Page:VCH Northamptonshire 1.djvu/217

 ROMANO-BRITISH NORTHAMPTONSHIRE places standing ten or eleven feet high. The individual remains — houses of various types, pottery works, numerous portable objects — are abun- dant and notable. There is perhaps no other site save Verulam through- out the non-military parts of Roman Britain which would better repay extensive, systematic and scientific excavation. It is greatly to be hoped that those interested in local archaeology will see that such an excavation is soon made. As a whole the site may be described roughly as extending four miles from east to west and two from north to south (fig. 4). The wind- ing Nene bisects it, the Roman high road from the south to Ancaster and Lincoln cuts straight across it in its unswerving north-westerly course and, as it crosses the river, throws out a branch which runs due north into Lincolnshire. Along these roads and east of them lie those portions of the site where the houses are thickest and the semblance of a town most definite. The rest of the site, particularly its western part towards Sibson, Wansford and Sutton, was less densely occupied. We may conveniently divide our closer description into three parts : the ' town ' south of the Nene, that north of it near Castor, and the re- mainder which we may provisionally style suburbs. On the south side of the Nene, between the river and the modern highroad, lie some fields known as 'the Castles,' where the visitor can yet discern the outline of a once fortified enclosure and the traces of Roman occupation, brick and tile and potsherds, lying on the surface. In shape the enclosure is an irregular hexagon, more oval than round ; in area it measures 45—50 acres ; the Roman highroad traverses it from end to end. It has nearer been explored. Stukeley tells us that the foundations of the south gate, constructed in hewn stone, were discovered by drainers in 1739 ; he adds, apparently on similar evidence, that the place was girt with a stone wall and a ditch 50 feet wide, but his observations were hardly minute enough to do more than confirm the fact, still obvious enough, that the enclosure had some sort of rampart and ditch around it. In the interior were dwellings. Artis marks the sites of twenty-two scattered up and down the area, and indicates in one of his plates that there were hypocausts and rooms whose walls were lined with thin slabs of local marble from Alwalton — sufficient signs of domestic comfortable life.' Close outside the north end of the enclosure and the north gate- way were other buildings — houses and potters' kilns, and three note- worthy objects have also been found here. An inscribed fragment, MARTO, was discovered, as Artis tells us, in ' removing a part of the old wall on the north side of the fortified ground.' Unfortunately he does not add whether this old wall was the Roman rampart or some other structure.* A rude bas relief of Hercules, also recorded by Artis, was TroUope erroneously transfers the south gateway to Castor. Alwalton marble resembles Purbcck marble, but is more shelly in texture, lighter in colour and perhaps more durable ; it was used in the Middle Ages, e.g. for Peterborough Cathedral. 169
 * Stukeley's Letters (Surtees Society's publications), iii. 60 ; Artis, pi. xxiii, xrvi. (i) and (2).
 * Artis, pi. XV. (1) ; Corf us Inscriftionum Latitiarum, vii. 79 ; Efhemer'is epigr. vii. 841.