Inventory of Roman London: The defences, structural remains

STRUCTURAL REMAINS.

The Town Wall.

N.B.—The positions of the various items, numbered consecutively in this section, are shown by the same numbers prefixed by the letter W on the large map A at the end of the volume. The structural remains now visible are scheduled on p. xii. All other remains are either buried or destroyed.

(1). Wardrobe Tower, Tower of London. A length of about 10½ ft. of wall (Fig. 11) remains standing to a height of 4¾ ft. at the back of the Wardrobe Tower. The thickness above the plinth is 6 ft. 11 in., and the external facing above the sandstone plinth consists of four courses of squared rag, three courses of brick carried through the wall and two more courses of rag. On the internal face of the wall an offset of three courses of brick corresponds in level to the plinth. The line of this fragment, which is still exposed, if produced southwards would strike the modern Lanthorn Tower which stands some feet to the N. of its predecessor. The known line of the wall to the N. of the Wardrobe Tower indicates that at this point there was a slight angle in the wall itself [Journ. Brit. Arch. Assoc., XXXVII, 280; XXXVIII, 130]. Excavations made in 1904 to trace the course of the wall to the S. of this point showed that all remains of the Roman structure had been removed [Arch., LX, 239].

(2). Bowyer Tower, Tower of London. Early in 1911 a small excavation in the floor of the Bowyer Tower revealed the inner face of the Roman wall. Only a small portion was uncovered, but included brick bonding-courses. It was carefully preserved and can be inspected by raising one of the stones of the modern floor [Arch., LXIII, 259].

(3). Trinity Place, S. A length of about 50 ft. of the town-wall is still standing here to a height of about 15 ft. The internal face is open to view and appears to be entirely of mediæval construction with the possible exception of the lowest visible courses of squared rag-stone which may be of Roman date. An engraving (Plate 22) in Roach Smith's Illustrations shows the Roman work still surviving on the external face to a considerable height. It consists above the plinth of four courses of squared rag-stone, a triple bonding-course of brick, six courses of rag, a double bonding-course, five courses of rag, a second double bonding-course and seven courses of rag; above this point the face has gone and the wall is perhaps mediæval. Between this point and Tower Hill a considerable stretch of the wall is incorporated in a warehouse and other buildings. The external face is visible in places and appears to be of mediæval date. Projecting on to Tower Hill is a narrow tenement which, no doubt, stands on the Roman wall.

(4). Trinity Place, N. A length of 73 ft. of the Roman wall (Plate 25), immediately adjoining (3) was destroyed for the construction of the Inner Circle Railway in 1882. Drawings and photographs of the wall from the Gardiner Collection are now in the possession of Dr. P. Norman. The drawing of the wall by H. Hodge shows the external face with four courses of squared rag above the plinth followed by three brick courses, six courses of rag, two of brick, four of rag and two of brick.

(5). Cooper's Row. A length of 110 ft. of the wall (Plate 23) was uncovered in 1864 on the rebuilding of Messrs. Barber and Co.'s warehouses. It was standing to a height of 35 ft. above the ancient ground-level; the upper part of the construction was of later date and included two round-headed embrasures, perhaps of the 12th century. This section of wall was retained in the new building and can still be inspected. The triple course of bricks on the internal face, corresponding to the external plinth, is 6 in. below the existing basement floor-level. A portion of the outside face of the wall is visible in the garden of No. 8 the Crescent, Minories [Arch., XL, 297; LXIII, 259].

(6). Southend (formerly Blackwall) Railway, S. In the course of demolitions for the Blackwall Railway in 1841, a portion of the Roman wall was uncovered 7½ ft. thick and standing to a height of 6 or 7 ft. It had a double course of bricks surmounted by five courses of squared stone, and a double bonding-course carried through the wall. The wood-cut shows also two courses of squared stone below the lower courses of brick and three above the higher [Knight, London, I, 163].

(7). Southend (formerly Blackwall) Railway, N. In 1881 a further stretch of the wall immediately adjoining (6) together with bastion No. 3, was destroyed in a widening of the railway on the N. side. The stretch was 40 ft. long and 8½ ft. thick above the plinth. The best record of it (Plate 24) is a drawing by H. Hodge (Guildhall Library Add. Prints, p. 98), which shows on the external face three courses of squared rag-stone below the plinth and four courses above followed by a triple bondingcourse with a set-back above the second brick; then six courses of squared rag-stone and a double bonding-course with a further set-back immediately above it; finally three courses of squared rag-stone. On the internal face the offsets are reproduced with the usual triple levelling-course opposite to the plinth. The top section of the wall is shown 7½ ft. thick, and the footings are 9 ft. in width.

Loftus Brock records the composition of the wall (Fig. 11) at this point somewhat differently. He states that the width at the base was 8½ ft., and his section shows no faced masonry below the level of the plinth. Above the latter the wall was 7½ ft. thick and stood to a maximum height of about 8 ft. His section, like that of H. Hodge, shows offsets on both faces of the wall [Journ. Brit. Arch. Assoc., XXXVI, 463; and section in XXXVIII, 132]. A. A. Langley who was in charge of the work confirms Hodge's section, particularly in regard to the exceptional offsets on the external face, and to the fact that faced masonry was carried below the plinth-level. He shows also that the foot of the faced walling was 18 ft. below the modern surface and at about this level there was a roughly rectangular drain of Roman brick carried through the wall [Antiq., III, 62].

(8). America Square. In 1908 the demolition of Nos. 15 and 16 America Square, a short distance N. of the piece last described, revealed a stretch of about 65 ft. of the wall (Fig. 19), and a further special excavation was made to determine its character. Above the plinth it stood to a height of 7 ft. and was 8¼ ft. thick. The base was 16 ft. below the street-level of America Square. The external face showed four courses of squared ragstone above the plinth, followed by a triple bondingcourse, six courses of squared rag-stone and a double bonding-course. The internal face showed the usual offsets [Arch., LXIII, 261].

(9). Roman Wall House. In 1905, on clearing the site of Nos. 18–20 Jewry Street and No. 1 Crutched Friars, a length of about 40 ft. of the inner face of the wall (Plate 26) was uncovered. Its maximum height was 8–9 ft., the base being 8½ ft. below the present ground-level. The usual triple levelling-course of brick was surmounted by four courses of squared rag-stone and a triple bondingcourse, six courses of squared rag-stone and a double bonding-course with the usual offsets. A large portion of this fragment is preserved in the basement of the modern building [Arch., LX, 191].

(10). The Cass School. About 1900, on rebuilding the Cass School at the corner of Jewry Street and George Street, the foundations of the wall were uncovered in the lower part of the site. The wall itself had been previously destroyed [Arch., LX, 193].

(11). Jewry Street, N. end. In 1861 a considerable stretch of the wall on the E. side of Jewry Street, immediately S. of Aldgate, was uncovered in the rebuilding of the premises of Messrs. Moses. It lay immediately beneath the frontage of Jewry Street, and according to Loftus Brock, the foundations rested on massive piles. A drawing of this stretch of wall is preserved at the Society of Antiquaries, Red Portfolio, London, I, but contains obvious inaccuracies [Journ. Brit. Arch. Assoc., 163; Gent. Mag., 1861, I, 646].

(12). Duke Street. In 1887 the widening of Duke Street on the N.E. side exposed a long stretch of the Roman wall. It lay partly beneath the foot-way of the old street and partly beneath the frontage of the demolished houses. Loftus Brock described the wall as being similar to the normal type of the structure as in Camomile Street, but no exact particulars are given [Journ. Brit. Arch. Assoc., XLIII, 203].

(13). Bevis Marks, E. end. In 1880 a stretch of about 70 ft. of the wall was exposed and removed in the rebuilding at the back of No. 31 Houndsditch. It formed the boundary at the back of the houses in Bevis Marks, and was standing to a height of 11¾ ft. Loftus Brock gives a section of this piece of the wall (Fig. 11) from which it appears that it was nearly 8 ft. thick and differed slightly from the normal section already described. On the outside face there were two courses of squared rag below the plinth and four courses above it followed by a triple bonding-course and one course of rag, above which the face had been destroyed. On the inside face, above the triple levelling-course of bricks were four courses of rag, the triple bonding-course without a set-back, five courses of rag and a second triple bonding-course [Journ. Brit. Arch. Assoc., XXXVII, 86; XXXVIII, 132–5, with section].

(14). Bevis Marks, E. of Goring Street. In 1923 a stretch of the wall about 120 ft. long and 8½ ft. thick immediately S.E. of Goring Street was uncovered and destroyed. The external face was of the normal type with sandstone plinth and triple bonding-course. The base of the plinth was 7½ ft. below the modern ground-level. The wall stood on 5 ft. of brick-earth lying above the gravel. [Journ. Rom. Studies, XII, 258 and P.N.].

(15). Bevis Marks, W. of Goring Street. A drawing, probably by H. Hodge, in the possession of Dr. P. Norman shows a plan and section (Plates 31, 32) of the wall at this point discovered together with the adjoining bastion in 1884. Though there is no title to the drawing the mention of Castle Street (now Goring Street) in the levels is sufficient to fix its position. The wall rested on a foundation of flint and puddled clay and consisted on the external face of two courses of rag-stone, the red sandstone plinth, four courses of squared rag; a triple bonding-course of tiles and five courses of squared rag. On the inside face the usual three courses of brick lined with the external plinth. The base of the plinth was about 7¼ ft. below the modern pavement-level and the wall was 8 ft. 10 in. thick [Antiq. Journ., VII, 518; Antiq., X, 134].

(16). Camomile Street, E. In 1905 a stretch of the wall was uncovered at the back of Nos. 58 and 60 Houndsditch and also adjoining the churchyard belonging to the parish of St. Martin Outwich. The bottom of the plinth lay at a depth of 8 ft. 4 in. below the street-level, and the total height of the fragment (which projected above ground) was 14½ ft. above the base of the plinth. The external face consisted of four courses of squared rag-stone above the plinth followed by a triple bonding-course of brick, and two more courses of squared rag-stone. Above this point only the core of the wall remained, but it included three double bonding-courses. This would appear to be the highest fragment of the wall so far recorded [Arch., LX, 187]. In 1926 a further portion of this same section of the wall was revealed and destroyed.

(17). Camomile Street, middle. In 1876 a stretch of the wall 70 ft. long (Fig. 25) was uncovered and is described by J. E. Price in connection with the bastion. It was 8 ft. thick and was destroyed above the plinth J. E. Price, On a Bastion of London Wall, 1880].

In Woodward's Letter to Wren mention is made of the destruction of part of the wall near Bishopsgate. It is described as standing 10 ft. high, with a thickness of 9 ft. and foundations 8 ft. below the surface. The wall was built of courses of stone with double courses of bricks at 2 ft. intervals.

(18). London Wall, E. of All Hallows Church. In 1905 a small portion of the wall was uncovered at a distance of 45 ft. E. of All Hallows Church. The plinth and two courses of squared rag-stone were exposed [Arch., LX, 211, with section].

(19). London Wall, All Hallows Church. In the same year, during the excavations of the bastion beneath the vestry, the lower part of the wall was uncovered. It showed the plinth with four courses of squared rag-stone and a triple bonding-course [Arch., LXIII, 271].

(20). London Wall, All Hallows Churchyard. The city-wall still forms the N. boundary of the churchyard and the external face was uncovered in 1905 (Fig. 13). The Roman work remained to a height of about 12 ft., i.e. to about the present ground-level. The facing consists, above the plinth of four courses of squared rag-stone, a triple bonding-course, five courses of rag, a second triple bonding-course, six courses of rag, a double bonding-course, and three more courses of rag. Below the plinth the foundation was pierced obliquely by a brick-lined culvert (Fig. 12), 15 in. by 9 in., set in red mortar, in a hollow depression at a depth of 2 ft. 4 in. below the plinth. The fall of the drain was from S. to N. "The soil of this depression had, in the lower part, the appearance of the filling of a stream, being light sandy silt and contained Roman pottery, oyster shells, a human femur and other animal-bones. . . . On digging further in the lower portion of the stream-deposit, which continued to a depth of 3½ ft. below the plinth, many pieces of tile were found and the remains of a human skeleton" [Arch., LX, 207].

(21). London Wall, W. corner of Blomfield Street. In 1837, "In building the new sewer, at a few feet eastward of Carpenters' Buildings, an ancient sewer of Roman workmanship was cut through. It was embedded in a mass of rubble masonry 12 ft. wide. At 14 ft. southward of London Wall, it terminated in a mouth cut to the slope of the ditch into which it had discharged itself. The bank of the ditch was still covered with large quantities of moss. On the northern side it had been converted into a place of sepulchre. The remains of two human skeletons, with a large dog's skull and part of the stem of a stag's horn, were found therein, together with some Roman pottery, a small silver coin of Antoninus, and a copper coin of Faustina, and other ancient money. One upright and two sloping stout iron bars at 12 ft. north of the new sewer, Moorgate Street to Old Broad Street, closed the mouth of this tomb, and were in the most perfect state of preservation, still retaining their grey colour. At 11½ ft. northward the crown had been broken in. A coarsely wrought base of a column was among the rubbish. The bottom is flat and paved with two layers of large tiles, and the sides and arch of the sewer are built of small tiles with thick joints of mortar. The bed of this ancient work, and that of the new sewer being nearly coincident, they were connected on both sides. . . . The substructure of the City wall is rubble, banded at 3-ft. intervals with two thicknesses of large tiles" [Kelsey, Descript. of Sewers, 138].

Sir W. Tite's description of the same discovery adds some further details and is as follows:— "Eastwards of Carpenters' Hall, a mass of rubble masonry, of about 12 ft. in thickness, was cut through; and in the centre was found a culvert, or Roman sewer, in which were discovered three iron bars in perfect preservation, enclosing a human skeleton, the skull of a dog, and the stem of a stag's horn, together with a silver coin of Antoninus and a copper coin of Faustina. Beyond this point the crown of the culvert had been broken in, and a fragment of a rudely wrought column had fallen through the breach. As the ancient sewer passed under houses no further examination could be made in this direction, but on the south side it was not only found to be perfect, but even the mouth of it was discovered under a house at the north-east corner of Carpenters' Buildings. The sewer was constructed of small thin tiles, cemented together by very thick joints of red mortar, made of pounded tile, and having a large pebble inserted in the centre of each. From the top of the sewer to the opposite bank of a ditch into which it discharged itself were placed several pieces of timber scantling in a sloping direction, and a considerable quantity of long moss, undecayed and still retaining a greenish colour, was taken from between them. The ditch receiving the contents of the sewer was made on the south side of the remains of a strong work like part of a fortification, about the site of Little Moorgate or the entrance of Bloomfield Street. As the depth from the present surface to the bottom of the sewer was 18 ft. 4 in., and the open ditch of the fortress was still deeper, it is evident that at the time when they were constructed the adjacent ground was dry and substantial, for the later accumulation of soil was so soft that at one part the bricks could scarcely be laid" [Cat. of Antiq. Roy. Exch., XXXI]. A City Sewers Plan [I, 124] gives the precise position of this culvert.

Roach Smith's account of the discovery of a second "aquaduct" in 1841 is as follows: "In London Wall, opposite Finsbury Chambers, at the depth of 19 ft. [to the extrados of the arch] what appeared to have been a subterranean aquaduct was laid open. It was found to run towards Finsbury under the houses of the Circus about 20 ft. At the termination were five iron bars fastened perpendicularly into the masonry. . . . At the opening of the work towards the city was an arch (Plate 27) 3½ ft. high from the crown to the springing-wall, and 3¼ ft. wide, composed of 50 tiles disposed as shown in the engraving. The spandrels were filled in with rag-stone to afford strength to the work." He estimates the total length of the enclosed as 60 yards [Arch., XXIX, 152]. Finsbury Chambers was the block on the W. corner of Blomfield Street, but its entrance would appear to have been a short distance W. of the junction of the two streets. The second culvert would thus have been very near, if not directly below, that discovered in 1837. The drawing shows that the channel was 24½ ft. below the present surface, or 5 ft. lower than that of the culvert first discovered. It is possible, therefore, that the higher culvert may have been inserted at a later date, when the lower channel had become blocked.

(22). London Wall, opposite Carpenters' Hall. In 1905 a shaft was sunk on the outside face of the wall at this point (Fig. 14). The base of the plinth lay at a depth of 13½ ft. below the street, and rested upon an unusually substantial foundation of rag-stone 5½ ft. deep and projecting 2 ft. from the face of the wall. Above the plinth were four courses of squared rag-stone, a triple bonding-course of brick, five courses of squared rag-stone, a second triple bonding-course and three courses of squared ragstone. In the lower portion of the shaft the relics were exclusively Roman, and lying in the sand overlying the undisturbed ballast were two skulls, one of which was partly imbedded in the mortar of the Roman foundations [Arch., LX, 170].

(23). London Wall, between Throgmorton Avenue and Moorgate Street. The greater part of this long stretch of wall was standing until 1817, when it formed the back-enclosure of Bethlehem Hospital. The destruction of 75 yards of it is recorded at that date, but without details as to its construction [Gent. Mag., 1817, I, 196]. The road was then widened towards the N., and now covers the site of the wall. A view of this section, dated 1812, is engraved in J. T. Smith, Ancient Topography of London, 28. In 1905 telephone-mains were laid, in the core of the wall, from Moorgate Street for a considerable distance eastwards. A manhole, sunk opposite No. 57, was carried down 15¼ ft. through the wall, the base of which would appear to have been one foot lower [Arch., LX, 170].

(24). London Wall, Copthall Avenue. A branch of the Walbrook passed under the wall a little W. of Little Bell Alley (now Copthall Avenue). J. E. Price's description of the remains discovered here is as follows:—"It was at this point adjoining the Swan's Nest Tavern . . . which yet stands, that in the year 1835 an interesting discovery of remains was made. A pit or well was disclosed which had been carefully planked with boards, and which was found to contain a store of earthenware vessels of divers patterns and capacities, together with a coin of Allectus. Some interesting indications of a red brick arch for the transit of water have been observed, but are now entirely gone. This structure was in Bell Alley, and will doubtless be observed in other places. In height it measured nearly 6 ft. and 4 ft. in width. It was supported on either side by massive piles of elm between which the river ran. These were firmly driven into the natural soil and were 6 ft. long, the total depth of the structure being nearly 18 ft. from the level of the street. . . . The black soil which marks the river-bed abounds in bones of animals, including Bos longifrons, etc. The objects found are deposited at Leathersellers Hall" [Builder, 1889, II, 236].

Although the actual structure of the wall is not recorded to have been observed here, the culvert described would appear to be that which conducted the water of the W. branch of the Walbrook into the city.

(25). London Wall, immediately W. of Moorgate Street. In 1882 a stretch of about 43 ft. of the wall was uncovered on the site of a house said wrongly, by Loftus Brock, to be No. 55, and was found to underly the street-frontage. Loftus Brock records that this section was similar to the general type. It was 9 ft. 2 in. thick, including 2 ft. of mediæval thickening on the internal face, and was standing 4 ft. above the surface and extended "quite 8 ft. below." The re-used material described by Brock has been thought to indicate the position of a former bastion, but, in view of the reconstruction of the adjoining stretch of wall described under 26, it is more likely to have formed a part of this later work [Journ. Brit. Arch. Assoc., XXXVIII, 424].

(26). London Wall, E. of Coleman Street. In 1911 excavations on the site of No. 123 revealed a portion of the wall; the outer face had been cut away. The wall reached to within 2 ft. of the pavement-level and extended 10¼ ft. below it [Arch., LXIII, 270].

In 1920 the demolition of No. 122 London Wall, at the E. angle of Coleman Street, exposed a short length of the wall (Fig. 15) which presented some unusual features. The core of the wall was not built with the usual care and contained numerous fragments of brick and roofing-tiles; a very haphazard extra layer of bonding-tiles could be traced under the third course above the plinth. This irregularity was thought not due to rebuilding as the normal double bonding-course was in position above it, but the whole section had been altered and the precise amount of rebuilding done was not easy to determine. The whole face of the wall was exposed at the W. end of the site, the base of the plinth being 14 ft. below the modern ground-level; above the plinth were five courses of squared rag-stone and a double bonding-course of brick; above this point the outer face of the wall had fallen away, and had been made good at some uncertain date by a battering plinth 6 ft. deep and resting on a rough foundation about 2 ft. thick laid against the surviving face of the original wall. Above the batter, the face of the wall resumed the vertical, the face being set back about 10 in. from the original face at the base of the wall. This later vertical face consisted of a double bonding-course surmounted by four courses of rag. Near the E. end of the site a human skull was found half buried, upside down, in the gravel, 2 ft. from the plinth [Arch., LXXI, 73].

(27). Aldermanbury Postern, N.E. side. "In the spring of last year (1857) excavations for the foundations of houses on the north-eastern side of Aldermanbury Postern laid open a portion of the wall of pecular construction, being composed of a series of blind arches, as shown in the annexed cut, prepared from a sketch made on the N. of London Wall, looking towards the street, the present level of which is indicated by the horizontal line below the temporary paling, upon the pavement. The view shows the wall as it appeared while being cut through and excavated up to the street. At first it was supposed there had been openings in the wall, but as the work advanced it was ascertained that the arches were merely constructional as they formed, throughout, part of the solid masonry" [Illus. Rom. Lond., 17]. (Plate 23.)

(28). St. Alphage Churchyard. A stretch of the city-wall still forms the N. boundary of the churchyard, but the whole of the structure above ground is of mediæval date. Roach Smith records the exposure of the N. face of the wall a few years before 1882, and states that nothing of Roman workmanship was visible. A further piece, also of mediæval date, was found adjoining Cripplegate [Brit. Arch. Assoc, XXXVIII, 426], and a third portion still adjoins the bastion in Cripplegate churchyard.

(29). E. of Aldersgate Street. In 1922, during excavations on the site of the Castle and Falcon Hotel and its yard in Aldersgate Street, a piece of the wall was found running E. and W. At the extreme E. end of the property about 210 ft. E. of the street was found a part of the angle-bastion and from this point the wall was traced to Aldersgate Street. It remained in parts undisturbed from the modern ground-level down to the foundations, a depth of about 10 ft., and over its whole length the outer face had been repaired in later Roman times. The outer face, for about 20 ft. W. of the bastion, was badly battered. No trace of a Roman ditch was found [Journ. Rom. Studies, XI, 220 and F.L.]. At one point a brick-lined drain pierced the wall.

(30). W. of Aldersgate. A section of wall at this point and running under the roadway was found in excavating for the French Protestant Church in 1841. Mr. W. D. Saull describes it as consisting of a foundation of flint and clay 1½ ft. thick and 11½ ft. below the surface of the ground, above this 4½ ft. of rubble, a double bonding-course, 2½ ft. of rubble, a second double bonding-course and more rubble above. It is not stated which face of the wall is described and illustrated, but as offsets at the bonding-courses are implied the internal face is no doubt intended [Arch., XXX, 522].

(31). St. Botolph, Aldersgate, churchyard, S. side. In 1887, in clearing a site for post-office buildings, a stretch of 131 ft. of the wall was exposed. The inner face of the wall now forms the N. side of the basement area. A total height of 14 ft. 4 in. of Roman work was seen. The face consisted of the usual triple leveiling-course above the footings, followed by five courses of squared rag-stone, a double bonding-course of brick, five courses of rag, a second double bonding-course, five courses of rag, a third double bonding-course, two courses of rag and at the same interval a fourth double bondingcourse visible in the core of the wall. At each levelling and bonding-course in the usual offset it was estimated that the original thickness at the base was 8 ft. [Builder, 1888, I, 315; Arch., LII, 609].

(32). Christ's Hospital site (A). In 1907–9 a stretch of wall near King Edward Street was exposed. The total height was 10 ft. 2 in., the plinth resting at a depth of 13 ft. 8 in. below the ground-level. Above this point one triple and two double bonding-courses remained. There were somewhat indefinite indications of a bank (Fig. 16) built against the inner face. It extended 16½ ft. from the wall, and remained to a height of about 5 ft. It was composed of an orange-coloured loam which was clearly distinguishable from the surrounding made-earth; where it had been covered by this bank, the face of the wall was remarkably well preserved [Arch., LXIII, 276].

(33). Christ's Hospital site (B). A portion of the wall (Fig. 22) at the back of the bastion (17) was uncovered at the same time. The base of the plinth was 9½ ft. below the surface, and was surmounted by four courses of squared rag-stone, a double bonding-course of brick, five of rag, a second double bonding-course and two courses of rag. The width above the plinth was 8½ ft. [Arch., LXIII, 277].

(34). Christ's Hospital site (C). A further long stretch of wall was uncovered W. of the above. It was of similar character. At one point just below the base levelling-course of brick was a flooring of large Roman tiles laid on a bed of puddle-clay 1 ft. thick, and extending for a distance of 10 ft. from the wall [Arch., LXIII, 280].

(35). Christ's Hospital site (D). At the N.W. corner of the wall, a fragment of the Roman structure (Plate 37 and Fig. 28) adjoining the angle-bastion. It was built on the curve, and the base of the plinth was 12 ft. below the surface. The substructure was upwards of 6 ft. deep. The outer face of the wall "bore evident marks of water having stood against it for a protracted period." On removing this coating the whole face was seen to have been pointed with pink mortar. Above the plinth were five courses of squared rag-stone, a double bonding-course of brick, five courses of squared rag, and a second double bonding-course. The thickness above the plinth was 7¾ ft. The wall had evidently tilted outwards at this point and badly cracked before the addition of the bastion (No. 19) [Arch., LXIII, 286]. This portion of the wall with the anglebastion is preserved in a specially constructed enclosure and can still be inspected.

(36). Immediately N. of Newgate a portion of the city-wall was uncovered in 1875 and recorded by Loftus Brock, though he failed to recognise its significance. It was 8 ft. thick and retained two double courses of bonding-tiles [Journ. Brit. Arch. Assoc., XXXI, 76; XXXII, 385].

(37). Newgate Prison site. On clearing the site for the New Sessions House in 1903, a considerable stretch of the wall (Plate 27), 76 ft. in length, was uncovered together with an isolated fragment farther S. Above the plinth the thickness was 8 ft. and the external face consisted of five courses of squared rag-stone, and a double bonding-course of brick; above this point the face of the wall had been largely destroyed, but a second double bonding-course was visible. On the inside the facing was preserved to above the second bondingcourse. The base of the plinth was 11 ft. below the pavement-level of Newgate Street [Arch., LIX, 125; LXIII, 295].

(38). Warwick Square, W. side. In 1922, during alterations of the premises of the Oxford University Press, the internal side of two pieces of the wall was uncovered; the northern fragment had been robbed of its facing, but the face of the southern was largely intact, showing a double bonding-course with four courses of squared rag-stone above and below it. A fragment is preserved in situ [Journ. Rom. Studies, XII. 258].

A further portion of the wall, near here, was found, before 1880, under the building of Tylor and Son, adjoining the Oxford Arms (now pulled down) [Price, On a Bastion of London Wall, 21n].

(39). Nos. 7–10, Old Bailey. In 1900 a fragment of the wall was uncovered at the rear of No. 8 Old Bailey. It was 8¼ ft. thick above the foundation and was standing 8 ft. high, the top being 18 in. below the pavement-level. Neither the description nor diagram, published at the time, are clear [Lond. and Midd. Arch. Soc. Trans., N.S., I, 354].

The remains of the wall behind Nos. 7, 8, 9 and 10, Old Bailey, were uncovered in 1907–8 on the demolition of the Old Bailey Sessions House, the original base of the wall continuing throughout the site. At the S.E. corner of the site the sandstone plinth was missing and the external face had been repaired in later times, the thickness at the first bond being only about 7 ft. [Arch., LXIII, 295].

(40). South of Ludgate. The line of the Roman wall from Ludgate to the Thames is badly recorded. Two portions, however, appear on the subjoined evidence to have continued the line from the Old Bailey southwards:—

(a) In Playhouse Yard. W. Chaffers, Junr., records that in sewer operations a portion of "old London wall " was exposed. From the context it appears to have run N. and S. in a line with Ludgate. It was 10 ft. thick and "composed of large unhewn stones embedded in a sort of grouting composed of powdered bricks, lime and gravel." The wall was tunnelled through but not destroyed [Coll. Antiq., I, 127].

A second account of the discoveries in Playhouse Yard is contributed by "E. B. P." He states that a wall 8 to 10 ft. thick was found near the W. end of Playhouse Yard near the Apothecaries' Hall and within 100 ft. of it (presumably farther E.) were two others of the same massive character, all three running N. and S. Which of these walls was that seen by Mr. Chaffers is uncertain but E. B. P.'s identification of the western wall with the town-wall is obviously wrong [Gent. Mag., 1843, I, 635]. It is possible that all three walls were part of the Blackfriars Convent.

(b) Under the "Times" Office. The position of a fragment is indicated on a sketch-plan in the Builder, 1855, 221 and 269, showing the line of the wall S. of Ludgate in its relationship to the Times building. Roach Smith describes this fragment as a very thick wall of three distinct constructions. "That of the Roman city-wall; a reparation of considerable solidity, which might be Norman or Early English work; and, above all, the remains of a passage or window which probably belonged to the Blackfriars Monastery." The section was examined during some alterations to the Times buildings [Journ. Brit. Arch. Assoc., V, 155]. The fragment has been subsequently destroyed.

(41). Upper Thames Street (A). Roach Smith records that in 1841 sewer excavations began at Blackfriars, but until the foot of Lambeth Hill was reached nothing was encountered; they "were there checked by a wall of extraordinary strength, which formed an angle with the hill and Thames Street; upon this wall the contractor was obliged to open his course to a depth of about 20 ft. so that the greater portion of the structure had to be overthrown. ... It extends, as far as I had the means of observing, from Lambeth Hill to Queenhithe, with occasional breaks; in thickness it measured from 8 to 10 ft.; the heights from the bottom of sewer was about 8 ft., in some places more or less; it reached to about 9 ft. from the present street, and 3 ft. from that which indicates the period of the Fire of London. . . . The foundation was made in the following manner: oaken piles were first used; upon these was laid a stratum of chalk and stones and then a course of hewn sandstones, from 3 to 4 ft. by 2 and 2½ ft., firmly cemented with the well-known compound of quicklime, sand and pounded tile. Upon this solid substructure was built the wall composed of rag and flint with layers of red and yellow, plain and curve-edged tiles. . . . Many of the large stones, above mentioned, are sculptured and ornamented with mouldings, which denote their prior use in a frieze or entablature of an edifice, the magnitude of which may be conceived from the fact of the stones weighing in many instances, upwards of half a ton. ... I observed also fragments of sculptured marble had been worked into the wall, and also a stone carved with an elegant ornament of the trellis-work pattern, the compartments being filled alternately with leaves and fruit ...." [Arch., XXIX, 150; Illus. Rom. Lond., 19]. The fragments of marble pilasters and the fragment with trellis-pattern (Plate 51) are now in the British Museum.

A piece of this wall (Fig. 17) was re-opened in October, 1924, in the construction of a sewer under Brooks Yard from Upper Thames Street, when the S. wall of the city was tunnelled through. The foundation was laid between two rows of contiguous piles the tops of which were 14 ft. below the roadway in Thames Street; the total depth of the tunnel being 16 ft. The wall is of a concrete of Kentish rag-stone with a course of bricks a few inches below the tops of the piles. A second course of bricks was found 2 ft. above that just described. Fifteen feet to the N. of the main wall, and parallel to it was a second wall 5 ft. thick, and with the foundation also between two rows of piles, but set apart. A thick bonding-course occurred just above the heads of the piles, and above this the wall was battered or coped back on both sides and finished with a flat top 2 ft. wide. On the S. face of this wall was a mass of puddled clay [Q. W. and Times, June 18th, 1925].

(42). Upper Thames Street (B). J. T. Smith records that "In June, 1839, the labourers engaged in deepening a sewer in Thames Street, opposite Vintners' Hall, in the middle of the street, at a depth of 10 ft. from the surface, discovered the perfect remains of an old Roman wall, running parallel with the line of the river. The wall was formed of alternate layers of flint, chalk and flat tiles " [Streets of London, 380].

(43). Upper Thames Street (C). Crossing Queen Street, Roach Smith saw a fragment of wall "precisely similar in character" to the length described under No. 41 [Arch., XXIX, 150; Illus. Rom. Lond., 19].

(44). Upper Thames Street (D). Under Cannon Street Station a wall 200 ft. long was discovered in 1868. It may have formed part of the city-wall, but the position and direction are not definitely recorded [see Inventory, p. 113].

(45). Upper Thames Street (E). In 1927, between the ends of Bush Lane and Little Bush Lane, a foundation of chalk blocks was encountered and an indeterminate edge on the S. side seemed to trend more N. of E. than the line of the trench. This foundation may represent either the foundation of the river-wall or the debris fallen outwards.

(46). Upper Thames Street (F). In 1863, at or near the S.E. angle of Suffolk Lane, a wall was found which was regarded as part of the river-wall, as described by Roach Smith and Tite [Arch., XL, 48].

(47). Lower Thames Street (A). Under the frontage-line of No. 125 Lower Thames Street and the adjoining pavement, a portion of the wall (Fig. 18) was exposed in 1911. The wall rested on the ballast at a depth of 24 ft. below the present surface. "Large roughly-squared timbers, 12 ft. long and about 8 in. square, were first laid on the top of the ballast, across the thickness of the wall, these being held in position by pointed piles driven in at intervals. . . . On these timbers were laid large irregular sandstones and rag-stones bedded in clay and flints. Three layers of these stones showed on the face above which was a bond of two (the drawing shows three) rows of yellow tiles. Some chalk with other stone formed the core, the whole being cemented with red mortar. The total height of the masonry remaining was 3 ft. and its width 10 ft. Some of the stones were apparently re-used though no moulded stone appeared in the small piece uncovered " [Arch., LXIII, 309].

(48). Lower Thames Street (B). Under the roadway immediately S. of the Coal Exchange a wall about 7 ft. thick, which may have been part of the city wall, was encountered in 1859. It was built of rag-stone but no other details of its construction are recorded [Journ. Brit. Arch. Assoc., XXIV, 295].

The Ditch.

(N.B.—The positions of the items, numbered consecutively in this section, are shown by the same numbers, prefixed by the letter D on the large Plan A at the end of the volume.)

(1). America Square. In 1908 excavations on the site of Nos. 15 and 16 America Square revealed in two places the section of the Roman city-ditch cut in the gravel (Figs. 19 and 20). The inner edge was 12 ft. in advance of the face of the wall, and the ditch was about 10 ft. wide and 5 ft. deep. It was of the usual V-shaped section, the lower portion being filled with a clear clayey deposit containing fragments of Roman tiles and pottery, including one piece of "Samian." Above this layer was a black band in which was a tightly packed mass of minute snail-shells. Apparently overlaying the ditch was about a foot of disturbed gravel; above this again was about 6 ft. of clean light loam, containing plentiful remains of the Roman period but nothing later. The description here summarized does not precisely tally with the section which accompanies it. The second section made at the N. end of the same site showed how rapidly the original surface rose towards the N., there being a difference in level of 3¼ ft. in a distance of 45 ft. [Arch., LXIII, 262].

(2). New Broad Street, W. A considerable stretch of the Roman city-ditch was examined in 1906 to the S. of this street. Eight sections in all were made showing that the ditch (Fig. 21) was of uniform character, 16 ft. wide, 4½ ft. to 5 ft. deep, the inner edge being 15 ft. in advance of the face of the wall. The form was V-shaped and the filling was a light sandy soil, containing a fairly abundant quantity of "Samian" and other Roman pottery, and no relics of a later age. This was held also to prove that the marsh-conditions which subsequently obtained in this district were not present when the ditch was filled up [Arch., LX, 212]. The footings of the bastion under the vestry of All Hallows Church oversailed the inner edge of this ditch (Fig. 27), the filling beneath being composed of black mud, chalk-rubble and rubbish in which oyster-shells, Roman pottery and tile frequently occurred [Arch., LXIII, 272]. Slight traces of a second and later ditch seem to have been discovered at the extreme W. end of the site; they corresponded with the remains of the second ditch described under Item 4 [Arch., LXIII, 279].

(3). Aldersgate Street. The ditch to the W. of the street was examined and recorded in 1888 by G. E. Fox. The inner edge was at a distance of 10¼ ft. from the wall. "The total width of the ditch across the top was 74½ ft., the flat bottom 35 ft., and the total depth 14 ft. 1 in. The sides sloped at an obtuse angle. Both sides and bottom had a clay-puddling 6 in. thick. Another section of the ditch was obtained close to Aldersgate Street. This revealed a curious feature. In the bottom of the ditch appeared a slightly raised mound of unknown length, as it ran under the street and could therefore only be traced for a short distance. It was 2¾ ft. high and 7 ft. 2 in. broad at the top and 11 ft. 10 in. at the base, and was traceable for a length of about 10 ft. The surface was puddled like the rest of the ditch. It was not placed in the middle of the ditch, but was nearer the outer than the inner margin." It was supposed to have formed the base to a wooden trestle-work forming a support to a bridge. The ditch was dug through a stratum of clay, and penetrated 2¾ ft. into the gravel beneath [Arch., LII, 615]. Though accepted as the Roman ditch by more than one recent authority, the unusual dimensions throw grave doubt upon this attribution.

(4). Christ's Hospital site. During the excavations of 1907–9, two sections of the first Roman ditch were uncovered to the E. and W. of the first bastion W. of King Edward Street. The first section (Fig. 22) was of the usual V-shaped form 12 ft. wide, 6½ ft. deep and about 11 ft. from the external face of the town-wall. The foundations of the bastion were carried down through the inner half of this ditch, the outer edge of which had been destroyed. The ditch was dug in the brick-earth, and was filled with black earth. The second section (Fig. 23) uncovered disclosed the remains of a second and later ditch also of V-shape, about 25 ft. wide, 14 ft. deep and 11¼ ft. in advance of the town-wall. This ditch had entirely destroyed the earlier ditch with the exception of a narrow sector on the inner side which showed that at this point the earlier ditch approached to within 10¼ ft. of the town-wall [Arch., LXIII, 276].

(5). Newgate Prison site. Very doubtful indications of a ditch about 25 ft. wide and some 55 ft. from the external face of the wall are recorded to have been found on this site in 1903–4 [Arch., LIX, 137]. A subsequent account, however, states that no regular ditch had been cut here, and that the surface of the gravel lay unevenly all over the site at depths varying from 17 ft. to 23 ft. [Arch., LXIII, 297].

The Gates.

(N.B.—The positions of the gates, numbered consecutively in this section, are shown by the same numbers, prefixed by the letter G on the large Plan A at the end of the volume.)

(1). Tower Postern is said by Stow to have been formerly on the main line of communication from E. to W. of the city. It was undermined by the digging of the tower-ditch, c. 1190, and partly fell down in 1440 [Stow, Survey (ed. Kingsford), I, 28]. There is no evidence of the original date of this structure.

(2). Aldgate. This gate is first mentioned temp, Canute [Cal. of Letter Books, C, 217]. The plan of the mediæval gate, perhaps that re-built from the foundations by Norman, Prior of Holy Trinity 1108–47 [Guildhall MS., 122, fol. 13, cited in Stow's Survey (ed. Kingsford), II, 274], is partly preserved in the Elizabethan plan of Holy Trinity, Aldgate, at Hatfield House [Home Counties Mag., II, 46]. It was of rectangular form with two semi-circular towers projecting on the external face. This gate was pulled down in 1606 and re-built in 1610 when "two heads done after antique models" were found [Soc. Antiq. MS. Min. VIII, 25a]. In 1907, in driving a sewage-tunnel under the roadway on the S. side of Aldgate High Street, some solid masonry was encountered at a depth of 16½ ft. below the surface. It consisted of work of two periods, one built against the other, and had to be tunnelled for a distance of 16 ft. The portion directly under the houses was comparatively modern and contained mediæval material; the other portion was of rag-stone very solidly built with hard white mortar, and containing fragments of Roman tile. At 10 ft. from the house-fronts, under the roadway, a built face of dressed stones, varying from 9 in. to 2½ ft. in length was found running diagonally in a south-easterly direction, but was not uncovered for a length of more than 2 ft. to 3 ft. It seems probable that this may have formed part of the base of a flanking tower [Arch., LXIII, 266]. The style of building appeared to be similar to that employed in the base of the bastions of the wall, and it is possible that the remains were those of the late Roman gate [V. C. H. London, I, 53].

(3). Bishopsgate. This gate is first mentioned in Domesday Book. The fact that it is approximately on the line of the Ermine Street seems to imply its Roman origin. In 1905, in connection with operations for laying telephone-mains, a manhole was formed near the N. angle of Wormwood Street and Bishopsgate Street, near the site of the gate. Here at a depth of 5 ft. a mass of rubble masonry was encountered extending 2 ft. into the ballast which here lay at a depth of 8 ft. On its S. side were indications of a carefully built face. The materials were rag-stone rubble with some portions of Roman tile, and below the whole mass was a puddling of flint and clay. Cutting into this mass of masonry and resting on it was a culvert of rag-stone, 2½ ft. wide and 1¾ ft. high; the floor was slightly hollowed, and the covering was of single slabs of stone. The presence of thin tiles in this structure indicated its mediæval date. From this fact and the presence of the puddling, it seems fairly certain that the mass of masonry was of Roman date and formed part of the S. face of a gate projecting some 20 ft. within the inner face of the town-wall [Arch., LX, 185]. A fragment of walling which may also have formed part of the gate was found in 1921 on the N. side of No. 108 Bishopsgate Street, i.e. on the E. side of the internal projection of the gate. Here was found at a depth of 3 ft. "Roman work 3 ft. thick and 4 ft. wide, with large hard stones and well-rammed clay 3 ft. lower." It appeared to be part of a wall about 5 ft. thick, and was apparently at right angles to the town-wall, though the published description is not easily intelligible. The wall contained red bricks and the stones were apparently squared [Trans. Lond. and Midd. Arch. Soc, N.S., IV, 332].

(4). Moorgate. This gate was an enlargement, made in 1415, of an earlier postern of uncertain date [Stow, Survey (ed. Kingsford), I, 32; II 274]. A cutting made by the Metropolitan Water Board in 1925 encountered a mass of concrete some 10 ft. in advance of the town-wall at this point, and composed of broken bricks (mostly 1½ in. thick) and tiles, rag-stone and septaria. It stood upon a timber raft, 3 in. thick, resting on 9 in. of rammed chalk; the total depth was 14¼ ft. below the pavement. It would appear probable that this was some portion of the mediæval gate.

(5). Aldermanbury Postern is said to have been formed in 1655, but the existence of something of the same sort in Roman times has been conjectured owing to the proximity of the curious wall-arches recorded by Roach Smith in 1857 [V.C.H. London, I, 62]. There seems to be little basis for this conjecture.

(6). Crippiegate. This is one of the three gates mentioned in the Laws of Ethelred, c. 1000 [Thorpe, Ancient Laws and Institutes, 127], and it may consequently go back to Roman times. No remains of this age have, however, been recorded.

(7). Aldersgate. This gate is mentioned as Ealdredesgate in the laws of Ethelred, c. 1000. In the excavations of 1887 (see Town-Wall No. 31) Mr. Fox recorded the existence in the middle of what he assumed to be the Roman ditch, close to Aldersgate, of a raised mound (p. 94), which was thought to have formed the support for a wooden trestle-bridge. If this evidence be accepted, it of course implies the existence of a Roman gate at this point. In 1924, a shaft was sunk for a new sewer on the line of the city wall in Aldersgate Street; concrete was found at a depth of 11 ft., and below this was a foundation 7 ft. thick composed of lumps of chalk and below this again were piles 8 or 9 in. in diameter [Q.W.].

Excavations on the E. side of the street in 1922–3 revealed the existence of what was probably a mediæval barbican with a polygonal turret at the N.E. angle, and a tunnel extending under the road, immediately to the W. of it. The barbican projected some 36 or 37 ft. in advance of the city-wall, but contained no work of a date recognizably anterior to the 15th century [A.C.].

(8). Newgate is almost certainly to be identified with the Westgetum of a charter of 857 [Stow, Survey (ed. Kingsford), II, 276]. Remains of the Roman gate (Fig. 24) at this point have from time to time come to light, providing sufficient evidence to reconstruct the general plan of the building. The discoveries made in 1875 were recorded but misunderstood by Loftus Brock; they consisted of portions of all four walls of the northern guardroom which measured 32 ft. by 30 ft. externally, and 22 ft. by 15 ft. internally. The walls at the N.W. angle still retained a double bonding-course of tiles. These remains were incorporated with the masonry of the mediæval gate which projected considerably farther on the outside of the town-wall [Journ. Brit. Arch. Assoc, XXXI, 76; XXXII, 385; Arch. Journ., XXXII, 477]. The remains found in 1903 and recorded by Dr. P. Norman consisted of a portion of the E. wall and plinth of the S. guardroom, including the S.E. angle. The base of the plinth was 6½ ft. below the pavement-level, and the plinth itself was composed of an oolitic stone closely resembling Barnack. Under the plinth was 1 ft. 10 in. of rag-stone walling resting on a foundation of puddled clay with fragments of rag-stone, nearly 5 ft. deep [Arch., LIX, 130]. In 1909, a portion of the plinth of the W. wall of the N. guardroom was uncovered and recorded by Messrs. Norman and Reader; it was similar in depth to the portion found in 1903 [Arch., LXIII, 294]. The results of these various discoveries are sufficient for the reconstruction of the plan of the gate which follows a normal Roman type with square flanking guardrooms. A double entrance is indicated by the distance, 35 ft., between the guardrooms. The N. guardroom projected 16 ft. within the town wall as compared to the 20 ft. at Bishopsgate, but the S. guardroom projected only some 8 ft. There can be no doubt that this gate was of Roman date, but that it was considerably later than the Roman town-wall is indicated by the difference in level of the respective plinths, that of the town-wall being 4½ ft. below the plinth of the gateway.

(9). Ludgate. There is no very early reference to this gate, which must, however, have been one of the seven double gates mentioned by Fitzstephen, temp. Henry II. That it was of Roman origin is indicated by the presence of burials in the neighbourhood of Fleet Street. No structural remains of this age have been found, but the discovery of the Roman sepulchral stone (Inscriptions No. 15) in the immediate neighbourhood may indicate a late building or rebuilding of the gate.

The Bastions of the Town-Wall.

(N.B.—The position of the bastions, numbered consecutively in this section, are shown by the same numbers, prefixed by the letter B on the large Map A at the end of the volume.)

(1). Wardrobe Tower in the Tower of London. The mediæval tower incorporates the base of a Roman tower (Fig. 11) of semi-circular plan and apparently hollow. This base consists of a mass of rubble masonry 5 ft. high, consisting of stone and broken Roman brick, with brown and red mortar, quite distinct from the white mortar of the mediæval reconstruction [Loftus Brock, Journ. Brit. Arch. Assoc., XXXVIII, 130].

(2). E. of Tower Hill. In 1852, during excavations on the eastern or outer side of Trinity Place, was found what was described as a quarry of 125 stones or a mediæval buttress, resting against the Roman wall which was in remarkably good condition. Roach Smith states that the buttress was in a great measure composed of stones which had belonged to Roman buildings of importance and to sepulchral monuments. Pink mortar is said to have been found on the face of the wall. From sketches in the Builder, it appears that the stones included cornices, column-drums and the monumental inscription, No. 6, p. 171. Although Roach Smith called the structure mediæval, its composition and situation leave little doubt that it was one of the usual semi-circular bastions added to the wall. An engraving (Plate 22) by Fairholt in Roach Smith's Illustrations shows the remains of the bastion still in situ against the face of the wall. This bastion is not shown on Ogilby and Morgan's survey [Builder, 1852, 562; Journ. Brit. Arch. Assoc., VIII, 241; Roach Smith, Illus. Rom. Lond., 15]. What was probably a surviving portion of this bastion was destroyed in the excavations for the Inner Circle Railway, 1882–5; it contained sculptured stones [Antiquary (1885), XI, 33].

(3). America Square, W. side. On the widening of the London and Blackwall Railway in 1880, a bastion in this position was destroyed. It is shown on Ogilby and Morgan's survey, but the only record of its form and situation is an unpublished drawing (Plate 29) by Henry Hodge, 1881, now in the Guildhall Library [Add. Prints, p. 98]. From this drawing it appears that the bastion, which was preserved to a height of 1½ ft. above the excavationlevel, was of slightly horseshoe-form, 21¾ ft. in diameter, and projected 14¾ ft. from the face of the wall. It was "built with rag, flint, chalk, brick, etc., grouted with grey gravelly mortar like Thames ballast. The facing was of rag, flint and limestone —all very smooth stone and random work. The core or main structure was also of small material," and contained two fragments of shaped coping in oolite. On each side of the bastion, 1 ft. from the ground-level of the excavation, was a patch of pink mortar extending about 2 ft. into the structure.

(4). N. of John Street. The bastion here is known only from Ogilby and Morgan's survey.

(5). Jewry Street. A bastion immediately N. of the present site of the Cass School is indicated on Ogilby and Morgan's survey. It is, no doubt, also that described by Maitland [Hist. of London, 1756, 31] as the basis of a "Roman tower about 8 ft. high which supports a new building," at the lower end of a street called the Vineyard. He adds that a tablet on the building stated that when the upper part of the tower, three storeys high, fell down no one was hurt. This part of the street is now called Little George Street.

(6). Duke's Place, E. This bastion is shown on the Elizabethan survey of Holy Trinity, and also on Ogilby and Morgan's Plan as of semi-circular form. It is described by Maitland as being 21 ft. high and of similar construction to No. 7, that is to say, with brick bonding-courses, "the bricks being as sound as newly laid," though the stonework was in bad condition. The foundations of this bastion were probably those seen by Loftus Brock in 1887, but his account of the position is complicated by an apparently wrong reference to the Jewish Synagogue in Bevis Marks. He describes the bastion as "built of large blocks of freestone worked to smooth surfaces, some being so well rounded as to warrant the belief that they were shafts of columns [W. Maitland, History of London (1756), I, 31; Journ. Brit. Arch. Assoc., XLIII, 203; Arch., LXIII, 340.]

(7). Duke's Place, W. This bastion also is shown semi-circular in the two plans mentioned under No. 6. It is described by Woodward in 1707 as follows:—"Tis composed of stone with layers of brick interposed, after the Roman manner .... being about 26 foot in height." The position of Woodward's bastion is fixed by Maitland, who says it was almost opposite the end of Gravel Lane, and that the back fronted a passage into Duke's Place. An etching of this tower by E. F. is preserved with a copy by Gough, dated 1763, in the Gough MSS. (Plate 28), and was engraved long subsequently by Fairholt. The etching shows a semi-circular tower of stone with triple bonding-courses of brick (four in all) at approximately regular intervals. The top part of the tower is shown polygonal in form, and was evidently of later date. The entirely misleading reproduction of Fairholt has led to the mistaken supposition that the tower was rectangular [Woodward's Letter to Hearne; Maitland, History of London (1756), I, 31; Roach Smith, Illus. Rom. Lond., 16; Gough MSS., Bodleian Lib. Gen. Top. 62, 16 and 16a and Map 21, 41].

(8). Bevis Marks, near E. end. During the rebuilding of No. 31 Houndsditch in 1880, the base of a bastion was found at "the N.E. of these excavations" (the compass-point should probably be S.E.), and it is described as "of later date [than the wall] and rougher but still probably of Roman work." It projected 18½ ft. from the outer face of the city wall. The width is said to have been 40 ft., but this is obviously an error. Its face was "a flat segment of a circle." Built up into it were some fragments of Roman architectural work, including the base of a circular column, a shaft, 9 in. in diameter, with trellis ornament in relief, and an inscribed stone (see p. 174, No. 23). Red mortar "was observable in some part of the bastion, as if used sparingly, and not as if it had adhered to the stones on their removal from some other building." A massive channel of solid stone, 1½ ft. broad and 1¼ ft. deep, led from the centre of the bastion to the ditch, and "traces of a raised earthen bank like an external vallum to the ditch" were found [see Loftus Brock, Journ. Brit. Arch. Assoc., XXXVII, 86; XXXVIII, 132–5].

(9). Castle Street (now Goring Street). The base of a bastion on the W. side of Castle Street was uncovered in 1884. The only record of it is a careful survey (Plates 31 and 32) probably by Henry Hodge, in the possession of Dr. Philip Norman. The foundation consisted of flint and puddled clay surmounted by a bed of chalk. The bastion itself was 26 ft. wide and projected 15½ ft.; it was standing nearly 11 ft. above this foundation which was 5 ft. below the base of the plinth of the adjoining town-wall. The facing of the bastion consisted apparently of coursed rag-stone battering outwards towards the base, and standing on a projecting footing-course. At a height of 8½ ft. above the foundation was a double course of bricks not carried through the wall. The bastion, as far as it was standing, was solid, the filling containing numerous lengths of shaped stone coping, a cornice and other moulded stones, a fragment of an inscription and a fragment of frieze carved with swags and running hares (Plate 21). The inscription (see Inscriptions No. 21), the frieze and probably also the cornice are now in the Guildhal Museum [Antiq. Journ., VII, 518; Antiq., X, 134].

(10). Camomile Street. In 1876 building opera tions revealed the remains of a semi-circular bastion (Figs. 25 and 26) 20 ft. in diameter and projecting 14 ft. 9 in. from the face of the Roman town-wall, below the plinth of which its base was carried about 4 ft. "It rested upon the natural soil of London clay, which had been simply levelled by compressing together masses of chalk into the clay, for a thickness which varied from two to three inches. With the exception of huge blocks of oolite and green sandstone, which formed the nucleus of the structure, the stone employed in building was the familiar Kentish rag-stone rubble, with a facing in random courses of the same material. The size of the blocks of which this facing was composed varied from three to eight and a half inches thick and from five to fourteen inches long. . . . Though shown to be of later date than the erection of the [Roman] wall, and separated from it in places by an intervening space filled in with rubble, there were yet signs that the masonry of the bastion had been toothed or chased into the wall for the purpose of acquiring solidity and strength." To its surviving height of 10 ft. the tower was solid. It had a projecting stone footing of about 8 in., and its structure included large blocks of re-used pink cement and masonry (Fig. 26)—fluted pilasters, shafts of half-columns, portions of canopies, cornices, door-jambs, etc.—together with the sculptured figures of a soldier and a lion, and a human head of colossal size, all now in the Guildhall Museum (Plates 7, 11, 15, 16 and 17).

J. E. Price thought that the tower was mediæval on the ground that the handle of a pitcher of green-glazed ware was found "beneath the lowest bed of stone, and near to the centre of the structure." But even if this observation was accurate, the sherd may well have been Roman. No bastion is shown here on Ogilby and Morgan's survey [J. E. Price, On a Bastion of London Wall, 1880].

(11). All Hallows Vestry. In 1905 the original masonry of the bastion (Plate 30 and Fig. 27) was found by excavation beneath the structure of the vestry, and subsequently the whole of the external face of the bastion was uncovered. Its diameter was 19 ft., and its projection 15 ft. The main structure survived to a maximum height of rather more than 8 ft., of which 3 ft. extended below the top of the chamfered plinth of the Roman town-wall. It was of random rubble with white mortar, and rested upon a plinth of re-used ashlar of larger size, which in turn rested upon a rectangular platform set in, and in places covered by, pink mortar. The platform also consisted of re-used ashlar, mostly of thin stones laid flat; and it oversailed the original Roman ditch, which had been filled for the purpose with chalk, flint and broken stones. The re-used masonry included an angle-pilaster with a moulded cap (Plate 38), several blocks with a marginal fillet, and many with sockets known as lewis-holes. Very few relics were found in the soil from just above the surface of the platform, "but it did not appear to have been disturbed since it was laid down. There were some portions of roofing tiles and a few fragments of Roman pottery, including some Samian, also oyster shells and bones. Similar relics were found in the ditch. Under the layer of chalk and flint, the ditch-filling was black mud, containing snail-shells and remains of rushes. There was also a horse's skull and a human femur." Beyond the extent of the platform the ditch had not been filled with chalk and stones; "it has clearly remained open for some time after the building of the bastion, accumulating mud and rubbish against the obstruction of the bastion footings." This bastion is not shown on Ogilby and Morgan's survey [Norman and Reader, Arch., LX, 200; LXIII, 271].

(12). St. Giles's Churchyard. This bastion, situated at the angle of the wall, is hollow and has an external diameter of about 37 ft. Excavation about 1900, showed that the base of the tower (Plate 33), the upper part of which is still visible here, extended to a depth of 18 ft. below the present surface, giving a total height of 31 ft. to the structure. "The foundations (which are on ballast) and, indeed, the lower portions of the wall to a height of about 4 ft., are in a good state of preservation, and judging by the appearance of the materials used, particularly the mortar, this portion is probably Roman work. Above this height the work was of a different character, intermixed with pieces of Roman tiles and flints, and in some instances the stones had been wedged up with several layers of oyster shells, the mortar being of an inferior quality to that found at a lower level, and there is not the slightest indication of this portion of the bastion being the work of the Romans, although full of their materials" [J. Terry, Lond. and Midd. Arch. Soc. Trans., (N.S.), I, 356].

(13). Between Well Street and Monkwell Street. An entry in the records of the Barber Surgeons' Company states: "5th February, 1607. This day it is ordered that a courthouse be erected upon the Bulwarke behind the Hall of this Company for the Mrs. or Governors to kepe the Courte at the charge of the Company." This building was destroyed in 1864, and the site is now covered by a rounded warehouse. A view of the bastion, dated 1800, is included in Smith's Antiquities of London, but the external face was then, apparently, rendered in cement.

(14). Windsor Court, Monkwell Street. No. 2a, known as "Bastion House," occupies the site of a bastion found in 1865. It is described as a semi-circular tower, about 40 ft. high. "The materials are rough flint and Kentish rag-stone, but .... high up in the elevation, as shown in the engraving (Plate 33), is a bonding-course of tiles resembling Romas bricks, upon which is a floor of modern rooms. In the face of the tower were appertures which have been filled up with brickwork" [Illus. Lond. News, Aug. 19th, 1865].

(15). East of Aldersgate. In 1922 the base of the bastion shown by Ogilby and Morgan in the re-entrant angle of the town-wall here was uncovered, but there is no detailed record of its structure [Journ. Rom. Studies, XI, 220]. Mr. F. Lambert states definitely that it was hollow and that only about two-thirds of the curve were uncovered, the rest turning under a modern wall on the edge of the site; it was not explored to its base. Dr. P. Norman states that the W. face of the bastion was flattened, indicating that the building was of horse-shoe form.

(16). Near King Edward Street, beneath the houses on the E. side of the street were discovered in 1887, the foundations of a hollow semi-circular tower. They were 5¼ ft. wide, and composed of Kentish rag with some chalk and a few fragments of old building-material. The internal measurements were 17¼ ft. by 16 ft. "Some pieces of worked stone discovered [in the foundations] showed traces of Norman mouldings and of foliage of the Early English Period." The staircaseprojection forming part of the existing post-office building marks the site. This tower is not shown on Ogilby and Morgan's survey [G. E. Fox, Arch., LII, p. 610].

(17). Site of Christ's Hospital, E. end. In 1908–9, part of the base of a solid bastion of slightly horseshoe plan was found here. It was about 26 ft. in diameter, projected 16 ft., and was built of random rubble, consisting of rag-stone, flints, fragments of Roman tile, etc. The foundations were carried down 7 ft. below the base of the plinth of the town-wall, and rested on undisturbed ground [Arch., LXIII, 276].

(18). Site of Christ's Hospital, middle part. The same excavation (1908–9) uncovered the remains of a second bastion (Plates 34 and 35), hollow and of horseshoe plan. Its wall was 5½ ft. thick, and the internal diameter was 13 ft. The base of the structure lay at a depth of nearly 10 ft. below the level of that of the Roman town-wall and was without footings [Arch., LXIII, 281].

(19). Site of Christ's Hospital, angle bastion (Plates 34, 36 and 37, and Fig. 28), now preserved below the courtyard of the G.P.O. It is of horseshoe plan with a projection of 26 ft., and is hollow, with walls 7 ft. thick at the base which is irregular (conforming to pre-existing hollows in the ground), and reaches a maximum depth of 7 ft. below the base of the plinth of the Roman townwall. The masonry of the tower was rag-stone set in good white mortar; no re-used stones were found. The external face was carefully pointed and smoothed, whilst the internal face was irregular and unpointed. The filling of the bastion below a level of 10 ft. from the present surface, with the exception of an easily distinguished sump-hole of the 16th or 17th centuries, contained only Roman objects. The upper part, which was apparently the artificial filling of the bastion, included masses of opus signinum flooring, fragments of rag-stone with mortar adhering, roofing-tiles, much Roman pottery, etc. The lower part had apparently not been disturbed by the building of the bastion, and represented surface-deposits antedating it. It contained a few fragments of Roman pottery. This bastion is shown on Ogilby and Morgan's survey [Norman and Reader, Arch., p. 286].

(20) and (21). Between Newgate and Ludgate, two bastions are shown on the map by John Leake, etc., engraved by Hollar in 1666. The more northern of the two was opposite Fleet Lane.