A HISTORY OF LONDON character very similar, being arched passages with flat bottoms, built of Roman tiles with red mortar and embedded in amass of rubble masonry. The foundation of the wall where these culverts occur is found to fall rapidly to a much lower level than that near Allhallows Church. The reason of this is that it crosses a stream which was of considerable dimensions when the wall was built, and in the fiUed-up bed of which flowed the smaller stream, known in the Middle Ages as the Walbrook. Although not recognized at the time of their discovery, it is now clear that these culverts were inserted under the wall to form a passage for the stream. The more easterly of them was described by Richard Kelsey, the City Surveyor, in 1837^ (Plan C, 29). The depth from the surface to the bottom of the culvert was 18 ft. 4 in., but the dimensions of the opening are not recorded. To the north were dis- covered one upright and two sloping iron bars, while the south entrance was found at a distance of 14 ft. from the wall, where it discharged into a ditch. West of this was the other culvert, which is described and figured by Roach Smith*- (Plan C, 30). It was found to begin at a point 20 ft. north of the wall, having an opening 3 ft. 6 in. high and 3 ft. 3 in. wide, and from there it ran in a southerly direction for 60 yds. The crown of the arch was 19 ft. below the surface, so that its base would be at a depth of 22 ft. 6 in. (Fig. 24). During operations for laying telephone mains in 1905 an excavation was undertaken by the Society of Antiquaries in order to obtain information as to how the wall had been constructed across the stream.*^ The nearest point available was just opposite Throgmorton Avenue, under the pavement on the north side of London Wall (street) (Plan C,3l). A shaft was carried down the outer face of the wall, the latter being met with about 5 ft. below the surface, to which level it had been broken down for the structure that stood on it until the rebuilding and widening of the street. The work was of the usual Roman character, with the face exceptionally well preserved, showing two bonding courses, each of three tiles, while the chamfered red sandstone plinth rested at a depth of i 3 ft. 3 in. This depth is about the same as that at Allhallows Churchyard, and as the present level from the church to this point falls between 2 and 3 ft. it follows that the plinth was laid to that extent out of the horizontal in a distance of about 3';oft. Below the plinth came a solid mass 5 ft. 8 in. deep, of irregularly shaped ragstones projecting 2 ft. in front of the face of the wall, and resting on the usual flint and clay puddling (Fig. 20). The total depth to the bottom of the stream at this point was thus 19 ft., while that of the centre of the stream, as represented by the lower culvert, is 22 ft. 6 in. It would seem then that the shaft was sunk about as far to the west of the centre as the higher culvert was to the east of it, their depth below the surface being about the same. The sub- structure of the wall may be presumed to have been carried across the hollow bed of the stream, the culverts being formed in it to carry the water, much as in the case of the smaller stream noticed near Allhallows Church. It is probable from the width and position of the stream deposit that more than two culverts were employed and others may yet be brought to light by future operations. From this point the wall passes from under the pavement until it reaches Moorgate Street, where it is nearly in the centre of the roadway. Part of
- ' j4rci. Ix, 237 ; Cat. Antiq. Roy. Exch. xxxi. " jirch. xxix, 152. " Ibid. Ix, 169.
60
