What of London? The first reliable Anglo-Saxon Chronicle entry referring to the city is dated AD 604, when Augustine sent Mellitus to be bishop at a new Episcopal see in London, which appears to have come under the control of the East Saxons by this time. It has been argued on the strength of a lack of Early Saxon cemeteries that the civitates of the Catuvellauni and Trinovantes formed an enclave occupying Essex, Hertfordshire, Middlesex and London, and maintained their independence from the Saxons until at least 500.
There is archaeological evidence for 5th-century activity from several sites in and around London. The Billingsgate bath-house produced a stratified late 4th to 5th century sequence associated with its final occupation and later abandonment. The 370-400+ dated pottery assemblage from squatting in the derelict building is large but derived from relatively few part-complete vessels, including a large Alice Holt/Farnham greyware storage jar, and was accompanied by a scattered hoard of Theodosian nummi with issues dating up to 402-3. Above this occupation debris on the bath-house floor was a mass of tiles from the collapse of the roof of the building, containing a Saxon saucer brooch of a type dated to the 5th century (Gerrard 2011, 190).
Above the collapsed roofing was a series of black, pebbly riverine silts incorporating small quantities of highly abraded late Roman pottery as well as some large fresh fragments from a deep handmade dish in sandy ware fired black. I have identified the source of this as a small late Roman pottery industry in the valley of the River Cray near Orpington (Lyne 1994, 214).
The 3rd-century riverside wall of Londinium was repaired and modified at the south-east corner of the city (Parnell 1985). The construction of this new wall at the Tower of London can be dated to after 388-92 by a coin in a constructional context and may have taken place as late as the early 5th century. I saw some of the latest pottery from this site soon after the excavation and was struck by the abnormally large quantities of fresh fragments from horizontally rilled Overwey/Portchester D fabric (PORD) jars. The random percentages of such wares in the latest pottery assemblages from sites in south-east Britain suggests that they were made by itinerant potters at various places around the edge of the Weald from the 330s until well into the 5th century. Most of the occupation sites with high percentages of this kind of pottery are near the Lower Greensand and Gault clay outcrop around the edge of the Weald. Because of this, there is very little variation in fabric except at a postulated production area in East Sussex where a finer version has been found alongside the classic fabric in late Roman pottery assemblages from Bishopstone and elsewhere. Bishopstone has also produced a so far unique jug in the finer version of the fabric (Green 1977, fig. 75, 77), as well as variants of the standard jar form but without rilling (Green 1977, fig. 75, 93).
Elsewhere in London, outside the walled city, there is more evidence for 5th-century occupation. Excavation on the north side of the church at St Martin-in-the-Fields, adjacent to Trafalgar Square, revealed part of a Middle Saxon cemetery relating to the contemporary emporium of Lundenwic, occupying the present Strand and Covent Garden areas and founded during the 7th century. The cemetery also yielded an inhumation in a stone sarcophagus radiocarbon-dated to c. 390-430, with a late Roman tile-kiln nearby (Swain 2007). Church dedications to St Martin are usually very early and Bede mentions another church with the same dedication founded outside the east wall of Canterbury during the Roman occupation and prayed in by Ethelbert's Christian Queen Bertha in AD 597.