¹ Goodchild, Richard, ‘A Byzantine Palace at Apollonia,’ Antiquity, xxxiv, 1960, pp. 246–258CrossRefGoogle Scholar.

² Concerning the authors of this report: Mr. Richard Goodchild, as Director of the Department of Antiquities, oversaw the entire operation; Mr. Widrig assisted during the summer of 1958 and was in charge of the sorting out process the following summer. Mr. Widrig's presence in Libya was in part financed by New York University directly and also by the Phyllis Lambert Architectural Research Fund of New York University. The original architectural survey and plans were done by A. Abdussaid of the Department of Antiquities. They have been prepared for publication by Aurelio Tassinari. Fig. 1 was drawn by Philip Oliver-Smith.

³ F. W., and Beechey, H. W., Proceedings of the Expedition to Explore the Northern Coast of Africa from Tripoly Eastward, London, 1828Google Scholar.

⁴ The staircase seems to lead not to an upper level of the church but to a once-existing platform set into the Hellenistic defence wall of the city. There would be more need for a strengthening of defences, before rather than after the Arab invasion, and from this it may be assumed that the staircase is Christian in date.

⁵ Conceivably these walls might relate to the strengthening of defences mentioned in footnote 4, on the other hand, since they interrupt a well defined corridor giving direct access to those sections of the forecomplex which have a specifically religious function, they may well belong to the post Christian or Arab period, when secularisation of church and forecomplex eliminated the ritual need for intercommunication.

⁶ Raids by desert tribesmen might necessitate repair to the church, but would not result in its secular conversion. The Arab invastion of A.D. 642 certainly destroyed the urban organization of Cyrenaica, Barka (El-Merj) becoming the new capital in place of Apollonia. The ‘Central’ and ‘East’ churches of Apollonia were deliberately destroyed, as can be seen from the undercutting of their columns; and nowhere in Cyrenaica is there evidence to suggest that the ritual observance of Christianity was allowed to continue into the Arab period.

⁷ Since this wall cuts across the room at a point which continues the line of the west wall of the porch, it undoubtedly was involved in a reroofing scheme for the east end of the church.

⁸ At the same time the eastern respond wall of the left colonnade seems to have been rebuilt and a door inserted to allow access into the newly created room; probably the first opening in the colonnade was blocked, although positive evidence is lacking.

⁹ Again indication of a new roofing scheme on the east.

¹⁰ Since the plan of the church is somewhat irregular, the figures given here are necessarily approximations of the mean. For exact measurements from one point to another consult the drawing of the plan (pl. XXXIII).

¹¹ Also no gallery fittings were unearthed. Ruling out galleries for the church almost forces the conclusion that the staircase in the right side aisle has to do with the defence wall and not the church itself.

¹² Probably it was filled with rubble at the same time the walls external to the church on the north were constructed, since these walls would have closed off the northern corridor.

¹³ This door likely was blocked at the same time as the staircase was inserted into the side aisle.

¹⁴ It is unlikely that the area under the synthronos would have been paved.

¹⁵ Enough fragments of the marble chancel screens remain to be able to identify their patterns, and in three cases even their heights (90 cm.). A raised Latin cross superimposed on a raised tondo, a rosette with tendrils pointing to Latin crosses on either side, and small Latin crosses with nothing else are the motifs represented. Some screens must have had perfectly plain surfaces.

¹⁶ The reconstructed colonnette would be almost identical with those found at Latrun, 35 km. east along the coast, where there is no question as to their function. The Latrun material for the present is unpublished.

¹⁷ These capitals are very similar to the unpublished ciborium capitals of the ‘Central’ Church at Apollonia.

¹⁸ Conceivably these could have been very different in design and execution, as well as in diameter.

¹⁹ The marbles here present are the following: cipollino (a few of the chancel screens); Proconnesian (other screens, the chancel posts, the columns, and the colonnette); a coarse-grained white marble with small grey veins (the combined plinth and base); and a grey-white marble (the capitals).

²⁰ A similar, but larger, monolithic ambon step, cut down from an inscribed pagan marble base, was found in the ‘East’ Church of Apollonia during its excavation in 1921; but there, as in the present case, there was no evidence of its precise original position. Nor have the other excavated churches of Cyrenaica produced evidence of ambons such as certainly existed in some of the churches of neighbouring Tripolitania.

²¹ These capitals are cut to the same pattern as capitals 1, 2 and 4 of the nave colonnades, except that their height compared to diameter is compressed to almost half.

²² It is reasonable that this door should not have been blocked until the Arab period when the area in front was incorporated into the forecomplex.

²³ In situ, sandstone troughs to catch the water from the roof extend out from the exterior face of the south wall of the left side aisle and definitely establish the area here as ‘out-of-doors.’

²⁴ Before the addition of the long room the left side aisle itself, with the possible niche in its west wall, might have substituted for a true sacristy.

²⁵ Ward-Perkins, J. B., ‘Christian Antiquities of the Cyrenaican Pentapolis,’ Bull. Soc. Arch. Copte ix, 1943, pp. 126–131Google Scholar.

²⁶ Consult IV Congr. Arch. Crist.; also Perkins, J. B. Ward and Goodchild, R., ‘Christian Antiquities of Tripolitania,’ Archaeologia, vol. 95, 1953, pp. 1–84CrossRefGoogle Scholar.

²⁷ Romanelli in IV Congr. Arch.Crist., pp. 274–279; Ward-Perkins, op. cit. Ward-Perkins and Good child are now preparing a corpus of the Cyrenaican churches.

²⁸ Coins were few and without telling locations. Their range was from Constantine the Great to Heraclius, but their concentration fell in the last half of the sixth century (this last fact does carry meaning). Pottery finds revealed no more than the coins.

²⁹ Because the capitals of the Central Church are a complete set it is reasonable they were always intended for this churth (besides even their earliest possible date makes them too late to be spoils). They could have been cut at the quarries or even right in Apollonia by foreign craftsmen accompanying the shipment. Either way, a surplus might have been created by a change in plan, a not unlikely occurrence, after execution of the capitals.

³⁰ The blocks of marble for capitals 2 right and 3 left and right could have already existed in Apollonia; it is doubtful whether they were specially imported for the job. Two of the blocks are of extremely poor quality (those of capitals 3 left and right).

³¹ Ward-Perkins, op. cit.

³² The third to fourth century A.D. The extensive use of marble revetment, true opus signinum floors, pebble-concrete floors, and the way of mounting columns here found are not characteristic of fifth to sixth century Christian building practices in Cyrenaica.

³³ Obviously the limestone blocks are reused fragments. On the surface of one such block there was an inscription in Greek, listing names.

³⁴ This is the only instance where the complete destruction of an old wall may be involved: see above.

³⁵ Discussion of the materials, iconography, and date of this mosaic will appear in a forthcoming publication on all of the Cyrenaican mosaics now being prepared jointly by the Cyrenaican Department of Antiquities and the British School at Rome.

³⁶ Ibid.

³⁷ The northern external corridor remained the principal means of reaching the baptistery even after the remodelling. It was only in the Arab period that the suite of rooms containing the baptistery was made once again to connect with the interior of the rest of the forecomplex. Yet the northern corridor could never have been the main path of entry to the church itself. Before the remodelling this might have been from the south.

³⁸ This is a gap only in terms of what is preserved of the Roman-period walls. Certainly there was a wall here before the remodelling since the space to the west was an interior space. At the time this gap was filled the Roman wall on either side was renewed; so also was that part of the Roman exterior east-west wall between the street and room 3 (the baptistery). Thus the angle of the Roman ‘L’ was reconstituted in its same location.

³⁹ The Christian Period I western extension of this wall (the south wall of room 1) was also rebuilt and in the process the door from room 1 blocked.

⁴⁰ A sounding in the courtyard produced no pavement or floor at a lower level.

⁴¹ The bases are round and cut with a concave moulding between two fillets, above and below which are torus mouldings; at the bottom of each base is a circular drum probably meant to be embedded in the wall of the tank.

⁴² Where these walls touch it is perfectly clear that the southern wall (the north wall of the long room) is built up against the northern one (the continuous south wall of the atrium and its preceding room); this proves conclusively its later date.

⁴³ In this small rectangular space there were two poorly constructed east-west walls of uncertain date.

⁴⁴ The apse floor, a few centimetres higher than that of the room, is set apart by a vertical marble sill across the opening of the apse.

⁴⁵ Again it is necessary to await the forthcoming publication on the Cyrenaican mosaics for a complete discussion of materials, iconography and date.

⁴⁶ The narrowing of both of the immediate approaches to the baptistery must signal its end as such and the secularisation of the suite of rooms.

⁴⁷ With the raising of the floor level in room 1 the water channel running past the passageway and formerly blocking it could be gotten over with out great difficulty.

⁴⁸ The capital is of grey-white marble and represents a type similar to, but not from the same set as, the nave capitals 1, 2 and 4; the abacus of this capital is very large in proportion to its body.

⁴⁹ The arch proves beyond all doubt that the space here is an interior space (this is already suggested by the treatment of the doorway); however, this may not have been an interior space until the Arab period.