Copyright © The Malta Historical Society, 2005.

Source: Melita Historica. 10(1988)1(69-87)

[p.69] Two archaeological sites - Ras ir-RaĦeb, Malta, and Ras il-Wardija, Gozo

Mario Buhagiar

Ras ir-RaĦeb

At Ras ir-RaĦeb (MR 395737), also known as RAS IL-KNEJJES, a remote, windswept headland, in the NW corner of the Rabat-Dingli plateau, precariously close to the sheer cliff-lip, 45m above the sea, are the scant ruins of a building the nature of which has still to be satisfactorily determined. The archaeological importance of the place has been known since at least the late sixteenth century. Bosio [1] and Marc Antonio Axiaq [2] , basing themselves on Cicero’s account of the fanum Junonis, which had described its location as not far from the city of Melite (ab eo oppidum non longe), identified the ruins with those of the celebrated temple. Abela [3] did not agree and limited himself to a passing reference to rovine e anticaglie di fabriche. The two Semitic toponyms, Ras ir-Raħeb (Headland of the Hermit) and Ras il-Knejjes (Headland of the Churches) may indicate that the site had some religious significance in the late Middle Ages.

In 1922 the remains were brought to the notice of Themistocles Zammit by the proprietor of the land, the Conte Stagno-Palermo. A square pavement of “very regularly laid” small, baked tiles, had been laid bare by the farmer who tilled the shallow soil, and Zammit, who examined the place on 1 December, was able “to trace a frame of stonework”; the remains of a door-sill were clearly visible. A conspicuous scatter of household pottery made him suspect “a numerous settlement.” He, therefore, noted that “the site should be kept in mind as one that could be studied and excavated with profit.” [4]

The site was partially dug up in 1961-1962 by a team of officers from the Royal Navy under Capt. D. Scott and Sgn. Comdr. P. Pugh. The brief report in the MAR [5] is most unsatisfactory and the published plan (fig. 1) is inaccurate. The almost complete absence of contemporary deposits, due to erosion which had also eaten away the exposed masonry, made the interpretation of the archaeological material difficult.

Captain Scott tells the story in a type written account (National Museum of Archaeology, Valletta, Excavations at Ras ir-RaĦeb 1961-1962) that he addressed to David Trump who was at the time curator of Archaeology at the Valletta Museum. Work started in November 1961 and ended on 13 May 1962 when a party of naval divers explored the sea bed off the promontory, reaching a depth of over 60m and noting a few scattered stone blocks lying in the sand.

[p.70] The surviving remains suggest a building complex of some importance which commanded extensive sea views and probably faced W. It was surrounded, on at least three sides, by an esplanade that was bounded by a semi circular sweep of wall which extended from E. to W. and can be followed for a length of about 62m. [6] Only the foundation stones, of regularly squared limestone blocks, measuring on an average, c.1.54 x 0.56m, survive but it is possible to reconstruct the plan of a cluster of rectangular rooms grouped round a central courtyard that had a pavement of small white marble cubes embedded in concrete. A few diamond-shaped tiles survived in one of the rooms until 1977 [7] but I could find no trace of them in February 1988, when I inspected the site. Roughly in the centre of the main room, Scott noted two “curious well dressed oblong stones each 4 feet eight inches long by two feet six inches” which, he suggests could be some form of altar. He and his team were of the opinion that the place “was used as a sacred site from neolithic times to about the 4th century A.D.”

In the NE corner, is a long and narrow rectangular cistern. It has a circular orifice (outer dia. 55cm; inner dia. c.41cm) dug into a block of limestone, 1.30 x 1.00, that has an opening for rain-water at one end (pl. 1b). Zammit [8] moreover, drew attention to the fact that “although no water appears in the immediate neighbourhood, the well known and permanent spring of Għajn Bierda which flows a few hundred yards to the SE, may have once been led to the lower ground” and could, therefore, have been “fully utilized” when the building was in use.

Harrison Lewis’s suggestion that one of the rooms was a bath [9] is not supported by the surviving evidence. Equally conjectural is his claim that “a large rectangular stone, partly buried, with two square-holes cut into it” could have been the base for an oil press. There is nothing to indicate that the building was either a country retreat or part of an agricultural set-up.

The main interest of the building is the way in which it integrated within its structure two rough stone megaliths, each measuring c.3.00 x 2.44m, which are quite obviously the relics of an earlier, presumably, prehistoric structure (pl. 1a). The only other known instance, in Malta, of a such a deliberate re-utilization of the remains of a prehistoric building, in a late Roman architectural context, is at Tas-Silġ where the rear apse of a megalithic temple was incorporated in the sacred precincts of the Romano-Punic sanctuary. There are, moreover, a number of similarities in the layout of the two sites, such as, for example, the position of the courtyard in relation to the rest of the building. It is, therefore, possible that, as suggested by Scott, the Ras ir-Raħeb building may have likewise had a religious significance. A recent reassessment of Ptolemy’s Maltese co-ordinates [10] has tried to link this site with the temple of Heracles (Geog. 4.3.13) but the arguments are not conclusive. The interpretation of this intriguing site can only be aided by a new more scientifically organized archaeological investigation for which, in spite [p.71] of soil denudation and erosion, there is still much scope.

In a letter in the files of the National Museum of Archaeology, Valletta, dated 14 November 1962 and addressed to Capt. Scott who had meanwhile left the Island, David Trump describes how he had done his best to get Dr. B.S.J. Isserlin, who in 1962 was excavating at Motya, to inspect the site and suggest an interpretation, but the latter was unable to do so because of a family problem and pressure of work. Dr. Donald Harden was likewise busy. Trump, who at the time was taken up with field work at Skorba felt ‘simply at their mercy’. He had neither the time nor the expertise in Romano-Punic archaeology to give the site the attention it deserved. He did not, however, forget about it and eventually succeed in getting Dr. Isserlin to Malta and showing him around the place but he was of little help. Trump describes the visit in another letter dated 15 April 1963: “I showed him the site and the finds, and the general conclusion was that some of each might be Punic, none must be and much could not be...” Trump’s last hope was now the Missione Archeologica Italiana which was shortly to begin its campaigns at Tas-Silġ. Its work there would, he felt, throw light on the ‘Punic-Roman overlap period’ and might, therefore, help in the interpretation of the Ras ir-Raħeb remains. It is not known whether the Italian Mission was ever actually consulted about Ras ir-Raħeb; the site is not mentioned in their preliminary reports. As Trump suspected there seem to be certain similarities between the two sites and their associated finds but the matter should be investigated further.

A coin of Constantius II (AD 337-361) found during the 1962 dig seems to hint at a long life for the building and its possible utilization in early Christian times. The “great quantity of pottery” recovered from the deposit in the cistern represented, as a matter of fact, the whole spectrum of the Roman occupation. [11] Four other coins from the site were in an advanced stage of corrosion; two of them were classified “Siculo-Punic” though what this exactly means is not stated.

Of greater interest was an ivory plaque, 2.6 x 5.1cm, with a low relief of a crouching boar that is unclassical in inspiration and may, as suggested by the Museum Annual Report indicate a “late surviving Punic influence”. Two clay satyr masks on vessel legs (8.8cm and 5.7cm high, respectively) are likewise seemingly unclassical but the remains of a clay figurine of a nude male youth, broken from the waist upwards (ht. as preserved, 82cm) is probably a late Hellenistic work. A small grotesque head (ht.3cm) of a bald and bearded old man, usually claimed to represent a satyr, and a fragment of a cloaked figure (ht.8.8cm) are more pedestrian works. Several other pieces of clay figurines, including another male nude and a draped female holding a small pyxes, were collected from among the debris. If the building did actually have a sacred character these figurines could fit into the category of ex-voti. Like the site itself they demand expert study. The finds, according to Dr. Isserlin (as quoted by Trump in his letter of 15th August 1963) showed “more signs of Punicity” than the structural remains “but whether Punic in date, or only later suvivals of Punic type, he was not prepared to commit himself.” Isserlin was particularly interested in the ivory plaque and in the cloaked [p.72] figure which, he cautiously suggested, might represent Heracles in lion skin. If this interpretation is correct, this statuette would gain in significance especially in view of the recent arguments associating the site with Ptolemy’s Temple of Heracles.

Finally mention ought also to be made to a small quarry of unknown antiquity to the immediate S. of the site. It is possible that it provided the stone for the building of the complex.


The site of RAS IL-WARDIJA on the extreme SW tip of Gozo shares a lot in common with Ras ir-Raħeb. It is another lonely promontory remarkable for its wild beauty and spectacular sea and country views. The only access to it is by means of a narrow country lane that winds its way through terraced fields, up to a rugged hill of globigerina limestone on the top of which a few weathered blocks of ashlar mark the foundations of a British semaphore signal station. The remains (fig. 2) investigated by a team from the Italian Archaeological Mission to Malta, under the direction of Dr Catherina Caprino and Dr Anna Maria Tamassia, in the course of four campaigns, between 1964 and 1967, stretch at the foot of this hill and are reached from it down a steep path. The excavation of the site is minutely recorded in the lavishly illustrated Missione Reports for the years 1964 (pp. 167-176, figs. 11-13, pls. 73-83), 1965 (pp. 125-155, figs. 8-10, pl. 75-108), 1966 (pp. 81-111, figs. l1-14, 56-76), 1967 (pp. 87-94, fig. 9, pls. 38-43).

The excavation results pointed to a Punico-Hellenistic sacred compound with a rock-cut sanctuary. It is tempting to associate the remains with one of the four temples (among them one dedicated to Sadambaal and another to Astarte) mentioned in the well known, second century BC Punic inscription, reputedly found in Gozo around 1855. [12] Direct evidence is, however, lacking and the poverty of the finds, together with the absence of stratigraphied layers, makes an interpretation hazardous.

Early activity at the site is hinted by the presence of prehistoric cart ruts [13] which may suggest a Bronze Age fortified settlement of the Borg-in-Nadur type. [14] The strategically commanding situation of the conical hill does, in fact, recall such settlements. No Bronze Age material has, however, been reported. The earliest pottery found were, apparently third-second century BC, late Punic sherds. [15]

The rock-cut sanctuary (fig.2), which the Missione Reports describe as a nymphaeum, is entirely man made. It consists of a rectangular chamber, 4.67 x 5.80 x 2.10m, that faces+the sea and is dug on an E-W axis into the side of the hill. As a result of the very exposed location, walls and ceiling have suffered extensively from weathering and the floor was found concealed beneath a thick layer of fine [p.73] dust formed by the disintegration of the very friable-rock. [16] Its most interesting feature are five finely cut niches, ‘a-b-c-d-e’, that carry a moulded cornice (pls 2a-b). They are between 1.15 and 0.98m high, on the outside, and between 0.62 and 0.46m wide. Their depth varies from c.0.46 to 0.30m. The best preserved cornice, ‘b’, is 1.22m wide x 0.25m high. The cornice of ‘c’ may have been crowned by a pediment but the advanced state of corrosion makes it impossible to be certain about this. A grooved-out channel in the opposite walls of the niches may have served to carry a shelf that divided the inside space in two sections of approximately equal height. [17]

There are two niches, ‘a’ and ‘b’, in the N wall and two others, ‘c’ and ‘d’, in the E wall, but only one niche, ‘e’, in the S wall. In the back wall of ‘e’ was a deeply incised graffito of what looked like a figure with outstretched arms. It was variously interpreted as a crucifix, [18] a winged creature [19] and the sign of Tanit. [20] (pl. 4c)

Two cruder graffiti, both apparently representing very schematized winged figures are, respectively, lightly scratched in the space between niches ‘a’ and ‘b’, on the N wall and in the E corner of the S wall, outside niche ‘e’. The latter, which seems to represent a man with an outsized head is c.40cm high. The height of the other is only 19cm. These two graffiti may, of course, be quite modern and it is dangerous to try to read too much significance into them. The implied suggestion that they may be Punic because they “vaguely recall the Sign of Tanit” [21] is as improbable as the tale of supernatural happenings and sorcery told to the Italian archaeologial team by a local informer. [22]

The wall-space between Niches ‘a’ and ‘b’ contains also two incised crosses, the largest of which is 25cm high. Scratched crosses are commonly met with in Maltese caves and rock-cut chambers and they are often impossible to date. [23] Their presence here, together with the crucifix-like figure in niche ‘e’, suggested to Caterina Caprino that the place may, at a subsequent period in its history have been utilized as the retreat of a holy hermit. [24] This is not improbable because late medieval Malta abounded in cave-churches and dwellings [25] but direct evidence is lacking. In more recent times, the place may also have been used as an animalpen. [26] A number of holes are dug in the N and S walls, at a height of about 1.68 and 1.80m from the floor. Most of them are, as suggested, probably modern. [27] A [p.74] few (diam. c.9cm) can nonetheless be original and may have been intended to carry oil lamps. [28]

A wide rock-cut platform, c.0.42m high, runs round three sides of the chamber leaving only a narrow, trench-like corridor, ‘mm’, c. 1.00m wide in the middle. At the entrance, the corridor was, however, extended in a SW direction to include a small rectangular space, 1.10 x 1.05m, at the S. end of which was a smoothly cut ledge of rock, 1.05 x 0.42 x 0.24m, with carefully chamfered edges, that contained two conical holes, dia. c.0.26m x 0.19m deep (pl. 4b). Two similar structures were found in the sacred compound outside (infra). [29] A ritual function, possibly associated with the pouring of libations, seems likely. The Missione Reports call them “mense per le offerte”, or altar-stones for offerings. They are compared to the so-called libation holes usually found at the entrance of Malta’s prehistoric temples [30] but though they may, in fact, have served a similar purpose there is absolutely no evidence to support a continuity of liturgical traditions. [31] In the platform at the back of the ‘altar stone’ is a shallow rectangular cutting, ‘k’, 0.43 x 0.16m, which, it is suggested, may have served as the socket-hole for a stele. [32]

It seems probable that the entrance to the chamber was originally blocked by some sort of rock-cut screen or façade. This is indicated by the remains of a wall in the NW corner (pl. 2) It measures 0.91m long x 0.45m wide and is cut in a stepped formation having a maximum ht. of 0.99m and a minimum of 0.54m. Another length of wall, c.0.79m long x 0.60m wide x 0.62m high is preserved in the SW corner (p1. 4b). The rock at this end is roughly cut to form a short ramp that slopes sharply to the platform skirting the inside S. wall of the chamber and contains a couple of clumsily hewn out steps. The purpose of this odd arrangement is not known but the indications are that it may be quite modern.

Erosion may have been responsible for the destruction of the rock-cut screen that was, in time, substituted by a rubble-facade that sealed the entrance, except for a narrow doorway in the SW corner. Such rubble-built façades are characteristic of the late medieval and early modern troglodytic architecture of Malta and it may have coincided with the possible re-utilization of the chamber as a shrine or as a dwelling. At about the same time, corridor ‘mm’ was filled with sail and stone chips and covered with stone slabs to bring it level with the rock-cut platforms. In this way an even floor was provided for the chamber. [33] Recent disturbance was demonstrated by the find of a coin of George V in the packing. [34]

The area immediately outside the rock-cut chamber is neatly cut to form a NS passageway, ‘pp’, c.15.35cm long x c.1.00m wide, intersected almost in the middle [p.75] by a shorter passageway, ‘nn’, c.4.00m long x c.1.00m wide, aligned on the same WE axis as corridor ‘mm’ of which it is actually a continuation (fig. 2; p1. 3). At right angles to the SW corner of passageway ‘pp’, is a rectangular extension, ‘4’, c.1.35 x 1.00m, that contains a second ‘altar-stone’, 1.00 x 0.45 x 0.17m, with two cylindrical holes (diam. 0.26 x 0.17m deep). The ledges of rock that border the sides of passage way ‘pp’ have been cut to form continuous stretches of benches which have a prominently raised ridge running along their back sides (pl. 3). Great pains seem to have been taken to smoothen the rock. Where levelling was difficult, a fine grained white mortar was used. Patches of this mortar, noted in 1965, were thought to be Roman rather than Punic. [35]

At point ‘h’, at the W. end of the EW. stretch of bench, close to the ‘altar-stone’ recess, (fig. 2) is a graffito showing an X-sign which Giovanni Garbini [36] considered a Punic religious symbol. He drew attention to its similarity to a much larger symbol engraved on the most prominent part of the ledge of rock that surrounds the tophet of S. Antioco, near Sulci in Sardinia, [37] and to other possible analogies such as the signs scratched on the backs of clay female figurines from the island of Ibizia, in the Balearic Archepelago, and on Stelae from Nora in Sardinia. Of greater significance is a Phoenician seal of unknown provenance, published by Lidzbarki [38] which shows an almost identical symbol beneath the left hand of a four-wing deity who wears the double Egyptian crown and holds the ankh in his right hand. In Malta, Garbini had already noted the symbol engraved on a stone slab at Tas-Silġ and on a number of pottery sherds from the same site and from San Pawl Milqi. The most interesting was the fragment (6cm long) of the inside rim of a large plate from Tas-Silġ on which the symbol (ht. 3cm) appeared painted in red together with the letters תם which may stand for an abbreviated votive formula that still awaits decipherment. [39] Garbini concludes that, while the true significance of the symbol remains unknown, an eastern origin seems indicated and that in Malta, where it seems to have been associated with the cult of Astarte/Tanit, it retained its religious currency longer than at Sulci and other places under Punic cultural influence.

Another graffito, at point ‘f’, of the adjoining NS bench (fig. 2), has the appearance of an inverted Latin-cross. The Missione 1965 Report (p.131) cautiously suggests another Punic symbol but it is equally possible a Christian sign scratched by a late visitor to the site before the rock-benches were covered with field-soil.

To the SW of the area just described is a nearly square water-tank, ‘6’, c.3.67 x 3.60m, that is nearly 3m deep and had a water capacity of c.37.5 cubic metres (fig. 4). The bottom was reached down a precipitous flight of ten badly eroded rockcut steps hewn along the W wall for a length of 2.20m. There are no indications for stuccoing. The internal dimensions of the tank were at some time enlarged by the digging of deep recesses into the S and E walls leaving an overhanging ledge [p.76] of rock above these two sides. A neatly cut rebate, varying in width from 0.36 to 0.19 and c. 0.18m high, is cut round the edges of the four sides at the opening. Its purpose is unknown unless it was to support some sort of roofing arrangement, possibly made up of slabs of globigerina limestone carried on the backs of wooden beams. A water-channel, 3.05m long x 0.09m wide x c.0.33m deep, dug in the rock above the W wall, probably fed the tank with rain water. It was found partially covered with rough copping-stones. [40] Cisterns of this type have often been reported from Maltese Romano-Punic sites but the Missione Reports suggested instead a sacred pool for ritual bathing and drew a comparison with the water-tank within the precincts of the temple of Saturn-Baal and Tanit Caelestis at Siagu in Tunisia. [41]

A small bell-shaped cistern, ‘5’, is situated a few metres away to the NW on the gentle slope beneath the rock-cut chamber (fig. 5). It has a rectangular orifice, 1.52 x 0.67 x 0.42m, enclosed within a rock-cut well-head that is c.0.30m high. It is c.3.00m deep and has maximum measurements of 3.80 x 3.00m. The water capacity is estimated at 14.50 cubic metres but there are indications that it was originally substantially smaller. [42] A series of cuttings in the S Wall may have been footholes. There is no evidence for stuccoing.

At the lowest part of the promontory close to the edge of the precipitous cliffs, the scanty remains of a quadrangular building were discovered in 1964 [43] but no proper excavations took place before 1966. [44] The building was constructed of roughly squared blocks of globigerina limestone and had an EW axis (fig. 3). Only two lengths of wall at right angles to each other had survived but there was evidence for a fire and for at least a partial rebuilding in late Roman times. The longer wall, which must have been at least 11.20m long, contained a threshold 2.10 x 0.64 x 0.27m with pivot-holes for a door; a longish slab, 1.45 x 0.67 x 0.28m may have been the lintle. [45] The other wall could be traced for c.2.80m but only three stones were found in situ.

The floor inside the building may actually have been the rough bed-rock. There was nothing to suggest any other type of pavement. Nor did the excavations produce evidence for a roof and the Missione Reports suggest a temple open to the sky. The outside walls were covered with a fine grained white plaster made of marble dust while the interior may have been stuccoed with a coarse cement mixture of small marble chips and mortar. Fragments of red intonaco and a large piece of brownish-violet plaster seem to suggest some sort of painted decoration.

On the outside of the building, facing the threshold at a distance of c.2.20m, was another altar-stone with offering holes c.20cm in diam. x c.12cm deep. It rested directly on the bed rock and was placed at the entrance of an enclosure, constructed of rubble stones, that skirted the walls of the temple. The Missione Reports, [p.77] therefore, suggest that it also served the purpose of a threshold that the faithful had to cross when they entered the temple compound.

The pottery finds from this site consisted mostly of small sherds which were generally of a typeless variety that is hard to classify. [46] Most pieces were seemingly modern. [47] The more datable seemed to fit a time-range between about the third century BC and the first two centuries AD [48] and consisted mostly of domestic wares. The more interesting material included a few pieces of the Romano-Punic pink pottery decorated with narrow red bands that is characteristic of Malta [49] and some fragments of Punic-type bilychins lamps. [50]

Finer wares were represented by a part of a red clay eonochoe with a trilobate mouth [51] and by a number of Aretino-type sherds. [52] Two terra sigillata pieces seemed to belong to a variety produced in North Africa in the course of the fourth century AD. [53] One of the pieces, 4 x 3cm, which is part of a globular vase, is decorated with a motif that seems to be made up of vine tendrils. The other, 4 x 3.5cm, is a fragment of the base of a circular dish and carries an impressed decoration of concentric circles and a palm-frond. Three other terra sigillata sherds (the largest is 2.8 x 1.8cm) had a similar red slip and seem to have been parts of plates or circular dishes. [54] A sherd, 3 x 2cm, of a greyish-yellow paste with a black slip is vaguely called Hellenistic [55] while a fragment of a red clay disc lamp, decorated with two parallel grooved lines, was possibly Roman Imperial period and, perhaps, North African. [56] More significant is part of the mouth of an amphora that bears the potter’s stamp AKLI which is apparently unknown. [57]

All pottery was found out of context and may have reached the site as a result of modern dumpings. The four small coins listed among the finds are a further indication of recent disturbance. One (diam. 1.8cm) was too badly corroded to allow identification [58] while another, in a slightly better state was, possibly, a Sicilian-Aragonese denarius because one of the faces seemed to carry an eagle with a profile head. [59] The two other coins [60] were, respectively a mezzo tari piece of Grand Master Perellos (1697-1720) and an unspecified coin of George V (supra).










[1] G. Bosio, Historia della Sacra Religione et Illustrissima Militia di San Giovanni Gierosolimitano, di nuovo ristampata e dal medesimo autore ampliata et illustrata. (Rome, 1621-29), Part iii, 924.

[2] M.A. Axiaq, Relazione della nuova e grandissima devotione introdotta nella S. Grotta di S. Paolo nell’Isola di Malta (1610). Notitie di Malta. N L M Lib. 515, ff.21v-22.

[3] G.F. Abela, Della Descrittione di Malta (Malta 1647), 67.

[4] Museum Annual Report (Malta) 1922-3, v.

[5] Museum Annual Report (Malta) 1962. 6.

[6] The esplanade and its perimeter wall are neither mentioned in the Museum Annual Report account, nor are they marked on the excavations plan. There is however a reference to them in the typewritten report Excavations at Ras ir-Raħeb, at the National Museum of Archaeology, Valletta.

[7] H. Lewis, Ancient Malta - A Study of its Antiquities. (Bucks 1977), 92.

[8] Museum Annual Report (Malta) 1922-3, v.

[9] H. Lewis, 92

[10] F. Ventura. “Ptolemy’s Maltese Co-ordinates”, Hyphen (Malta) vol v no. 6 (1988), 253-69.

[11] Museum Annual Report (Malta) 1962, 6.

[12] CIS I, 132; M. Guzzo-Amadasi, Le Iscrizioni Fenicie e Puniche delle Colonie in Occidente, (Rome 1967), 23-5.

[13] Missione 1964, 167.

[14] On cart-ruts and fortified Bronze-Age settlements in the Maltese Islands, see J.D. Evans, The Prehistoric Antiquities of the Maltese Islands (London 1971), 200-4.

[15] Missione 1965, 138-49; 1966, 107-111.

[16] Missione 1964, 170.

[17] Strangely enough the Missione reports make no reference to these grooved-out channels.

[18] Missione 1964, 169.

[19] Missione 1965, 127.

[20] Missione 1966, 81. This graffito was removed from the wall in 1988 by someone, apparently, using a reumatic drill and it now, presumably, forms part of a private collection.

[21] Missione 1966, 81.

[22] Missione 1965, 126-7.

[23] M. Buhagiar, Late Roman and Byzantine Catacombs and Related Burial Places in the Maltese Islands (Oxford 1986), 355.

[24] Missione 1964, 169.

[25] M. Buhagiar, “Medieval Cave-Dwellings and Rock-Cut Churches in Malta”, Atrium, 3 (Malta 1983), 17-22; id., “Medieval Malta: Its Hypogea, Cave-Churches and Ecclesiastical Buildings”, Architecture in Malta I: Historical Aspects, SACES (Malta 1986), 39-49.

[26] Missione 1964, 170.

[27] Missione 1965, 126.

[28] This possibility seems to have been missed by the Italian archaeological team.

[29] These ritual stones are today unaccounted for. They may have been stolen at the same time as the graffito in niche ‘5’.

[30] Missione 1966, 101.

[31] Missione 1966, 101, also draws attention to the similarity of these tables to certain archaic and classical Greek monumental altars of the type found in Cyrene and illustrated in S. Stucchi, L’agorà di Cirene, vol. i (Rome 1965), 97 et seq., pl. XV3.

[32] Missione 1965, 129.

[33] Missione 1965, 125-6.

[34] Missione 1964, 171, 173.

[35] Missione 1965 , 132.

[36] Missione 1965 , 153-5.

[37] On the S. Antioco tophet, G. Pesce, Sardegna Punica (Cagliari 1961).

[38] M. Lidzbarski, Ephemeris für semitiche Epigraphik, vol. i (Giessen 1900), 137-9.

[39] Missione 1965, 59. See also M. Guzzo-Amadasi, op.ct. 41-2.

[40] Missione 1965, 137.

[41] Missione 1965 , 133-7; 1966, 83-5. On the water-tank in the temple of Tanit Caelestis at Siagu, see A. Merlin, “Le sanctuaire de Baal et de Tanit près de Siagu”, Notes et Documents, iv (1910), 24, 38n.

[42] Missione 1966, 82-3.

[43] Missione 1964 , 173-5.

[44] Missione 1966 , 94-111.

[45] Missione 1964, 175.

[46] Missione 1964, 175; 1965, 138.

[47] Missione 1965, 138.

[48] Missione 1965 , 142; 1966, 140.

[49] Missione 1965, 149; 1966, 109.

[50] Missione 1964, 175; 1966, 108.

[51] Missione 1965 , 138.

[52] Missione 1964 , 173.

[53] Missione 1966, 92-3.

[54] Missione 1965, 147.

[55] Missione 1965, 148.

[56] Missione 1966, 108.

[57] Missione 1966, 107-8.

[58] Missione 1965, 149.

[59] Missione 1966, 107. The identification of this coin is due to Capt. C.G. Zammit who pointed out its similarity to the denarii minted during the reign of James I of Sicily (1285-1296).

[60] Missione 1964, 171, 173.