Demolished 1950. Excavation after demolition showed the nave and chancel of the medieval church had Roman walls for their foundations. <1> Question, now unanswerable, is how much Roman walling was incorporated into the Medieval church. The first church comprised a nave and chancel only, adapted from a standing Roman building, presumably sometime in the Anglo- Saxon period. The Roman building was not a church (why not?) but was part of a major Roman public building of as yet undetermined use. <2> Hulls recorded remains of Roman buidings in the adjacent Insulae 29 and 30 separated by a street. In both the earliest buildings were C1 of clay blocks which had burnt, presumably in the Boudiccan sack. Hull suggested these early buildings were public buildings. The last Roman buildings on the site were stone, C4 in date and also public it seems. There was little evidence for the intervening sequence; 18" of yellow clay loam suggested by Hull to be from demolition of timber framed buildings with wattle and daub panelling. Hull thought that the first church had been built on the Roman foundations rather than reusing standing walls. Finds included an uninscribed, illegible coin (one of Nero was reportedly found by one of the contractors workmen); a large bronze rivet; worked stone and marble; tesserae and a lump of tessellated pavement; wall plaster; pottery and amphora sherds. <3> No Medieval walls were found resting on Roman work. Instead, only the remains of Scott's (1875-6) church were found, suggesting that the Roman walls had been raised above ground level in the Medieval period and had not simply served as convenient foundations as Hull believed. <4>