Hunting the origins of Roman London

It has never been entirely clear why the Romans founded London exactly where they did in c AD50-55. There seems to have been no pre-existing British settlement at London, and several sites along the Thames might have been equally suitable for a new Roman town.

The debate, which has simmered gently for decades, came to the boil over the past year in the pages of the London Archaeologist. Three writers examined the written accounts of the Claudian invasion of AD43, the topography of London and the archaeological evidence (especially Roman roads), and produced three conflicting theories of why and where the town began.

In January last year, Bill Sole, an amateur archaeologist, made the radical suggestion that the Romans were drawn to the site by an early Roman settlement in Mayfair, west of the City and north-west of Westminster, founded some years before c AD50-55 (the date when archaeological evidence begins to appear in the City and Southwark).

A riposte came in October from Nicholas Fuentes, a post-graduate student, agreeing that `proto-London' existed but claiming its location was in Southwark.

Both were challenged in January this year by David Bird, County Archaeologist for Surrey, who saw no reason to dispute the evidence that there was no Roman settlement in the area before c AD50-55. Others, writing to the magazine, commented that Sole's idea had been anticipated by Francis Celona in the 1950s and Graham Dawson in the 1970s.

The main positive evidence comes from six roads: Wading Street east to Canterbury; Stane Street south-west to Chichester; the route west to Silchester; Wading Street north-west to St Albans; Ermine Street north to Godmanchester; and the route east to Colchester.

Stane Street and Ermine Street are aligned on London Bridge and the City, and are therefore probably no earlier than them. But the two Wading Streets can be projected to meet at Westminster, suggesting to Sole and Fuentes a primary river crossing there.

Sole noted that this route crossed the east-west route near Marble Arch, which he therefore saw as the hub of the primary road system, with invasion routes fanning out west, north-west and east. Such routes, he argued, would have radiated from a military base, which would have become an administrative and urban centre as the invasion passed on.

Fuentes, seeing a main invasion base north of the natural barrier of the Thames as military folly, preferred to locate it to the south, in north Lambeth or Southwark. A military camp there would then have developed into a proto-town.

Against them, Bird argued that no evidence exists for a Roman crossing at Westminster, which would in any case have been as difficult to engineer as one at London Bridge. Evidence from Verulamium, he added, suggested that Watling Street north of the Thames may have been laid as late as AD55. He saw the primary route north as being up the Lea Valley to the east of London.

Much has been made of the historian Dio's account, in which the Britons, in retreat following the Roman invasion of AD43, crossed the Thames `where it empties into the Ocean and at flood-tide forms a lake'.

This point is thought to lie between Westminster and London Bridge, where the low-lying river-banks are prone to flooding. It has been assumed to be the location of the first Roman crossing, although some have suggested a location as far upstream as Staines or downstream as Greenwich.

Bird pointed out that, in military tactics, a good crossing for retreating Britons would have been a bad one for advancing Romans, and that there was in any case no need to postulate an early crossing in the London area at all.

To support his theory, however, Sole appealed to the apparent existence of an early (pre-1700) straight stretch of road at what is now Park Street, east of Park Lane, and cited topographical evidence - humps in Culross St and Blackburne's Mews - for the outline of a proposed Roman fort in this area. He dismissed Fuentes's claim that an 18th century map showed no evidence of an early feature on the line of Park St.

What can we make of this debate? The difficulties of using archaeological evidence to pin-point events only ten years apart are well known, not to mention the problems of relating written evidence to topography. The lack of excavated evidence makes Bird's view almost convincing, but leaves a nagging doubt about self-fulfilling prophecies. The debate about the origins of Roman London is, I suspect, still far from over.

Clive Orton is Editor of London Archaeologist