Skip to Content.

Kent Archaeological Review extract
 

Roman Dover -- Hide and Seek.
by Sir Mortimer Wheeler.

IN ROMAN times, a thousand years before Dover's majestic Norman castle first frowned above its cliff upon the narrows opposite Calais, a fortress already stood in the neighbourhood known by some such name as Dubrae.

Where actually this fortress lay was unknown until the summer of 1970, but over 40 years ago I collected, with local aid, the find-spots of casually discovered Roman fragments of one kind or another and, placing them upon the landscape, made a shameless guess at the answer. In fact I committed myself in publication to a plan showing precisely where the fortress should have been, at the same time admitting that the visitor would there and then find no vestige of Roman fortification.

But this hazardous conjecture was conditioned also by other factors than a scatter of buried masonry and pot-sherds. We know that Dubrae was primarily a fortified harbour, port-of-call for the Roman Channel fleet, which left inscribed tiles and recorded its status as a coastal station in late Roman times.

To anyone approaching from the Channel with an eye for harbourage, the only place where a naval fortress can effectively have been sited was at or near the mouth of the little River Dour, within a few hundred yards of the present harbour and sheltered by flanking hills whereon two Roman lighthouses (both still traceable and one indeed still towering boldly within the medieval castle) bracketed the scene as twin sea-marks.

To a tolerably experienced mind in 1929 all this amounted to something approaching certainty. More closely, the Roman fortress had lain in the vicinity of the Market Square and the former position of St Martins-le-Grand.

Consistently with these general considerations, the late Sir Alfred Clapham adduced a possible pointer from the Saxon period. There is a documentary assertion that King Eadbald of Kent founded a monastery before 640 'in the castle of Dover.' There is no evidence whatever for a castle on the Eastern Heights before the Norman period.

On the other hand, there was an early Saxon foundation on the site of the former St Martins-le-Grand within the conjectured and now proved area of the Roman fortress. To the surviving remains of this fortress in the 7th century AD the term 'castle' was no doubt still perfectly applicable.

To me, I will confess, these accummulative scraps of evidence were, and remained, reasonably satisfying. But in the intervening years they had a bad Press, which seemed to be supported by unrevealing excavations (which I did not see) carried out after World War II.

The situation at the beginning of 1970 was summed up in a learned journal: 'The thesis (of 1929) has now been dismissed. We now know that a fort in the suggested position was geographically impossible.' The proposed fort was a 'will-o'-the-wisp.' That might have been the end of the story; but not quite. Within a few weeks of the publication of the sentence of dismissal, in June 1970, Brian Philp and his colleagues began their skilful and productive rescue-excavation on the line of a new road near the Market Square.

Within a few days they had lighted upon a thick fort-wall within seven feet, I am told, of the hypothetical line of 1929, and now have recovered most of the outline of a third- or fourth-century fortress. Beneath that lay a second-century fort with many of its buildings intact on plan. A few yards outside the defences has emerged a residential or bathing establishment with wall-paintings, of unrivalled extent in Britain.

Brian Philp has led the way in this adventure of discovery and has added new chapters to the history of this traditional gateway of Britain. But even his leadership would have been short of effective without official and less official cooperation. Above all, his success underlines the potential wealth of historical and cultural treasures which is daily destroyed or threatened in our crowded country.

 
Close Window
Accessed this page via search engine or bookmark? Full K A R Index here.