The Roman riverside wall and monumental arch in London

Charles Hill, Martin Millett and Thomas Blagg   Download (54 MB)


During the summer of 1974 a rare opportunity arose to excavate a section across Upper Thames Street, a major and ancient thoroughfare, at a point a little to the east of the Mermaid Theatre and close to the south-west corner of the intra-mural City. The investigation showed that the street had been in continuous existence, albeit much encroached upon in the late medieval period, from the 12/13th centuries. It also showed that its earliest deposition overlay a collapsed portion of a Roman wall. Lying on the presumed shore-line of the early City, this masonry was seen as a remnant of a riverside wall which had long been conjectured as a logical extension of the more enduring landwall but which had never been verified archaeologically. On the other hand, documentary evidence was rather more positive, and the fact that below the wall were found dumps of 6th to 8th-century date seemed in particular to support William FitzStephen’s statement of the 1170s that such a wall had existed but had succumbed at some earlier period to river erosion.

Accordingly, close observation was maintained in 1975-6 on redevelopment activity over a much more extensive area to the east and south of the original controlled excavation. At intervals considerably more evidence of the wall came to light, in varying states of decay and exhibiting differing methods of construction, over a total length of some 115m. To the west, where the sub-soil was firm, less elaborate foundations were required but on the less stable ground to the east a 40m stretch of wall was constructed upon a chalk raft supported by tightly packed rows of timber piles. Carbon 14 and dendrochronolgoical dating of these piles indicated that the wall was built in the 4th century, probably after AD 330. Certain of its construction methods share similarities with the apparently late Roman eastern group of bastions added to the landwall, and a further, more recently discovered, portion of the riverside wall close to the inner curtain of the Tower of London at the opposite end of the City has been provisionally dated on numismatic evidence to the 390s.

One peculiar characteristic of the western section of the wall was the re-use of sculputured stone blocks which on close examination were found to possess a dramatic and independent interest of their own. Most of the 52 blocks recovered can be assigned to one or other of two major monuments, of which one was an arch, richly decorated with figures of classical gods, including Minerva and Hercules, in relief, and surmounted by a frieze of the busts of deities, possibly representing the days of the week. Rare in Britain - it is in fact the first time that a detailed reconstruction could be made of the original appearance of a Roman arch in this country - and unique to London, the arch was probably not erected earlier than the late 2nd or early 3rd centuries. The other monument was a screen of gods, 6.20m long, carved on both sides and at one end, which presumably formed part of some larger monument. The remaining blocks included a relief of four (at variance with the more usual three) Mother Goddessess probably of 3rd-century date; and two altars commemorating the restoration of temples, one to Isis and the other probably to Jupiter, one of which records the name of an hitherto unknown Roman governor of Britain whose term of office was most likely within the years AD 251-9.

During the post-Roman period the gradually rising level of the Thames caused the southern face of the wall on the eastern part of the site to erode to about half its original width, and by the mid 12th century a gravel foreshore had formed against the surviving masonry. But while there is now every reason to respect FizStephen’s claim that the destruction of the wall in general was due to erosion, it was also clear from the present site that this was not the only cause. One feature of the wall in the western part of the site was that it had collapsed northward or inland, a phenomenon which strongly suggests deliberate demolition. It is unlikely that this remote section of the City waterfront was developed much before the mid 12th century at the earliest, and the present conclusion is that this demolition took place in the late 12th century before the superimposition of the earliest Thames Street level and before the construction of the timber waterfront to the south in the first half of the 13th century.

[ Special Paper 3 (1980); abstract as published]