¹ Caputo, G., ‘Flavius Nepotianus’, Revue del Études Anciennes lii, 1951, p. 243Google Scholar.

² Published by Romanelli, P. in Leptis Magna, Rome, 1925, pp. 55–6Google Scholar.

³ Smyth's own account of his work at Lepcis was published by F. W. and Beechey, F. W. and Beechey, H. W. in Proceedings of the expedition to explore the northern coast of Africa, London, 1828, pp. 72–8Google Scholar. His plan of the ancient city appears as an inset to Admiralty Chart No. 247.

⁴ Cowper, H. S., Hill of the Graces, London, 1897, pp. 194–200Google Scholar.

⁵ Barth, H., Wanderungen durch die Küstenländer des Mittelmeeres i, Berlin, 1849, p. 313Google Scholar. The three transverse walls, built ‘in the true Phoenician style’, observed by Barth crossing the western mole of the harbour, are in fact part of the massive Severan harbour-works, and had no defensive function.

⁶ Nouvelles Archives des Missions Scientifiques x, 1902, p. 256 and pl. IGoogle Scholar.

⁷ Romanelli, op. cit. pp. 83–8.

⁸ Bartoccini, R., ‘Il recinto Giustinianeo di Leptis Magna’, Rivista della Tripolitania ii, 1925, pp. 63–72Google Scholar.

⁹ Guidi, G., in Africa Romana (Istituto di Studi Homani), Milan, 1935, pp. 240–1Google Scholar, and pl. XVI.

¹⁰ Notiziario Archeologico i, 1915, p. 64Google Scholar.

¹¹ Romanelli, op. cit. pp. 70–4.

¹² The antiquity of this bank has been questioned by Romanelli, (Archeologia Classica, iv, 1952, pp. 100–2Google Scholar), whose suggestion that it was thrown up by Italian soldiers in 1912 has been answered by the present writers (ibid. pp. 284–6).

¹³ IRT 357, of A.D. 119–120, found near the cisterns behind the Hadrianic baths. The baths themselves were not completed and dedicated until six or seven years later (IRT 361).

¹⁴ Photographs (now in the archives of the Department of Antiquities at Tripoli) taken during these excavations show that both outer and inner faces of the city wall were reasonably well coursed with large re-used blocks. A funerary inscription (IRT 692) of second-century or early third-century date was found built into the inner face.

¹⁵ Romanelli's description (Leptis Magna, pp. 83–5) was written when only the upper part of the arch protruded from the dunes.

¹⁶ JRS xxxviii, 1948, pl. VIII, 4Google Scholar (mistakenly attributed to the Severan Age; but see ibid. p. 160); Caputo, G., Archeologia Classica i, 1949, pp. 86 ff.Google Scholar; Fasti Archaeologici iii, 1948 (1950), 3485Google Scholar, fig. 82.

¹⁷ For the use of marble in Tripolitania, , see JRS xli, 1951, pp. 89–104Google Scholar.

¹⁸ Robertson, D. S., Handbook of Greek and Roman Architecture, Cambridge, 1945, p. 293Google Scholar.

¹⁹ Richmond, I. A., The City Wall of Imperial Rome, Oxford, 1930, p. 141Google Scholar.

²⁰ Below, p. 71.

²¹ This sector of badly laid, unpointed, and unsuitable material may represent a later repair, as it is distinctly inferior to the wall masonry between the West Gate and the sea.

²² This road was the ‘(via) ab oppido in mediterraneum’ laid out by the proconsul Aelius Lamia, c. A.D. 14–17 (IRT 930). Its importance was increased in the Severan period and later, when it became the main line of communication in rear of the limes.

²³ For the ancient topography of this extra-mural region, as revealed by recent air photographs, see Reports and Monographs of the Department of Antiquities in Tripolitania ii, 1949, p. 38 and plan 2Google Scholar.

²⁴ Gasr Shaddad, described by Romanelli (Leptis Magna, p. 163 and fig. 90) is the only relatively wellpreserved mausoleum in the eastern necropolis of the city within the great bank. In the same area there are the bases of numerous similar monuments, which were probably stripped of all usable materials in late antiquity.

²⁵ During the Vandal period, after the dismantlement of the earlier city wall, various structures in the main built-up area of the city seem to have been strengthened for purposes of local defence. One isolated defensive wall stands just outside the south wall of the Market, and numerous walls of similar character were found by Caputo during the excavation of the Theatre. The latter building, one of the most high-standing in the city, may have continued to serve as an extramural redoubt during the Byzantine occupation.

²⁶ This, and other Christian monuments of Lepcis, are described in Perkins-Goodchild, Ward, ‘The Christian Antiquities of Tripolitania’, Archaeologia xcv, 1953Google Scholar.

²⁷ For a fuller description, see Bartoccini, R., ‘Il recinto Giustinianeo di Leptis Magna’, Rivista delta Tripolitania ii, 1925, pp. 63–72Google Scholar.

²⁸ Krischen, F., Meyer-Plath, B., and Schneider, A. M., Die Landmauer von Konstantinopel, Berlin, 1938–1943, i, pl. 14–15Google Scholar; ii, pp. 65–9. Cf. Van Millingen, A., Byzantine Constantinople, London, 1899, p. 78Google Scholar. For Africa, cf. Gsell, , Monuments Antiques de l'Algerie, ii, 1901, pp. 344–84Google Scholar.

²⁹ Ward Perkins-Goodchild, op. cit. (n. 26).

³⁰ ibid.

³¹ In plan (fig. 7) this fragmentary Byzantine foundation might seem, together with the high-standing wall (B 5), to form a tower or gateway; but it must be noted that B 6 was dismantled or abandoned when B 5 was built, since there are no traces of its bonding with the exposed and well-preserved face of the latter. Recent excavations (1953) have exposed the inner face of B 5, confirming that B 6 represents the original project, never completed.

³² These late walls were first excavated by Professor Caputo before the war, and were re-examined during the British School's survey of the Severan Piazza in 1951. No dating evidence was found. A modern Decauville track covers the presumed course of the Byzantine wall in this area.

³³ First described by Romanelli, Leptis Magna, p. 150.

³⁴ An irregular breach in the masonry of the main wall behind the tower probably represents the site of an entrance doorway. A similar breach occurs in a corresponding position in Tower B 1.

³⁵ The plan (fig. 9 b) of this tower represents a level some two metres above the Byzantine ground surface.

^35a Excavations by the Italian Archaeological Mission (1952) have recently brought to light a tower midway between points B 13 and B 14.

³⁶ Ward Perkins–Goodchild, op. cit. (n. 26).

³⁷ It is uncertain whether the roughly built defensive wall at the seaward extremity of the eastern mole is earlier or later than the Byzantine circuit, as the point of junction is not visible.

³⁸ Aurigemma, (Africa Italiana iii, 1930, p. 84Google Scholar) refers to the quarrying of this ‘grosso muro in parallelepipedi di pietra, elevato con materiale raccogliticcio … a difesa del porto interno’; and a photograph in the archives of the Department of Antiquities at Tripoli shows the sappers at work.

³⁹ The north-west portico of the Colonnaded Street is not completely excavated at this point; it is possible that traces of the missing gate-tower may come to light later.

⁴⁰ The cutting was made by members of the 1950 Map of Roman Libya expedition, and Mr. John Spaul made the drawing on which fig. 10 is based.

⁴¹ It is curious that Wall B 5 escaped the dismantlement which the other sectors of the outer circuit suffered. Possibly dunesand had already begun to cover it before the new Byzantine defensive scheme was decided on.

⁴² It seems generally agreed (cf. Bury, J. B., History of the Later Roman Empire i, 1923, p. 255Google Scholar) that Tripolitania was retained by the Empire under the treaty of A.D. 442, and only passed into the hands of the Vandals at the death 01 Valentinian III in 455. Coherent municipal life came to an end, as inscriptions show, early in the fifth century.

⁴³ The remains of the walls of Punic Sabratha were identified by Miss K. M. Kenyon in 1948–9 on a line corresponding approximately to that of die north precinct wall of the East Forum Temple.

⁴⁴ Fasti Archaeologici iii, 1948 (1950), 3483Google Scholar.

⁴⁵ IRT 308, mentioning the third year of the proconsulship of C. Vibius Marsus. The monument itself remains unpublished.

⁴⁶ S.H.A., , vit. Sev. 18, 3Google Scholar: Tripolim unde oriundus erat contusis bellicosissimis gentibus securissimam reddidit. This statement does not necessarily imply that the ‘very war-like tribes’ in question (the Garamantes again?) had penetrated to the coast. It was the security of the Gebel that was restored by these campaigns, which resulted in the establishment of garrisons at Bu Ngem and Ghadames.

⁴⁷ Caputo, G., Rev. Ét. Anc. liii (1951), p. 244Google Scholar.

⁴⁸ IRT 880. Cf. Goodchild, and Perkins, Ward, JRS xxxix, 1949, pp. 91–2Google Scholar.

⁴⁹ Goodchild, , Reports and Monographs ii, 1949, p. 31Google Scholar, instancing IRT 284, 456 and 457. IRT 459 also refers.

⁵⁰ Manni, Eugenio, L'Impero di Gallieno, Rome, 1949, p. 58Google Scholar and n. 3.

⁵¹ IRT 460 (a).

⁵² Cf. n. 25.

⁵³ Goodchild, , in Fasti Archaeologici i, 1946 (1948), 2072Google Scholar.

⁵⁴ Mud-brick was more extensively used at Lepcis than is generally realised. Examples of its use at ground-level can still be seen in the Chalcidicum and near the Severan Arch.

⁵⁵ Cod. Iust. i, 27, 2, 1: Sancitimus itaque, ut dux limitis Tripolitanae provinciae in Leptimagnensi civitate sedes interim habeat.

⁵⁶ CIL viii, 101–2.

⁵⁷ Procopius, , bell. vand. ii 21, 3Google Scholar.

⁵⁸ S. Gsell and C. A. Joly, Khamissa, Mdaourouch, Announan, 1922, pp. 126–32.

⁵⁹ For the Arab sources relating to Lebda, see Romanelli, Leptis Magna, pp. 33–6. As Romanelli rightly points out (p. 34, note 2) Lebda controlled the coastal route to Tripoli.

⁶⁰ El-Bekri, , Description de l'Afrique septentrionale (ed. de Slane, , Alger, 1913), p. 26Google Scholar.

⁶¹ Romanelli (op. cit. p. 169) describes the fort on Ras el-Hammam as ‘sicuramente di costruzione bizantina’, and makes no reference to the Arabic inscription on the lintel of its inner doorway. Vida, Levi Delia (Annali Ist. Univ. Orientale di Napoli, n.s. iii, 1949, pp. 77–81Google Scholar) publishes the inscription without questioning Romanelli's dating of the fort itself. Yet there can be little doubt that the inscription refers to the initial construction of the fort which, in fact, has every appearance of being early Islamic rather than Byzantine.

⁶² Africa Italiana i, 1927, pp. 56–9Google Scholar. As Bartoccini remarks, the undermining of the Forum walls was not motivated by any desire to make use of their materials, and took place long before the large-scale exportation of marbles began. Possibly the aim was to destroy a notorious refuge of outlaws and bandits.

⁶³ The Italian defensive line on the south bank of the Wadi Lebda followed a course very different from that of the great earthen bank (see n. 12). Its line can still be traced to-day.