Nothing was known of a Roman bridge until last century. Then when the old stone bridge was destroyed evidence was found which convinced observers of the time that a Roman bridge had preceded it on the same line. Recently some writers, while accepting the Roman bridge as proved, have preferred to put it back to Stow’s line. Haverfield says: “No traces of a Roman bridge have yet been found (_Archæologia_, lx.): the oldest mediæval bridge (eleventh century) is said by Stow to have been near Botolph’s wharf (see plan).” This plan shows the bridge “temp. William the Conqueror” far to the east of Fish Street Hill (see also _V.C.H._). Exactly what Haverfield meant by saying that no traces of the bridge had been found is hard to say; it seems to have been as loose a statement as the one which seems to imply that the earliest mediæval bridge was of the eleventh century. Roach Smith, a cautious observer, was entirely convinced by the evidence that the mediæval bridge followed the course of the Roman bridge. “Throughout the line of the old bridge many thousands of Roman coins, with abundance of Roman pottery, were discovered, and beneath some of the central piles brass medallions of Aurelius, Faustina and Commodus. The enormous quantity of Roman coins may be accounted for by the practice of the Romans ... they may have been deposited upon the building or repairs of the bridge, as well as upon the accession of a new emperor.... The beautiful works of art which were discovered alongside the foundations, the colossal bronze head of Hadrian, the bronze images of Apollo, Mercury, Atys ... and other relics were possibly thrown into the river by early Christians” (_Archæol. Jour._, vol. i.). This seems substantial evidence. The charter cited by Stow only speaks of a wharf as being at the head of London Bridge; it does not tell us that the bridge ran into the middle of the wharf. The Roman bridge was linked up with an approach from the south over a raised causeway; the bridge-ends would have required much consolidation, and the foundations in the great tidal river must have been extremely difficult to construct. We should need very clear demonstration before we could believe that the early Saxons did more than patch up the work of skilled Roman engineers. Altering of the bridge to the Gracechurch Street line on the City side in 1176 would have meant replanning on a big scale. The ancient line of approach on the south side is guaranteed by the area of Roman finds (see _V.C.H._ plan). Gracechurch Street is known to have existed before the Conquest, and the positions of the ancient churches of St. Magnus’s and St. Olaf’s at each end of the bridge are significant: the bridge, I believe, was in the parishes of these two churches.