Source: Melita Historica. 9(1984)1(1-18)
[p.1]The Salina Hypogea at St. Paul's Bay
Salina Bay, which gets its name from the salt-pans built by the Knights in the 17th century, has considerable archaeological significance and there is evidence that it was a port, of some importance in Roman times.  Earlier activity is indicated by the presumably Bronze-Age cart-ruts  that can be followed for a considerable distance round the contours of the bay. Its most important antiquities are, however, the clusters of Paleochristian hypogea which because of their carved decorations, seem to imply a degree of local wealth. The site first made archaeological history in 1721 when, according to G.A. Ciantar  :
... a gold coin ... with the stamp of the Emperor Phocas was found in a field above the salt-pans. It was therefore guessed, that other valuable things might be discovered and after a diligent search a tomb was found with many skeletons piled haphazardly one on top of the other. There was also found a niche or an arch of ordinary stone with one skeleton, and a similar one dug in the rock with another skeleton, both of which were covered with stone slabs from which it seems that they were tombs of persons of distinction...
This may refer to one of the hypogea about to be discussed and a tomb that seems to fit the description was in fact excavated in 1937.
The necropolis is situated in a place known as Ta’ Latnija and it is reached by following a path across the fields to the south-east of the church of the Annunciation of the Virgin (fig. 1a). It consists of five small catacombs, about a score of floor-tombs and several rock-tombs. The catacombs and the rock-tombs are here discussed under the generic name of ‘hypogea’ and each is given a progressive number. The rock-tombs are mostly filled with stones and field soil and are therefore, inaccessible; a number were, apparently, destroyed during the construction of houses at the southern end of the site in the 1960’s and two were reputedly cemented over in the farmyard at the back of the church.  The entrances to the hypogea are tunnelled at different levels within a low hill that slopes gently to the sea. There is reason to believe that the natural terracing was, in post-Roman times, modified by quarrying possibly to determine field [p.2] boundaries.  The site suffered considerable damage as a result and a number of tombs were either destroyed or mutilated.
The tops of the entrances of at least some of the hypogea must always have been visible above the field soil and it is therefore, strange that they apparently, passed unnoticed until 1937 when they were discovered by C.G. Zammit, the Curator of Archaeology at the Valletta Museum.  According to one account  in 1937, Andrew Sammut, a local farmer
... realising that the catacombs were worthy of examination, personally cleaned the mud from them and was able to supply tangible evidence to the Museum at Valletta
If this is indeed true, valuable archaeological information must have been lost before Zammit started his survey in May. This was soon afterwards taken over by L.J. Upton Way who had previously done work for the museum. Upton Way tells the story in an article he wrote for the Times of Malta on 29th June 1937. He had arrived in Malta from Palestine after a serious car accident and was immediately persuaded by Charles Rizzo, the Director of Museum, to take charge of the excavations as soon as he was well enough. He started work about the second week of June and after a forthnight had cleaned two small catacombs in addition to the one that had already been investigated by Zammit. His newspaper article is the only excavation report. The account in the Museum Annual Report  refers the reader to it and describes in lesser detail the most important catacomb. It mentions, nonetheless, five instead of three catacombs, two new catacombs having apparently been investigated after Upton Way published his article.  The site was then abandoned and suffered considerable damage: the rock-tombs in particular were opened by inexperienced hands and their contents, if any, dispersed. The Archaeological Section of the Valletta Museum had intended to put gates in the entrances of the more important catacombs  but this was not done and in 1947 the Rev. J. Farrugia protested in the Times of Malta that the largest catacomb was badly mutilated and left in a ruined condition. The Museum Department was at that time undergoing re-organisation after the disruption of World War II and the site could not be attended to because of other pressing work. In 1963 a number of letters and reports in the local press once again urged the government to take steps to preserve the hypogea from further “profanation” and in January of the following year the Rev. George Serracino Inglott contributed a detailed report to the Sunday Times of Malta.  The most noteworthy [p.3] catacomb was, as a result, fitted with a wooden gate but the rest of the cluster is still unprotected. In 1979 the site received the attention of George H. Musgrave who was studying St. Paul’s shipwreck in Malta. His conclusions are not always acceptable but this account was the first attempt at a systematic analysis. 
Description of Remains (fig. 1b)
The most conspicuous part of the necropolis is an apparently man-made rectangular opening, c. 8 x 4m, cut into the upper part of the hill. It contains the entrances to Hypogea ‘1-2-3-4.’ Upton Way calls it a small amphitheatre-shaped bay in the rock and it is possible that it was intended as a place of congregation for a liturgy of burial or commemoration of the dead. The regularity of the cutting with neat right angled corners suggests that it was excavated according to a plan and Musgrave is probably quite right in rejecting the theory of a disused quarry adapted into a cemetery.  One should perhaps interpret it as a development of the narrow open courts in front of the hillside hypogea at Bingemma and elsewhere.  There may have been a wall across its front with some sort of doorway in the middle but the vegetation and accumulated debris on its floor and the subsequent quarrying make it impossible to reconstruct its original appearance. The mutilation is most noticeable in the right hand side where a number of tombs above the entrance to Hypogeum 1 have been destroyed. The walls were clearly somewhat higher and the possibility of a roof cannot be excluded.  It is, however, unlikely that the whole space was roofed over and one ought, perhaps, to think in terms of a projecting ledge or rock canopy (possibly supported by pilasters) running along the sides of the ‘rectangle’ and forming a sort of cloistered walk in front of the entrances, to the catacombs. A hypothetical analogy would in this way be created with the main catacomb at Abbatija tad-Dejr, in Rabat, which was entered through arched openings at the back of a narrow atrium. The ‘rectangle’ does indeed pose intriguing problems which can only be solved by a proper archaeological dig. According to Upton Way it was found filled with material and only “the top of the entrances” of the several hypogea “appeared above the soil.” The debris was, apparently, indiscriminately removed and though the original floor-level was reached no record of finds seems to have been kept. Its important features are now once again hidden by modern rubbish but there are indications for at least two floor-tombs and for a rock-bench in the north-east corner. Outside the perimeter are two parallel channels in the rock which may either be associated with a drainage system or, perhaps, with some form of entrance arrangement.
[p.4] Several shallow floor-tombs with a rock-pillow and a single head-rest.  are noticeable in the area outside the ‘rectangle.’ In 1964 Serracino Inglott counted twenty-five such tombs most of which are covered with soil when the fields are under cultivation.  He wrongly refers to them as “locular-graves” and is also mistaken in thinking they are pre-Christian. They are in fact, apparently, a late form of Maltese Paleochristian tomb. 
Hypogeum 1 (fig. 2)
This small catacomb tunnelled into the north-east wall of the ‘rectangle’ is the largest in the necropolis. It faces north-west and has a spacious roughly rectangular plan. It is reached down a few steps through a square-headed opening deeply recessed in the hillside but it is clear that the original entrance arrangement has been considerably disturbed. The rectangular socket-hole in the threshold suggests that it was closed by one of the usual stone pivot-doors.  Immediately inside the entrance to the left is a well preserved agape-table with a triclinium 44cm high.  The curved wall at its back contains a bi-partite window tomb, ‘2’, and a roughly cut, arched depression, ‘1’, which, probably, marks the start of another tomb. 
The central tomb of the hypogeum is the large baldacchino, ‘9’, facing the entrance, at the far end, which is detached on three sides and has a deep trough-like grave with two head-rests. On the right hand pilaster of the front facade, is a deeply incised motif, 18cm high, that looks like a one-armed cross (fig 2b). The remaining tombs consist of four arcosolia, ‘3-4-5-6’, and two other partially detached baldacchini, ‘7-8’. Arcosolia ‘3-4’, in the south-east wall, communicated with one another and with the agape-exedra  through narrow arched windows. That opening on the exedra is cut in a neat rectangular frame (fig 2c) which suggests that it may have been closed by a stone plug. Arcosolium ‘4’, which has two bi-partite troughs, is noteworthy for the two pilasters, 32cm high, with rudimentary base and capital, carved in very low relief on either side of the arched opening (fig. 2d). Arcosolia ‘5-6’, in the south-west wall, are also linked by a small arched window. These two tombs are in a bad state of preservation and ‘6’ is mutilated. Baldacchino-tombs ‘7-8’, in the north-west corner of the same wall, facing the agape-table, are even more mutilated and their arched canopies are almost entirely destroyed.
[p.5] Upton Way does not report on this catacomb and it is not known in what condition it was found. There is, however, reason to believe that some at least of the mutilation has been caused by modern vandalism. Outside the catacomb two floor tombs are dug into the raised ledge of rock on either side of the entrance and, at a higher level, in the corner between it and the opening to Catacomb 2, is a large loculi with a wide rebate for the sealing-slab cut all round it.
Hypogeum 2 (fig. 3a)
This is a smaller catacomb with an L-shaped plan formed by the junction of two short corridors, ‘aa’ and ‘bb’. It is reached down three steps from an arched window-opening, 58 x 82 x 15cm, which was probably closed by a stone pivot-door for socket-holes are noticeable just inside the entrance. Corridor ‘aa’, 0.67m wide x 1.68m high, contains a table-tomb, ‘1’, and an arcosolium ‘2’, which face each other on opposite walls. Both are bi-partite. The arcosolium is well preserved and has a flattened arch, 81 cm high. Inside it are two pyramidal lampholes. Similar lampholes are also noticeable above one of the head-rests of the table-tomb which is detached on two sides.
Corridor ‘bb’ is more spacious than ‘aa’ and has a maximum width of 1.39m. At its north-east end a mutilated arched doorway leads to a small box-like space, ‘c’ which probably marked the start of a projected extension. It has two noteworthy tombs, an undetached baldaccino ‘5’ and an Arcosolium ‘3’. Baldacchino, ‘5’', cut parallel to table-tomb ‘1’, to which it is connected by a narrow arched opening, has a large trough, 46cm deep, with, two head-rests. Arcosolium ‘3’ is cut along the same wall as Arcosolium ‘2’ and the two arcosolia are connected by another of the small arched openings that are one of the characteristics of the Salina hypogea. The platform of ‘3’ is 83cm high and is recessed in the usual flat arched niche 79cm high, in the back wall of which is a child-loculi ‘4’ containing an arched lamp-hole. Dug in the platform is a single burial-trougn with one head-rest and a narrow trench, 1.04 x 0.23m, which possibly served as an ossuary.
In the back wall of ‘bb’ facing the entrance, are two adult-loculi, ‘6-7’, ‘7’, 1.20m above ground, measures 1.89 x 0.41 x 0.43m, and has a rebate for the sealing-slab cut round three sides. Beneath it is a narrow cutting, 50 x 12 x 7cm deep, which may have been intended for a thin marble slab with an inscription.  Patches of a smooth intonaco with daubs of red pigment suggest that the wall beneath it may have been painted. Loculi ‘6’ has almost identical proportion and beneath it is another small cutting which may likewise have served as a bed for a marble inscription.
[p.6] This small hypogeum was one of the three described by Upton Way according to whose report it
... was crammed to the roof with skeletons, piled one upon the other, the lowest resting upon the lids of the original graves, the lids of three of which have never been lifted. These are to remain unopened ... there is no carving, but fixed by cement to the wall opposite the entrance is a nicely cut stone slab which appears to be a later form of the table of the feast ...
The piling of skeletons seems to fit the description of the catacomb discovered in 1721  but I think it would be wrong to assume that it was the same hypogeum. In an age, like the early 18th century, when the cult of antiquity was becoming a fashion it would have been, to say the least, strange for a tomb to be sealed up again exactly as found without its contents being explored and dispersed. There were probably, therefore, at Salina two hypogea which at an unknown period were used as ossuaries. Upton Way suggests that this may have happened
... during an epidemic or when the saracens overran the island ... but only a study of the bones (if they can still be identified in the stores of the Valletta Museum) can help solve the mystery.
The hypogeum has suffered some mutilation since 1937. The graves have all lost their lids and, more regretably, the “nicely cut stone slab” opposite the entrance has disappeared. The laconic description points to a square agape-table carved out of an independent block of stone similar to the one subsequently discovered in the Taċ-Ċagħqi Secondary School Hypogea.  Significantly it seems to have stood in the most spacious part of the hypogeum, against the painted wall beneath tomb ‘7’. This presumed table is the only appartently Christian feature of the hypogeum but on the wall outside the entrance to the left is a deeply incised Latin cross of questionable antiquity.
Finds from the hypogeum included: “Four lamps and jug of human bones”.  The “jug” (sic) poses an intriguing problem. There is no reference to cremated remains which would have been incongruous in a presumably Christian hypogeum. The “jug” might, therefore, have contained small bone splinters and served the purpose of an ossuary.
[p.7] Hypogeum 3 (fig. 3b)
This is the smallest catacomb in the ‘rectangle’ and consists of a rectangular room, c. 1.69m high, with a large baldacchino ‘3’, in the middle. It was also excavated by Upton Way who reported that
... it produced a couple of lamps, a good many bones and a small amount of pottery fragments. All the graves have been broken open ...
it is cut in hard limestone which is difficult to work and the walls are very irregularly shaped. Entrance is down two steps through a large rectangluar opening, 1.29 x 0.72 x 0.26m, which was possibly closed by a wooden door that turned on iron pins set in the small circular socket-holes, c. 3cm in diameter, that are noticeable in the threshold and in the ceiling at the back of the lintel. Immediately inside the entrance against the right wall is a stone bench which might have served as a shelf for the tomb-pottery. The baldacchino-tomb is unfortunately mutilated and its front side has suffered considerable destruction. It has unusually thick walls and contains a deep burial-trough with two head-rests. Since it is free-standing it divides the internal space into four corridors to which it is linked by three round arches. The hypogeum contains three other tombs one of which, ‘4’, is unfinished. This was intended either as an adult-loculi or, perhaps, as an arcosolium to complement the arcosolium-tomb in the wall opposite. The latter tomb, ‘1’, is mutilated. Of greater interest is a finely finished bi-partite window-tomb ‘2’. This is cut at the back of a deep arched recess with an unusually elongated pyramidal lamp-hole, 15 x 32cm, in the left wall (fig. 3c). This tomb has plain pilasters 28cm high, at the entrance and, above the lintel of its rectangular window-door, there are two neatly drilled holes (fig. 3c) the exact purpose of which is unknown. Possibly they helped secure the sealing-slab in position.  Before leaving the hypogeum one may also mention a rectangular opening, ‘5’, that cuts through its front side. This is probably a modern addition.
Hypogeum 4 (fig. 4)
Of the catacombs of the ‘rectangle’, this is the one that has suffered most mutilation and (on account of the unsuitability of the hard limestone) it is also the least well-cut. As in the case of Hypogeum 1, it was investigated by Upton Way after the publication of the Times of Malta article; no account of the excavation seems likewise to exist. It is entered down a couple of steps through a small arched doorway, 84 x 64 x 15cm, and has a similar plan to Hypogeum 2, consisting of two short corridors ‘aa’ and ‘bb’ aligned at right angles. Corridor ‘aa’, 65cm wide x 1.58m high, is flanked on either side by two deep arcosolia, ‘1-2’, with three burial-troughs [p.8] with two head-rests each. A large part of the roof of ‘1’ has caved in and the troughs, are therefore, full of deciris. Arcosolium ‘2’ is better preserved and the flattened arch above is well proportioned. Each of its three troughs has, besides, a large arched opening in corridor ‘bb’ but this part has suffered considerable destruction. The middle trough has a large arched pottery shelf in the wall above the head-rests while the inner trough is segregated from the rest by another flattened arch which makes it look like a seperate canopied tomb. In the north-west corner of ‘aa’ is another arcosolium ‘3’ with a bipartite-trough. In the wall above its two head-rests is a small arched opening to arcosolium ‘1’. Corridor ‘bb’ has a maximum width of 102cm and a minimum of 69cm. It is very unevenly cut and contains a mutilated window-tomb with one head-rest, ‘4’. Outside the entrance to the hypogeum, to the left, is a well cut rock-bench with a loculi, ‘5’, above it.
Hypogeum 5 (fig. 5)
This is the finest catacomb at Salina and one of the most impressive paleochristian monuments in Malta. It is tunnelled into the hillside a few metres to the north of the ‘rectangle’ and in front of it is a narrow court reached down several steps. It was investigated by C.G. Zammit between May and June 1937 when Upton Way was put in charge of the work. It was found completely despoiled, except for two lamps and “a considerable quantity” of human bones. 
That the catacomb had considerable local importance is immediately suggested by the two tall, slender half columns that flank the square-headed doorway. They are roughly carved and have elongated capitals and fluted and banded shafts. Upton Way picturesquely likened them to “latin fasces” but the general effect is unclassical. Immediately inside the entrance on the right wall is the rude carving of a quadruped of undetermined species, apparently with its head lowered in the act of grazing or drinking. Musgrave  suggests a stag but the antlers are seemingly missing and a lamb is more probable. Both were common early Christian symbols.
The plan of the catacomb is different from that of the ‘rectangle’ hypogea for it consists of a single short gallery, 6.10m long. On the left hand side, on entry, is a well-preserved agape-table flanked by two low seats carved from the triclinium. A plain cornice in the form of a corbel-like projection runs round the flat ceiling of the exedra and the flattened arch at its entrance is decorated by a small engraved latin cross. The curved wall at the back of the table contains two window-tombs, ‘1-2’. Window-tomb ‘2’ is more noteworthy and its rectangular doorway is cut in the back of an arched niche with corner pilasters. At point ‘x’ in the ceiling of the exedra, a concave depression suggests that another window-tomb was destroyed when the [p.9] catacomb was enlarged to provide space for the digging of a large saddle-backed baldacchino, ‘5’. 
Another saddle-backed baldacchino, ‘3’, is excavated opposite the agape-table. It is free-standing and forms an island surrounded by narrow corridors with finely proportioned arched openings. One of these corridors contains a window-tomb, ‘4’, with a single head-rest. Baldacchino ‘3’ is one of the most ornate tombs of early Christian Malta. The cubical rock platform with its large immovable sarcophagus-lid contains a bi-partite chamber with a rectangular window-opening on the side facing the table. The canopy above has a very flattened arch flanked on the inside, by two short, badly mutilated, rope-pattern half columns, while the side pilasters contain a beautiful graffito simulating a pilaster with a pseudo corinthian capital and a shaft with herring-bone decoration (fig. 5d).
Next to this tomb in the inner part of the gallery is another saddle-backed baldacchino, ‘6’, which is, however, only partially detached. It is less architecturally imposing than ‘3’ but more decidely Christian. On its front-facade, the left hand pilaster is decorated with a rough but deeply incised graffito of an X-cross, while the right hand pilaster has a carefully carved Greek cross inscribed in a compass-drawn circle. A similar but smaller cross is engraved on the side facing tomb ‘3’ The unusually well rounded arch of the canopy is here flanked by two pilasters and framed by a hood-mould which curls at either end to form a pair of rosettes. The pilasters are respectively fluted and rope-patterned. They have plain capitals but no plinths (fig. 5c).
The third saddle-backed baldacchino ‘5’, which is cut in the agape-table recess, is decorated with spiral volutes that curl in almost baroque profusion round the canopies of two of its facades. The facade rising above the triclinium platform is also ornamented with two rope-patterned half columns. Musgrave  thought he could make out
... what appears to have been a dolphin with a basket of leaves above it ...
This is however, highly contentious and the same applies to the “evidence of carved lettering” which he thought he saw “on the lower face” of the same side. 
In the back wall of the gallery is an apparent carving in relief, c. 92cm high. Its subject matter is difficult to determine except for the possible outlines of a standing figure in the middle and a branched twig in the top left hand corner. It was covered with mud which was removed in 1975 at the suggestion of.Musgrave.  Harrison Lewis who saw it before it was cleaned thought it represented Christ between SS. [p.10] Peter and Paul.  Musgrave imagined St. Paul. Both are very subjective interpretations for the ‘carving’ is either unfinished or so badly mutilated that any guess is as good as another.
This is a small hypogeum consisting of a short narrow passage flanked by two tombs on either side. It is situated a few centimetres to the south of Hypogeum ‘5’ and faces east. It is full of soil and loose stones and flooded with water most of the year.
Hypogea ‘7-8-9-10-11-12’ (fig. 1)
These six rock-tombs are tunnelled into a rocky ledge at the back of a field to the north-north-west of the ‘rectangle’. Only ‘7’ is normally accessible; the others are, at the time of writing, filled with field soil. Hypogeum ‘7’ (fig. 4b) is reached down three steps from an arched opening which may have been sealed by a plug-door.  It consists of a narrow passage flanked on either side by a flattened-arched arcosolium. The arcosolium on the left has not got a burial trough and is, presumably, incomplete. The other has two troughs with one head-rest each.
According to Musgrave, who may have investigated it, Hypogeum ‘11’ has a similar plan but the central passage is, apparently, longer and it is not clear whether the tombs are window-graves or arcosolia.  Hypogeum ‘12’ may also be similar. Musgrave detected a narrow gallery leading into the hill but “a sturdy vine growing across the entrance” prevented its exploration.  Judging from his plan,  ‘8-9’ are probably two window-tombs with an independent entrance while ‘10’ may either be the mutilated chamber of a window-tomb or, as Musgrave suggests, a loculi-tomb. 
Hypogea 13-14 (fig. 1)
These are excavated in a field to the north-north-east of the ‘rectangle’ and they are likewise at present inaccessible. Musgrave shows them as two shallow excavations in the hill-side. He describes ‘13’ as a loculi and suggests that ‘14’ may have been a second entrance to Hypogeum 15. 
[p.11] Hypogeum 15 (fig. 6)
Despite its diminutive size this is, after Hypogeum 5, the most ornate of the Salina Hypogea. It is situated in the first field after leaving the Annunciation church, in the lowest part of the hillside, beneath the ‘rectangle’ to the north-north-east. It was first reported by Serracino Inglott who apparently thought that it was cleared by Upton Way.  This is, however, incorrect and no details are, unfortunately, available about its investigation. It is entered through an arched opening, 60 x 57cm, protected by a low masonry wall built recently to prevent it from flooding during the rainy season. The floor, 50cm below, is reached down two steps which lead to a short gallery at the end of which is a now walled-up second entrance which Musgrave associates with Hypogeum 14  but which, more probably, is a modern opening through which the hypogeum may have been despoiled in recent times. There are only three tombs: two bi-partite-arcosalia, ‘1-2’, with very flattened arches, and a child-loculi, ‘3’. Arcosalia ‘1’ on the right hand wall has two rope-patterned half columns which are, unfortunately, mutilated. The platform is also partially destroyed. Arcosolium ‘2’ on the opposite wall is better preserved and more ornate. It is framed by deeply incised spiral volutes that curl round the top of the arch and it has a crisply carved rope-patterned half column in the left hand corner. The corresponding column on the right was left uncarved (fig. 6b).
The most interesting feature of the hypogeum is a small carving, 27cm high, in the confined space between arcosolium ‘2’ and child-loculi ‘3’. It represents a biped above which is a chi-rho monogram and an alpha and omega. Harrison Lewis and Musgrave think that the obviously allegorical animal is a lamb  but its species is not easily determined. It has a humped back and an elongated snout that looks like a bird’s beak (fig. 6c). Serracino Inglott is probably right in suggesting a pelican as at Ħal Resqun Hypogeum near Gudja.  The chi-rho monogram excludes a date prior to the mid-fourth century.
The entrances of two other rock-tombs show above the field soil closeby. They are a reminder that this important site is still only very partially explored. There is much scope for a full scale archaeological dig but it is imperative to act quickly. Already development (in the form of macadamised roads, hotels and modernistic villas) has encroached dangerously close and some tombs have already been destroyed.
Decorative and iconographical considerations suggest late date for some of the Salina hypogea.  The chi-rho monogram in Hypogeum 15 and the baldacchino-tombs in Hypogea 1, 3 and 5 all point to a post 4th century date but the unfortunate dispersal of the archaeological material and the absence of adequate excavation reports do not allow a more precise dating. It is, however, probable that the site had a long history. Harrison Lewis  claimed to have collected pottery sherds dating from the second to the 6th centuries A.D. but his dating is dubious.  More important is the gold coin of Phocas (602-610) mentioned by Ciantar which seems to suggest that the site was in use in the early 7th century. There is so far absolutely no evidence for Punico-Hellenistic or earlier tombs and the site may have been exclusively Christian.
It is not yet possible to establish a chronological sequence for the different hypogea but the bone deposits in Hypogeum ‘2’ and in the 1721 tomb suggest an early phase and a subsequent utilisation as ossuaries. Ciantar’s description of the 1721 tomb  makes an apparent reference to arcosolia but does not mention carved decoration. This is certainly absent in Hypogeum 2 and, in fact, in the three other hypogea of the ‘rectangle’. The two plain pilasters flanking arcosolium ‘4’ in Hypogeum 1 are the only exception. It does seem possible that the four ‘rectangle’ Hypogea belong to a different period from Hypogea 5 and 15. These have a different plan and share a preoccupation with iconographical symbolism and ornate decoration. The rock-tombs still await investigation but they seem to have much in common with the ‘rectangle’ Hypogea. The external floor-tombs, on the other hand, form a class by themselves and may represent the last phase of the site when, perhaps as a result of new cultural currents, the centuries old tradition of underground hypogea was gradually being replaced by above ground cemeteries.
 D.H. Trump, Malta: An Archaeological Guide, London 1972, p. 132; A. Bonanno, “Distribution of Villas and some aspects of the Maltese economy in the Roman period,” Contributions to Mediterranean Studies, Malta 1977, p. 78; id. “L’habitat maltese in età romana.” KOKALOS 22-23 (1976-77) p.30.
 J.D. Evans, The Prehistoric Antiquities of the Maltese Islands: A Survey, London 1971, pp. 202-204.
 G.A. Ciantar, Malta Illustrata ovvero Descrittione di Malta, vol. i, Malta 1772, pp.196-197.
 G.H. Musgrave, Friendly Refuge - A Study of St. Paul’s Shipwreck and his Stay in Malta, Sussex 1979, p. 90.
 The stone seems unfit for building.
 L.J. Upton Way, “Catacombs near Salina Bay,” in Times of Malta 29/6/1937.
 Musgrave, op. cit., p. 90.
 Museum Annual Report 1937-38, p.v.
 No excavation report of these last two catacombs survives.
 Upton Way, op. cit.
 G. Serracino Inglott, “The Salina Necropolis,” Sunday Times of Malta, 26/1/l9ó4, p. 9.
 Musgrave, op. cit., pp. 73-111.
 Ibid., pp. 102-104.
 M. Buhagiar, “The Maltese Catacombs - Characteristics and General Consideration.” Proceedings of History Week 1983, pp. 5-6. See also V. Borg “Malte: Une île et ses hypogées de l’ere des premiers chrétiens,” Les Dossiers de l'Archaeologie 19, Nov-Dec 1976, pp. 52-67.
 One possible hint is the occurence of lamp notes which would not serve any useful purposes in the open unless, of course, they had a ritual significance.
 On the occurence of rock-pillows and head-rests in tombs see M. Buhagiar, op. cit., pp. 4-5.
 G. Serracino Inglott, op. cit.
 M. Buhagiar, op. cit., p. 12.
 On these doors see M. Buhagiar, pp. 14-15.
 On ‘agape-tables’ and their significance see V. Borg op. cit and M. Buhagiar, pp. 18-21.
 The numbers marking the discussed tombs correspond to the numbers on the plan. A bipartite window-tomb is a window-tomb with two head-rests. All the tomb-types mentioned in this paper are discussed and illustrated in Buhagiar pp. 9-12.
 Agape-tables are usually found in apsed recesses or exedras.
 On the use of marble in Maltese catacombs see M. Buhagiar, p. 23.
 Supra, p. 1.
 Borg op. cit., publishes two photographs of the Taċ-Ċaghqi, ‘square table’ while the Secondary School hypogea are discussed in V. Borg and B. Rocco, “L’Ipogeo di Taċ-Ċaghki a Malta”, Sicilia Archaeologica - Rassegna periodica di studi, notizie e documentazione a cura dell’E. P. T. di Trapani, Anno V nn. 18-19-20 Guigno-Settembre-Dicembre 1972, pp.61-74. See also Buhagiar, op. cit., p.19. Another square table existed in Catacomb 21 in the SS. Paul/Agatha Cluster (see E. Becker, Malta Sotteranea - Studien Zur Altchristlichen Und Jüdischen Sepulkralkunst, Strassburg 1913, p.32; Becker, who was unfamiliar with ‘square-tables’ thought it was a seat).
 Upton Way, op. cit.
 On the sealing of window-tombs see M. Buhagiar, op. cit. p.13.
 Upton Way, op. cit.
 Musgrave, op. cit., p. 97.
 Discussed hereunder.
 Musgrave, p. 97.
 Ibid., p. 98.
 Harrison Lewis, Ancient Malta - A Study of Its Antiquities, Bucks 1977, p. 148.
 A thick sealing-slab with a plug-like projection at the back that tight fitted into the window opening of the tomb. For further details on plug-doors see Buhagiar, op. cit., p. 13.
 This tomb is marked ‘J’ on Musgrave’s plan. Musgrave, p. 89.
 Musgrave, p. 93. It is marked ‘K’ on his plan.
 Here reproduced with modification and corrections as fig. 1 b.
 Musgrave, p. 92. These three tombs are marked ‘I’ and ‘H’ on his plan.
 Musgrave p. 89 (where they are marked as ‘L,M,N’ on the plan) and p. 92.
 Serracino Inglott, op. cit.
 Musgrave, p. 92.
 Harrison Lewis, op. cit., p.148; Musgrave, p. 91.
 G. Serracino Inglott, op. cit. On the Ħal Resqun Hypogeum at Gudja see T. Zammit, “An Early Christian Rock-Tomb on the Ħal Resqun Bridle Road at Gudja” in Bulletin of the Museum (Malta) vol. 1 no. V, February 1935, pp. 189-195.
 Dating considerations are discussed in Buhagiar, op. cit., pp. 24-25.
 Harrison Lewes, op. cit., p. 147.
 In a letter to the Sunday Times of Malta (5/6/1977) Professor Pasquale Testini of the Pontifical Institute of Christian Archaeology, Rome, describes how Mr. Harrison Lewes called upon him with some sherds and insisted on a dating “and was unable to understand that neither I nor my assistants were in a position to give any judgement without an adequate study of the sherds. Eventually and with considerable uncertainity he was given an approximate date but with the added warning that it was nothing ... but a first summary impression”.
 Supra, p. 1.