker hills remained comparatively empty, there is also ample evidence of them in the valleys, on the gravel banks of streams and rivers. For example, a hamlet examined, though not fully excavated, by Professor Haverfield showed similar, but ruder, features. It lay close to the Thames, at Northfield Farm between Abingdon and the Oxfordshire Dorchester, and consisted of largish circular and rectangular inclosures among which lay fragments of wattle and daub walling, rudely coloured wall-plaster, and roof-tiles and slates. Indications of similar sites have been noticed at nine or ten places in the neighbourhood, and it is probable that a row of hamlets occupied by peasants filled this part of the Thames valley. Similar hamlets abound along the river banks of Huntingdonshire, Northamptonshire, and Bedfordshire. Here we find hamlets planted in no obvious relation to any known country houses, and we cannot determine the position in which their inhabitants stood to the wealthy owners of large country seats. If the ‘villa’ system obtained in what is now Dorset or Oxfordshire, we might place the coloni in such hamlets.
   The local government of the province was left, according to Roman custom, to local authorities. We can distinguish three units of administration. The five municipalities mentioned above each possessed and ruled a territory of its own, which may have been as large as an average English county. The Imperial Domains also formed independent areas under Imperial officials. Their extent in Britain is uncertain. Perhaps they were smaller here than in many other provinces, but the mines were normally Imperial property, and some slight traces occur of other Imperial estates. The rest of the land— presumably the larger part of it—was left to tribal or cantonal authorities. These authorities represented the native chiefs and nobles of pre-Roman days. But they bore sway under Roman forms and titles; they were often styled duoviri, like true municipal magistrates, and their local council was called ordo, like the municipal senate. We may suppose that this council and the magistrates ruled both the cantonal area and its chief town—that, for instance, the ordo of the civitas Atrebatum, the canton of the Atrebates, administered both the cantonal area and the chief town, Calleva Atrebatum. If this assumption is correct, the local government of Britain resembled that of northern Gaul. But the cantons were smaller and less important in Britain than in Gaul, and have left fewer clear traces of their existence.
   One feature, not a prominent one, remains to be noticed—trade and industry. Here the first place is due to the agrarian industry of the landed estates which yielded wheat and wool in sufficient quantities to be exported to Gaul and even farther. This industry must have provided their occupations for the larger part of the population and their incomes for the landowners. Mining was also pursued actively in some districts during at least the first two centuries of the Empire. Lead was sought in Somerset, in Shropshire, in Flintshire, in Derbyshire, and iron in the Sussex Weald and the Gloucester-shire Forest of Dean. But the gold mentioned by Tacitus proved very scanty and employed only a few miners in the Welsh hills, while the far-famed Cornish tin seems, according to present evidence, to have been worked comparatively little and late. The chief commercial town was, from the