An Inventory of the Historical Monuments in London, Volume 3, Roman London. Originally published by His Majesty's Stationery Office, London, 1928.
This free content was digitised by double rekeying and sponsored by English Heritage. All rights reserved.
Terra Sigillata Prior to A.D. 60. By T. Davies Pryce.
The red-glazed pottery, "terra sigillata" or "Samian" ware, which is the characteristic ceramic product of the earlier centuries of the Roman Imperial period, is abundantly represented in London. Indeed, no site in Britain has been so productive of this class of ware, whether it be regarded from the point of view of quality, quantity or variety. All the variations of technique as applied to this fabric are forthcoming, i.e. moulded, marbled, incised, applied, rouletted and stamped ware. Owing to the conditions under which excavations have been made in the city, much of this fabric has found its way into the hands of dealers and private collectors, and a considerable proportion has been scattered throughout the country. Some of it will also be found in the offices of City Companies, etc. Notwithstanding these leakages, this ware is well represented in the Guildhall, British and London Museums, and to a lesser extent in the Bethnal Green, South Kensington and other Museums.
No discussion of the whole of the terra sigillata of London is undertaken in this context, but an endeavour is made to isolate the earliest examples with a view to determining the date of the initial occupation of the site. The absence of any direct historical or epigraphical evidence on this point gives a special value to the ceramic materials, and in particular to this well-known red-glazed ware.
This fabric, as occurring in London, naturally falls into the categories of Italian (chiefly Arretine) and Provincial (chiefly South Gaulish) ware. Before discussing the evidence under these headings, it is necessary to point out that in the case of London we are dealing with the results of desultory excavations which have been the subject, with one or two exceptions, of very imperfect record. Considerable reservation is therefore necessary in their topographical application. Thus, one locality may bulk large as the result of the accident of excavation or of a fuller and better record, whilst in the absence of these circumstances other areas of perhaps equal importance may remain more or less silent.
(1) Italian Ware.
This fabric is the prototype of Provincial ware. The flourishing period of the Italian potteries, especially those of Arretium, can be assigned with some confidence to the last third of the 1st century B.C. and the first two decades of the 1st century A.D., but the industry continued down into the Nero-Vespasian period. (fn. 1) During the earlier part of this period—the Augustan age—this ware was extensively exported into the provinces. At Haltern, which was occupied from 11 B.C. to A.D. 16, this ware alone is represented. During the latter part of this period, i.e. circa A.D. 40 to 80, this ware, to some extent, supplied the home market, but even here it was exposed to the competition of an increasing importation from the provincial potteries of Southern Gaul, and it is highly probable that the exportation of the Italian fabric, as a contemporary product, through the ordinary channel of commerce had become exceedingly slight by the accession of Claudius (A.D. 41). (See below, p. 181.)
(1) AMAR: Cup of Loeschcke type 8 (fn. 2) and Ritterling type 5 (fn. 3) Fig. 91, A.l. Found in Leadenhall market; now in the London Museum. The rim and the wallmoulding lack the rouletting which is common to Augustan examples of this form as found at Haltern; it is therefore probably later and should be dated to the first half of the 1st century A.D. No exact parallel for this stamp can be discovered but those of AMA and AMAB occur at Arezzo and Rome, respectively (CIL. XI, 2, 1; 6700, 25; XV. 2; 4950). (fn. 4)
(2) ATEIVS: ATEI. in planta pedis, Cup, Loeschcke type 12 and Dragendorff (fn. 5) type 24/25, Fig. 91, A.2. Found at London Bridge Station; now in the Bethnal Green Museum. One hundred and two stamps of the potter Ateius and his slaves have been found at Haltern in the Augustan period (11 B.C. to A.D. 16). The stamp of ATEI/XANTHI occurs at Mont Beuvray, which was abandoned circa 5 B.C. He is represented at Xanten and in the early period at Wiesbaden, both of which sites date to the reign of Augustus. Twenty-one of his impressions have been found at Sels (ante A.D. 41). (fn. 6) Rarely, and then probably as a "survival," this stamp occurs on a site which was first definitely occupied in the reign of Caligula, e.g. Grimmlinghausen (A.D. 40). (fn. 7)
In Britain the stamps of ATEIVS and his slaves also occur ten times at Silchester, once at Pleshey in Essex and at Foxton in Cambridgeshire. (fn. 8) The London stamp, with its label in the form of the sole of a foot, dates to the earlier third of the 1st century A.D. (fn. 9)
(3) CORNELIVS: CORNEL, on a crater, Fig. 91, A3. Found in London, 1837, probably in Southwark (cf. Brit. Mus. Cat., L160). The vessel has an everted rim and its wall is divided into two zones by a rouletted moulding. On the upper zone are two conjoined dolphins which also overlap the moulding. Beneath them are the stamp and a wolf to L. Similar dolphins occur at Arretium (cf. Oswald and Pryce, op. cit., XXIV, 7). This well-known potter worked at Arezzo in the Augustan period, and later. His stamp occurs at Haltern (11 B.C. to A.D. 16). His usual signature is P. CORNELIVS, but stamps in which the P. is omitted, as on the London vessel, occur at Arezzo (CIL. XI, 2, 1; 214, 215, 259). As some of these are in planta pedis they should be dated to the first third of the 1st century A.D., to which period the London crater is probably attributable.
(4) HILARVS: HILAR, form not stated. Found in London (see Arch., XXVII, 152, and XXIV, 202). This stamp occurs in the Augustan period at Haltern, where it is classed amongst the older Arretine fabric. In the pre-Claudian period at Sels it is found in combination with the stamp of Ateius, i.e., CN.ATEIVS.HILARUS, ATEI HILARI (CIL. XIII, 3, 1; 49).
The stamp of Hilarus also occurs in combination with those of M. PERENNIVS (CIL. XV, 2; 5420), C. MEMMIVS (CIL. XI, 2; 382) and ANNIVS (CIL. XIII, 3, 1; 21), all of whom are well-known Arretine potters. Perhaps, his latest impression is that of HILAR FECT (CIL. XIII, 3, 1; 140), in the Rodez Museum. Probably more than one potter worked under this name.
Closely similar stamps occur in the Augustan period at Haltern SECV NDI, and at Rome SEC VNDI and SECV NDI (CIL. XV, 2; 5560b); also in the early period at Xanten (Bonner Jahrbüch., 116, 330, cup Loeschcke type 8b). No evidence of a Claudian provenance is forthcoming.
(6) ZOILVS: ZOII (the Z retrograde) on a cup Loeschcke type 11, viz., a Dragendorff form 27 with rouletted wall, Fig. 91, A.6. Found in Lombard Street; now in the Guildhall Museum. It has a truncated, flat rim, decorated with two applied dolphins. ZOILVS was a slave of ATEIVS (cf. CN. ATE ZOII (the Z retrograde) at Sels). The Z is frequently reversed by this potter, as at Haltern (Loeschcke, No. 114).
The chief points of chronological significance in this list of stamps are that those of Ateius, Cornelius, Hilarus, Secundus and Zoilus occur in the Augustan period at Haltern, that of Ateius/Xanthus at Mont Beuvray ante 5 B.C., those of Ateius, Hilarus and Zoilus at Sels in the pre-Claudian period and that of Ateius/Xanthus at Grimmlinghausen in the period Caligula-Claudius.
(7) Crater: Fig. 91, A.7; see also Brit. Mus. Cat., L159. Found in London (C. R. Smith, Journ. Brit. Arch. Assoc, IV, 16). The slightly everted rim is rouletted (cf. Oswald and Pryce, op. cit., II, 1,2, 4; XXVI, 1, 2), and the wall is divided into two zones by a rouletted moulding (cf. Oswald and Pryce, op. cit., II, 1, 2, 4). Beneath the moulding is an ovolo, surmounted by a row of large beads. Then follow the figure of a nude man with club in 1. hand and an indeterminate ornament. The same figure occurs on a fragment of Italic ware, from Arezzo, now in the Ashmolean Museum. Rouletting of the rim and the wall-moulding is characteristic of the decoration of many craters of the Augustan period.
(8) Crater: Brit. Mus. Cat., L161, Fig. 35. Found in London. The fluted and everted rim is defined below by a raised moulding. On the rim, Eros asleep. Decoration similarly situated is not infrequent in Arretine ware (G. H. Chase, Cat. of Arretine Pottery in the Fine Arts Museum, Boston, Figs. 140–142).
(9) Crater: Oswald and Pryce, type C (XXVI, 2), Fig. 91, A.9; Brit. Mus. Cat., L162. Found in London 1837, probably in Southwark. The rouletted moulding is succeeded by a bead-row, beneath which are depicted an eagle to r. and a floral and fruit ornament. Although indistinct, it is evident that this ornament is of a similar class to those used by CORNELIVS and PANTAGATVS (cf. Oswald and Pryce, op. cit., XXIV, 1, 2).
(10) Crater: Fig. 91, A.10. Found in London (Brit. Mus. Cat., M. 2363). Glaze, pitted externally, smooth internally; poor workmanship. The upper border of the design is composed of a series of concentric circles with two intervening beads. Closely similar upper borders are seen on two vessels by CORNELIVS (Brit. Mus. Cat., L55, 56). Then follows a double line, composed of repeated cuneiform leaves (rod-chain) and a row of beads. From this line depend festoons of the same type. On the wall are depicted a draped woman to front and a winged figure to r., both interrupting the middle of a festoon (cf. Chase, op. cit., Figs. 1, 69, M. PERENNIVS and TIGRANVS for similar interruptions of "rod-chain" festoons by figure-subjects). The design is closed by a "rod-chain" and a bead-row. Notwithstanding its poor workmanship the vessel is probably Italic. On the other hand it may, possibly, be a Belgic imitation of decorated Arretine, examples of which have been found in the Augustan pottery at Xanten (Bonner Jahrb., 122, PI. LIV, 3; J. Hagen, Einzelfunde von Vetera, 1910–12, PI. LIV, 1, 2, 3; LV, 3).
(12) Crater: Brit. Mus. Cat., L166, Fig. 36. Found in New Street, E.C. The raised moulding around the circumference of the bowl is a frequent feature of Italic craters. Above the moulding is seen a crater, closely similar in form to those on Arretine fragments in the British Museum (L98) and the Boston Collection (Chase, op. cit., Fig. 15). Probably Italic.
(13) and (14) Two plates with flat bases (Brit. Mus. Cat. L164–5), Figs. 92, A.13, 14, restored after Dragendorff's form 22. On the wall of one is a dog, on that of the other a dolphin. Similar figures in applied relief are not uncommon on Italic "plain" forms. An alternative restoration is that of a Tiberian plate with footstand, as at Aislingen (Oswald and Pryce, op. cit., XLII, 9).
(15) Fig. 92, A. 15. Found in London; now in the London Museum. The vessel has the inbent rim and internal groove of Loeschcke's type 4b. On the exterior wall is the spiral "handle" so commonly found on "plain" Arretine forms (cf. Loeschcke, Pl. X, 9, 12, 15a, b). Beneath the handle is a mask in applied relief.
(16) Fig. 92, A.16: London Museum, A. 17498. Found in Tooley Street, Southwark. Good, brownish-red, smooth glaze. The rouletted rim is followed by a narrow plain moulding. The wall is decorated with a mask and a wreath of repeated bifid leaves, in the form of a low-curved festoon. For closely similar masks on Arretine ware, see Chase, op. cit., Figs. 108, 142. The decorated wall is bordered below by a narrow plain band and a basal moulding which shows traces of rouletting. The rim is defined internally by a circular groove. (fn. 10)
An examination of these pieces demonstrates that they cannot be assigned to the early flourishing period of the Italic industry. Typologically, they belong to the first four decades of the 1st century of our era; perhaps the fragment, Brit. Mus. Cat., L166, may be even a little later.
In this connection it is interesting to note the evidence from dated sites in the north-western provinces. Thus at Haltern, in the Augustan period, Arretine ware is alone represented. At Sels, a site dating to the reign of Augustus and evacuated in that of Caligula, both Italian and Provincial sigillata are forthcoming; upwards of a hundred stamps of Arretine potters and many other Italics sherds have been recorded.
At Hofheim, which was occupied in the reign of Caligula (A.D. 40), no stamp of an Italian potter has been found, whilst this ware is only represented by three fragments of an Italic character. (fn. 11) At Aislingen, which was occupied from late in the reign of Tiberius, the Arretine ware is restricted to the stamp of L.GELLI, in planta pedis, and two other fragments. (fn. 12) In the legionary fortress at Neuss, which was first occupied c. A.D. 25, only a single example of Arretine ware has been found—a plate, stamped XANTHI in planta pedis (Bonn. Jahrb. 111/112, XXXVI, 24). At Grimmlinghausen, occupied in and after A.D. 40, Arretine ware is represented only by a single example stamped ATEIVS/XANTHVS (CIL. XIII, 10009, 55). In Britain the evidence is less clear. Occasionally the ware of a potter who was working as late as the Nero-Vespasian period found its way to this province, e.g., L.R.PIS(ANVS), found near Cambridge (CIL. VII, 894). At Richborough, on the other hand, the small area at present excavated, though certainly occupied in the Claudian era, has failed to yield a single fragment of Arretine. At Colchester, notwithstanding the presence of typologically pre-Claudian examples of Provincial fabric, Arretine ware is only represented by a solitary cup (Oswald and Pryce, op. cit., XL, 14). At Silchester, however, which is known to have been occupied before the Claudian period, probably as early as the last third of the 1st century B.C., (fn. 13) 32 pieces of Italic type have been recorded. On the balance of the evidence, it would seem probable that some of the 15 or 16 fragments from London found their way to the site previous to the conquest, thus indicating something of a pre-Claudian settlement of Italian and other traders. (fn. 14) Nevertheless, in view of the suddenness with which Claudian London sprang into prominence as a commercial centre, and the comparative abundance of Claudian pottery from the site, it is safer to leave open the possibility that the Arretine sherds may have been brought in on the first stream of trade after the conquest and before the complete extinction of traffic in this ware. The question whether these fragments of Italic ware reached London in the ordinary way of commerce and are consequently of pre-Claudian date, or whether they are of the nature of "survivals" brought over by the Claudian invaders or their immediate successors, does not therefore admit of a definite answer.
(II) Provincial Ware.
Space permits only of a brief consideration of the early provincial Terra Sigillata or Samian ware, found in London. The subject may be conveniently discussed under the two headings of potters' stamps and early decorated types, located examples of which are plotted on Map, Plate 64. (fn. 15)
(a) Stamps of Early Potters. (fn. 16)
These potters constitute more than 50 per cent, of those who are known to have commenced work in South Gaul in the reign of Tiberius, circa A.D. 20. Their decorated types are, almost wholly, those which are characteristic of the reigns of Tiberius, Caligula and Claudius. More particularly, the potters Amandus, Balbus, Bilicatus, Maccarus, Regenus and Salve worked, almost exclusively in the Claudian and preClaudian periods; the activity of the potters Scotius and Senicio may have continued, in some degree, into the reign of Nero, but their decorated work is almost entirely of a Claudian or pre-Claudian type (cf. Knorr, Töpfer und Fabriken verzierter Terra-Sigillata des ersten Jahrhunderts, 1919, Pls. 70–72, 75–78). The terminal date of the activity of the potters Ingenuus and Licinus cannot be so definitely assigned, for although the major part of their decorative work is typologically Claudian or pre-Claudian (cf. Knorr, 1919, op. cit., 40–42, 45–47), there is some evidence that they continued to work in the reign of Nero.
The potter Cocus made some very early and rare South Gaulish plain forms, e.g. Dragendorf 27 with rouletted wall (cf. Loeschcke's Arretine type 11) and Ritterling type 5 (cf. the Arretine cup by AMAR, Fig. 92, G. 1). His decoration is typologically Claudian rather than Neronian. Potitus was associated with the Tiberio-Claudian potter VOLVS, e.g., VOLVSII ET POTITIO at Vaison and VOLVS SII ET POTITIO at Nimes.
It should be noted that the stamps of the potters Amandus, Balbus, Bilicatus, Ingenuus, Maccarus, Regenus, Scottius and Senicio have been found in the pre-Claudian period at Sels, and that of CaligulaClaudius at Hofheim (A.D. 40–51 ) (fn. 17). Finally, it is pointed out that many potters (whose stamps have been found in London), whose work is largely characteristic of the Claudian period, have been excluded from this list because there is evidence that their activity was continued into the reign of Nero. Amongst them may be mentioned AQVITANVS, ARDACVS, BASSVS, GALLICANVS, LABIO, MASCLVS, and MVRRANVS.
In order to obtain an approximately accurate conspectus of the earliest Sigillata found in London the decorated types located on the Map (Plate 64), have been restricted to those which are characteristic of the reigns of Tiberius, Caligula and Claudius or such as have parallels in the Claudian period at Hofheim It is sometimes impossible to draw a hard and fast line between Claudian and Neronian decoration, but no ornamental piece has been accepted for which Claudian or pre-Claudian parallels are not forthcoming. On this account many examples of decorative types which fall into the period Claudius-Nero have not been plotted on the map.
(1) Crater: Found in Angel Court; now in the London Museum. The ovolo is full and its tongue is plain, both features usually met with in Arretine ware. The vessel is decorated with an upright plant ornament which terminates below in basal arcading, a class of ornament frequently met with in the earliest provincial sigillata (cf. Knorr, 1919, op. cit., 41 G, crater at Sels; 40a form 29, by INGENVVS, at Sels; Ritterling, Hofheim, I, op. cit., XXVII, 22, 23, both on forms 30). This type of decoration is an imitation of an Italic prototype (cf. Oswald and Pryce, op. cit., XXIV, 10, by NAEVIVS of Puteoli). The Gaulish crater is a direct imitation of an Augustan Italic prototype (cf. Oswald and Pryce, op. cit., II, 3, by ATEIVS). Five examples of this form have been recorded in Britain; two at Richborough, one at Tong, Kent, and two in London. Fourteen examples, dated to the reign of Tiberius, have been found at Bregenz (Knorr, 1919, op. cit., 1C, 2D, E, F, G, H, J, K; 3 M, N, O, P, 4 V). The Gaulish crater occurs at Sels in the period TiberiusCaligula (Knorr, 1919, op. cit., 41 G, J), and at Hofheim in the Claudian period (Ritterling, op. cit., Abb. 46, 1, 2; 47; 52, 4a and b). Fig. 92, G.1.
(2) Crater: Found in Cheapside, now in the Guildhall Museum. Ovolo the tongue of which has a central depression as on a crater at Hofheim I (Ritterling, op. cit., abb. 46, 1). Scroll decoration, very similar to that on Claudian forms 30 at Hofheim (Ritterling, op. cit., XXVI, 5; XXVII, 19). Fig. 92, G.2.
(3) Form 29: The distinguishing feature of this class of bowl is the central moulding which divides its external wall into two decorated friezes. Found in Leadenhall Street [London Museum, A21033]. Fig. 93, G.3.
Upper Frieze: Continuous scroll, the stalks of which terminate in sceptre-like leaves and rosettes. This scroll is an exact facsimile of that on a form 29, found at Claudian Hofheim, by the potter LICINVS (Ritterling, op. cit., XXIII, 2).
(a) The upright plant has lateral spirals ending in rosettes and terminates above in a sceptre-like leaf. Below, it forms a basal arcading, as in the crater, Fig. 92, G.1, and the lower frieze of a form 29 found at London Bridge Station, Fig. 93, G.6. The curve of the arcadingis continued upwards and terminates in an upright, pointed and serrated leaf.
In the field are two hares with striated rumps as used by BILICATVS (Knorr, 1919, op. cit., 14A); also a large rosette closely similar to that on a Tiberian form 29 at Sels (Knorr, 1919, op. cit., 91 C). The glaze, contour and decoration of the bowl are typologically Tiberio-Claudian. It is probably representative of the early work of LICINVS.
The rim is boldly out-curved, but materially differs from the high everted rim of Nero-Flavian examples of this form; it is probably copied from an Augustan Italic crater with a similarly curved rim (cf. Oswald and Pryce, op. cit., XXVI, 1 Haltern). The wall of the vessel has a rounded contour.
On the upper frieze is seen a scroll with sessile bifid leaves, and rosettes in the field. A closely similar scroll occurs on forms 29 of Tiberian type at Hod Hill (Brit. Mus. Cat., M208) and Silchester (May, XI, 2). This class of scroll with sessile bifid leaves is highly characteristic of the Tiberio-Claudian period and occurs in the early work of the potters ALBINVS, LICINVS and VRVOED and on forms 29, in the style of MACCARVS (Knorr, 1919, op. cit., 1B, 46D, Text Fig. 28, 88A, C). The central moulding is bordered by rows of large and well-spaced beads, in the Italian manner (cf. Fig. 93, G.7 and Knorr, 1919, op. cit., Text Fig. 4, M.PERENNIVS). The repeated gadroons of the lower frieze are full in character, as in early examples of this ornament.
(5) Form 29: Found on the National Safe Deposit Co's premises, Queen Victoria Street; now in the Guildhall Museum. Fig. 93, G.5. The fragment shows the rouletted central moulding frequently, but not invariably, found on Tiberian examples of this form. It is a copy of the rouletted moulding of Augustan Italic craters (cf. Oswald and Pryce, op. cit., II, 1, 2, 4; see also Fig. 93, G. 7). Occasionally this technique is found on sites which were first occupied in the reign of Caligula (cf. Ritterling, Hofheim, op. cit., XXV, 7a).
(6) Form 29: Found at London Bridge, Railway Approach; now in the London Museum. Fig. 93, G.6. Portion of the lower frieze showing an upright plant ornament with basal arcading. The chronological significance of this decoration has been referred to under No. 1.
(7) Form 29: Found in Gracechurch Street, west side. Now in the London Museum. The repeated sessile, lyre-shaped leaf of the upper frieze is derived from an Italian prototype (cf. Oswald and Pryce, op. cit., XXIV, 10, by NAEVIVS of Puteoli), and was used by some of the oldest South Gaulish potters (cf. Knorr, 1919, op. cit., 1A ALBINVS, 6B AMANDVS, 11A BALBVS, 14C BILICATVS, 51J MACCARVS, 77K SENICIO, 78A SENO, 80 VAPVSO). It occurs in the Claudian period at Hofheim (Ritterling, op. cit., XXV, 12b, XXVII, 21 and XXXII, 16, on a crater) but has not been recorded on subsequently occupied sites. Fig. 93, G.7.
(8) Form 29: Found at London Bridge, Railway Approach; now in the London Museum. Part of the lower frieze on which is depicted a large straight-wreath, composed of alternating sessile and stalked leaves. The stem of the wreath consists of repeated coalescent beads, as frequently occurring in the work of the older potters (cf. Knorr, 1919, op. cit., 1A ALBINVS). The bold character of the design places it amongst the work of the earliest potters (cf. Knorr, 1919, op. cit., 7Aa, Bb, Cc. 14B BILICATVS, 72H SCOTTIVS). Fig. 93, G.8.
(9) Form 30: the cylindrical bowl: found in Gracechurch Street, west side; now in the London Museum. The ovolo is surmounted by a series of "grouped" beads, as on an early form 30 at Hofheim I (Ritterling, op. cit., XXXII, 18). The wall is decorated with a scroll the stalks of which terminate in large leaves (occupying nearly the whole depth of the wall), spiral buds, seven-lobed buds and a tri-lobed leaf with three detached "berries." The bifurcations of the scroll are masked by bifid "tendril-unions" with basal beads, as on the Hofheim vessel referred to above.
The central element of the large leaf has a triangular termination, a feature which only occurs in the earliest examples of this leaf (cf. Knorr, 1919, op. cit., 2E. 3M, O. all on craters of Tiberian type of Bregenz; 21D; 32, a form 29 by FIRMO the elder; 41J crater at Sels). In later examples of this leaf the central element has a rounded termination. Large-leaf scrolls are characteristic of the decoration of some of the earliest examples of the cylindrical bowl (cf. Ritterling, op. cit. XXXII, 18, and pieces in the Museums of Cologne and Wiesbaden). Fig. 93, G.9.
Whilst it cannot be stated that none of the examples of Provincial sigillata, recorded on the map, was imported during the reign of Nero, the detail-evidence is sufficiently cumulative to warrant the conclusion that during the course of the Claudian period, London had already attained the position of a flourishing community. This conclusion is reinforced by the relative frequency of Italic ware found on the site, whether it be regarded as a "survival" or as suggestive of a pre-Claudian occupation.
(III) Distribution of Early Types.
(3) An area immediately to the W. of the Walbrook, (fn. 18) at the eastern termination of Queen Victoria Street.
Thus, the stamps of the potters MACCARVS and SENICIO have been found at Baltic House in the N.E., that of BALBVS on London Wall in the N., that of SALVE in Monkwell Street in the N.W., and that of MACCARVS in Paternoster Square, in the W.
From this wide-spread distribution of early types it may, with some probability, be inferred that at the date of Boudicca's rebellion (A.D. 60), the area within the enclosure of the later walls was largely occupied (see p. 31).