A HISTORY OF NORTHAMPTONSHIRE the midland districts, the west and centre of Northamptonshire, should have possessed a less richly developed civilization than many other parts of the Roman province. 2. Towns of Romano-British Northamptonshire (a) castor The most interesting and the most important of these towns is in the east of the county, but it does not lie wholly within it. The remains discovered at Castor on the north bank of the Nene, four miles west from Peterborough, form only part of a larger whole which stretches out south of the river and belongs no less to Huntingdonshire than to Northamptonshire. Here as elsewhere the student of Roman Britain must ignore the territorial divisions of later England. For our present purpose it is a mere accident that the Nene at Castor separates two English shires. The remains on its north bank cannot be sundered from those on its south bank : the two together constitute one extensive straggling settlement. The Roman name of this settlement is generally and confidently asserted to be Durobrivae. That was the view of Camden, and it has been the dominant view, if not the universal view, ever since. Probably it is also the true view. But the arguments adducible in its favour are in reality very unsatisfactory and demand some examination. They rest on two pieces of evidence, (i.) one supplied by the Antonine Itinerary and (ii.) the other by an old English appellation of Castor. (i.) The Itinerary mentions a route from London by way of Col- chester and Lincoln to the north, and inserts as ' stations ' between Colchester and Lincoln the following : Villa Faustini, Icini (or Iciani), Camboritum, Durolipons, Durobrivae and Causennae.' The determina- tion of these places and of the route connecting them is a well known problem in Romano-British topography. We should expect the route to run north-west from Colchester and then skirt the Fens by way of Cambridge, Huntingdon and Peterborough. But no Roman road can be traced issuing from Colchester in the direction of Cambridge ; none of the Roman names are otherwise known to us, and the mileage of the Itinerary is irreconcilable with any reasonable identifications of them. If however we start in the north we can trace a road running south from Lincoln and passing Roman sites at Ancaster, Castor and Godman- chester near Huntingdon. Its further course is complicated and obscure ; but so far it may well represent the Itinerary route, and the Roman sites along it may be the Itinerary ' stations.' That is to say, Ancaster may be Causennae and Durobrivae may be Castor. Certainly this road 1 ///». Ant. 474, 475; Ravennas, 429, 12-7, probably names some of these stations, but with very distorted orthography. Some of the places (e.g. Villa Faustini and Icini) may belong to a branch route (see Victoria History ofNorfolk, i. 300). It used to be thought that the similarity of names fixed Camboritum at Cambridge and thus gave us a definite point on the route. Mr. Skeat has however shown that the names Cam and Cambridge are comparatively modern and for our purpose useless {Placnama 0/ Cambridgeshire, p. 30). 166
