Sicily 2017

Gerace 2017 – Summary of the Results of the UBC Excavation

The fourth campaign of the UBC excavations at Gerace in central Sicily (fig. 1) lasted for four weeks, from 8th May to 4th June 2017. Excavation was conducted in four separate areas of the site, Areas D, E, F and G. The first two represented a continuation of work started in 2016, while Areas F and G were opened for the first time. In Area G a trench measuring 5 m by 6 m was excavated to explore a geophysical anomaly noted in the 2012 survey, which suggested the presence of a large dump of tiles in the centre of this trench. The tiles excavated did not constitute a major deposit and did not, it appears, belong to any identifiable building. In the western part of the trench a north–south wall, not earlier than the second half of the 4th century AD, was uncovered. It was not clear if it belonged to a building or whether, as is more likely, it formed part of the boundary wall of the estate. On its west side a further wall had been added at a later date, in early Byzantine times (6th century AD?) on the basis of associated roof tiles; but the structure of which it formed part lay outside the trench to the west.

Fig. 1 The setting of the archaeological site of Gerace, in the orchard in the foreground of the photograph

In Area E, in the southern part of the site, the kiln partly identified in 2016 was examined in detail. It proved to be not circular but of ‘keyhole’ shape, with an oval chamber in which the ceramic goods were placed for firing, 2.35 m long and 1.87 m, wide, and an additional long access trench (a ‘dromos’), which served as the pit from which the fire under the oval chamber was lit and controlled (fig. 2). The dromos was 2.12 m long and 0.85 m wide at its narrowest point. Its side walls were constructed of greenish bricks and tiles, the customary colour for over-fired ceramic material. At one point part of the east wall had clearly collapsed and a repair had been made, detectable where the wall bends slightly out of line. Among the material used in this repair was a very large but damaged brick, 49 cm by 34.3 cm and 8.5 cm thick, possibly a rare survival at Gerace of one of the large bricks used for spanning the gap between pilae in rooms with hypocausts (see below). The southern part of this dromos, that closest to the entrance to the combustion chamber, was excavated down to natural: this was yellow sand, underlying yellow sandstone, at a depth of 1.60 m below the surface. The floor of the combustion chamber itself had been completely lost to the plough, which had also destroyed the central part of each of the six arches that supported it. Because of the latter’s fragility it was decided to leave this part of the kiln unexcavated.

Fig. 2 Area E, the kiln of Philippianus from the south, at any early stage in its excavation (scale: 1 m)

Fig. 3 One of the construction bricks used in the same kiln, with the monogram of Philippianus in relief (scale: 10 cm) 

The bricks used to build these kiln arches carry the monogram of Philippianus (fig. 3), whose stamped roof tiles have been ubiquitous at Gerace since the commencement of the UBC excavations in 2013. He is conjectured to have been the owner of the Gerace estate in the second half of the 4th century AD. Why he would he have bothered to have his name placed on the bricks (in relief) when they were never going to be seen? The answer is probably that the raised inscription provided an uneven surface which made the mortar adhere better between the bricks, and so improved the stability of the kiln. In the fill of the dromos immediately adjacent, examples of roof-tile wasters, distorted during production both in shape and colour, were found, and these too bore the name of Philippianus. It seems, therefore, very probable that the kiln was constructed by this estate owner for the firing of roof tiles, since both they and the kiln’s construction bricks both bore his name. This occurred sometime between about AD 350 and 400, the period of Philippianus’ activity at Gerace. The top of the kiln fill, however, included a group of pottery datable to the second/third century AD, a period up to now not well represented at Gerace. It seems likely that in this part of the estate, piles of debris attesting to a long period of industrial activity were lying about on site, and when the fourth-century kiln came to be filled in after its abandonment, material was thrown indiscriminately into it from the immediately surrounding area. The intensity of earlier production of ceramic materials here is further suggested by structures found immediately west of Philippianus’ kiln, where fragments of as many as four further kilns were discovered. The most coherent remains were those of a small kiln, with one arch that supported the floor of its firing chamber still intact; more of the structure lies under the baulk to the north. Its fill produced twelve pieces of stamped brick and tile, including several with the name of Philippianus, but none of these were wasters, and they might represent ordinary rubbish disposal rather than evidence for what the small kiln actually produced. To its south are remains of a further small kiln with a central pedestal as floor support, the pedestal of another former kiln lying horizontally after demolition, and part of yet another kiln further south again (making five kilns in all in Area E). The small kilns with central pier supports are likely to have been used for pottery production. Their chronology is uncertain, but all predate the large Philippianus kiln of AD 350/400, the construction of which partially cut all the earlier kilns.

Fig. 4 Area F, tile kiln as so far excavated, seen from the north. The furnace for the kiln lies off the photo to the right (scale: 2 m)

In Area F, 40 m east of Area E, the geophysical survey of 2012 indicated the presence of another area of firing activity, and excavation in 2017 revealed that this consisted of one large rectangular kiln, 4.95 m (east–west) by 3.54 m (fig. 4). The furnace area lay to the west. The exterior wall was made of stone, but up against it, at least on the east and south sides, was a thick bank of yellow clay, added against the stone wall at the time of construction to enhance the efficiency of thermal insulation. Internally the kiln was divided up by five north–south walls made of large mud-bricks, measuring 55 cm by between 37 and 43 cm, and 9.5 cm thick, which became baked into terracotta only after the first and subsequent firings of the kiln. As usual in such kilns, the central part of these north–south internal partitions was broken by an arch at the centre: both walls and arches were designed to suspend the floor of the large chamber above in which the products of the kiln were stacked. All this upper part, however, has long been destroyed by the plough, and oblique furrow marks digging deep into the surviving structure of the kiln were regularly observed across the entire area of F. The kiln had a broad central space for the circulation during firing of the hot air, which was also able to pass sideways into the narrow spaces, 75 cm wide, which separated each north–south internal wall from its neighbour. This is a common type of kiln arrangement, found widely throughout the Roman Empire.

What ceramic products were made here? The only evidence that we have comes from disturbed plough-soil immediately above the kiln, which produced tile wasters, and it seems therefore very probable that this is another kiln designed mainly for the manufacture of roof tiles. Its size and layout also support this hypothesis. There is not a scrap of evidence, however, to link this kiln in any way to Philippianus, and it may therefore be either later or earlier than his period of activity. A rim fragment of a type of open basin, found built into the wall of the kiln as part of its construction, is of a type found elsewhere at Gerace in contexts of the mid-fifth century and later. That might suggest that the kiln belongs to the full fifth century, but this tentative chronology needs verification through radiocarbon analysis of burnt material from the kiln, which is currently in progress.

In a second phase of use, the kiln was completely transformed. The central arches were demolished, and the whole of the back part of the kiln (towards the east) was filled with earth and stones because it was no longer required. A new but much smaller firing chamber was constructed in the west part of the former kiln, measuring only about 1.75 m by 0. 6 m, and of this the springers on either side of two smaller brick arches (a third pair lies under a tree) were excavated; but again the superstructure containing the floor of the chamber where the products were fired higher up is once again entirely missing. Probably contemporary with this secondary kiln insertion, the extremities of three of the former brick cross-walls of the phase-1 kiln were pierced by irregular, vertical shafts, one side of which was formed in each case by the original exterior stone wall of the original kiln. The irregularity was no doubt caused by the difficulty of the digging operation: trying to create a large hole in the middle of solid brickwork would not have been easy. Probably six such shafts were dug, of which the three more northerly ones were clearly detectable by the black charcoal deposits with which each was packed; of the three pits at the southern end of the same three cross-walls, one is clearly definable but the other two have been compromised by serious plough damage. Only one of these shafts has so far been fully excavated, on the north side: it measured 54 cm by 80 cm and was 81 cm deep. The charcoal it contained has been carefully collected. The most plausible explanation at present of this industrial process is that charcoal production was being carried on here, but in the absence of an air source at the bottom, bellows must have been used to start the fire as wood was being laid down in successive layers on top of it to allow for the subsequent slow smouldering process, necessary to turn wood into charcoal. Charcoal is a necessary fuel for metal working, but the evidence of how it was made rarely survives from the Roman period, since such charcoal-making ‘clamps’ were usually built on the surface, and their temporary nature means that they have normally disappeared completely. In view of the potential importance of the discovery, it is intended to excavate these deposits more fully in a future year, if further funding for the project can be obtained.

Fig. 5 Detail of the pilae in tepidarium 2 (foreground), and the connecting underfloor flue to tepidarium 1 (top), from the south (scale: 1 m)

In Area D, excavation continued of the bathing complex first identified in 2016. Excavation of a heated room of which a small part was uncovered in 2016 was completed (internally it measured ca. 2.40 m square), and an additional heated room, 2.42 m by only 1.52 m, was excavated alongside it to the north. These rooms were both tepidaria, and preceded the caldarium which was partially excavated last year: the new small heated room is tepidarum 1 and the one to the south of it tepdarium 2. Heat for tepidarium 1 was taken directly from the adjacent caldarium via an underfloor flue (fig. 6, top right); by contrast, tepidarium 2 had its own praefurnum (furnace), discovered last year. The walls of these rooms stand up to 2.25 m high, and the floors, as customary with heated rooms in the Roman world, were raised above hypocaust basements and supported on small brick pillars (pilae) (fig. 5). These floors, which were paved with geometric mosaics, had, as expected from our work in 2016, been broken into small fragments (see below), pieces of which were found lying on the basement floors (fig. 6). There is enough evidence from the fragments, however, at least for tepidarium 1 (fig. 7), to reconstruct the original design. The walls of the bath’s heated rooms were originally faced with a thin veneer of coloured marbles, of which only fragments were found in the debris. A dozen different marbles are represented, imported from Italy, Greece, Turkey and Tunisia: the ones illustrated here are pavonazetto from near Afyon in Turkey (fig. 8), portasanta from Chios in Greece (fig. 9), and verde antico from Larisa in Thessaly, also in Greece (fig. 10). Their presence betokens a high-status building. The bath-house, however, had been systematically stripped on decommissioning, an interesting example of Roman recycling. This included removing marble decoration from the walls, the hot air convectors made of terracotta (tubuli), also from the walls, and the large bricks used to span the gaps between the pilae in order to support the floor. It was in order to recover the last item, and also some of the smaller bricks of which the pilae are composed, that the mosaic floors, which were not recyclable, were so comprehensively smashed.

     Fig. 6 Tepidarium 1 from the west, showing smashed pieces of mosaic floor in situ (scale: 1 m)

Fig. 7 Two adjoining pieces of the mosaic after lifting and cleaning, from the north-east corner of tepidarium 1 (scale: 20 cm)

Fig. 8 A piece of pavonazzetto marble from Turkey, found discarded in the frigidarium        

Fig. 9 A piece of portasanta marble from Chios, Greece, found in tepidarium 

Fig. 10 Pieces of verde antico marble wall veneer from Larisa, Thessaly, Greece, from tepidarium 1

To the north of tepidarium 1, part of the cold room (frigidarum) was uncovered, a room of notably larger dimensions (5.82 m by 6.15 m) than the rest of the rooms in the bath-house. Part of its south wall used rammed earth (pisé) in its upper construction, chosen for what was thought to be its good insulation properties: the design of the baths placed the caldarium and frigidarium adjacent to one another. The experiment obviously failed, because a new north wall for the caldarium was built in a secondary period a short distance to the south, no doubt compromising its mosaic floor in the process. On the east side of the frigidarium were two pools with cold water (piscinae) side by side, of which only one (fig. 11) has so far been excavated (the other lies under a tree). At the base the pool measures 2 m by 1.60 m, and is reached by two steps from the frigidarium. Water arrived in two pipes, via holes in the east wall, from where it would have cascaded down into the pool. The walls of the pool ware covered by white lime-mortar, still intact, and its floor was originally paved in marble, of which a few tiny scraps remained. The water exit-pipe, of terracotta, is still in situ in the north-west corner of the pool (fig. 12). Another drainpipe to take away excess water was also located at the entrance to tepidarium 1.

Fig. 11 The excavated cold-water pool (piscina) in the frigidarium, from the south-west (scale: 2 m)

Fig. 12 Detail of the steps leading into the cold water pool. The choice of a limestone slab on the left (the rest of the floor was of marble) may have been for safety reasons, being less slippery than marble when entering the water. Note the ceramic outflow pipe to the right. Scale: 50 cm

The frigidarium itself is paved in geometric mosaic in a good state of preservation. The design consists of pairs of irregular hexagons (‘shields’), one superimposed on the other but at right-angles to its companion. The borders of the hexagons are composed of laurel wreath bands, one executed in white, yellow and red, the other in white, yellow and blue (fig. 13). While cross-‘shield’ patterns with guilloche borders, separated by (regular) hexagons and octagons between, are quite widely known in the Roman Empire, including one in Sicily (at Marsala), the precise design used here of tangent hexagons (without intermediate shapes) is, to the best of my knowledge, unique. The pavement was probably laid ca. AD 380/400. A mosaic inscription runs around all four sides of the room, a highly unusual arrangement that may itself be a unicum in the Roman world. As so far uncovered, it reads: ]IVM/PLVRA FABRICETIS MELLIORA DEDI/CETIS ASCLEPIADES SENESCAS CVM TVIS/ PHIL[, viz. ‘may you build more and may you dedicate better things. Asclepiades, may you grow old with your family. Phil[. It seems almost certain that the last name (fig. 15) refers to Philippianus, who constructed the baths (Asclepiades might have been his son): that he did so is made clear by the appearance of his name in monogram form at the centre of one of the roundels on the mosaic (fig. 14). It is the same monogram that appears on his bricks in the kiln (fig. 3). Monograms are rarely used in any context in the second half of the 4th century AD, and Philippianus was therefore highly unusual in using it: in mosaic, it is paralleled in just one other site in the Roman Empire, in a villa at Cuevas de Soria in Spain, which is approximately contemporary with the Gerace bath-house. The otiose second L in MELLIORA is presumed to be an error of the mosaicist, who perhaps was ignorant of the correct spelling: there is no evidence that two Ls was an acceptable late Latin variant of this word (fig. 16)

Fig. 13 Mosaic in the frigidarium, from the west, with the entrance to tepidarium 1 on the right, and the step into the cold pool at the top (scale: 50 cm) 

Fig. 14 Detail of the frigidarium mosaic, with the monogram of Philippianus in the floral roundel (scale: 20 cm)

Fig. 15 Detail of the south-west corner of the frigidarium mosaic, with PHIL[ippianus?] in the inscribed border at top right (scale: 20 cm)

Fig. 16 Detail of the frigidarium mosaic and part of its accompanying inscription (scale: 20 cm)

At a certain moment in the 5th century, not before ca. AD 450, use of the baths came to a sudden end. The reason was probably an earthquake – not that which had destroyed the large storehouse, attributed to one attested by Libanius as having destroyed the cities of Sicily during the reign of Julian (AD 361/3), but another, later one, not recorded in the sources. In the the east wall of the excavated  frigidarium pool there is a substantial fissure, and its line is carried through to a lesser break in the entrance wall and the steps of the piscina, as well as in the mosaic floor immediately adjacent. A vertically-lying piece of pavement found in one of the pools of the caldarium in 2016 also points to serious earthquake damage. Although there seem to have been attempts, initially, to patch up the baths – the west wall of the frigidarium was repaired, partially encroaching on the mosaic (fig. 15, top right) – the baths were soon abandoned, stripped of their reusable materials, and systematically filled in. The earliest material in this fill can be dated to not earlier than ca. AD 450 (fig. 17), and includes fragments of lamps, both African ones (fig. 18), and local Sicilian copies of much inferior quality (fig. 19). Later, in the sixth century, a low-status village grew up over the ruins of the Roman estate.

Fig. 17 Part of an African red slip dish, with stamped decoration on the bottom, found in the fill of the cold pool  (scale: 5 cm)  


Fig. 18 An African red slip lamp featuring a rooster on the central discus, fifth century AD


Fig. 19 A Sicilian-made white-slip lamp, of a type imitating an African lamp form, with a running animal (a deer? a dog?) depicted on the discus


My first thanks go to the proprietor of Gerace, Signora Antonella Fontanazza Coppola, without whose permission to enter her land, in order to conduct the excavation, this research would not have been possible; as always she followed our results with great interest. Nor would the excavation have been possible without the financial support of an Insight Grant made by The Social Studies and Humanities Research Council of Canada, to whom my debt is profound. Great gratitude is further due to the Regione Siciliana, and in particular to its Direttore Generale dei Beni Culturali e dell’Identità Siciliana, Professor Gaetano Pennino, who granted UBC the honour and the privilege of a concessione di scavo through the good offices of the Soprintendenza per i Beni Culturali di Enna. I would particularly like to thank the Soprintendente of Enna, Arch. Salvatore Gueli, and the Director of the Sezione Archeologica, Dott.ssa Pinella Marchese, for their constant help and encouragement. Dott.ssa Enza Cilia Platamone, the first excavator of Gerace in 1994, has been as ever an inexhaustible source of support, friendship and advice, and who followed the results of our excavation with the greatest enthusiasm. The constant support and practical assistance of the entire Stellino family at Agriturismo Il Mandorleto was as ever an unshakable central pillar in the success of the whole expedition; their kindness to us knows no bounds. Their next-door neighbour Dott. Carmelo Fontanazza was also a source of constant encouragement, help, practical support, and wise advice; he also kindly put an outbuilding at our disposal which served as our pot-shed. Dott. Salvatore Burgio was our capable conservator, and I thank him also for his congenial company. Both Tomoo Mukai, our pottery expert, and Sally Cann, who drew the finds, demonstrated exemplary skill and unending patience and good humour. Lorenzo Zurla once again wove his magic in creating magnificent aerial photographs of the excavations at the close of our four weeks of labour. Michael MacKinnon examined the animal bones with his customary enthusiasm, exemplary scholarship and endless good cheer. The separation of carbonized seeds and wood from selected soil deposits was carried out by Danielle Maerlander with professionalism, efficiency and an engaging chuckle. Mark Burch, our surveyor, was instrumental in laying out the trenches under difficult conditions, and helped greatly in the interpretation of Area F towards the end of the excavation. Lesley Dunwoodie and Antonietta Lerz were our site supervisors, whose service was as always sans pareil. The whole project owes more to them than words can express, and I am truly grateful to them for their outstanding skill as excavators, their unceasing hard work, their unquenchable sense of humour, and their brilliance as teachers too. Last but far from least, I was exceptionally lucky to have the outstanding dedication, determination, tireless energy and ceaseless good cheer of a wonderful student workforce, who worked through the pain barrier to shift an impressive amount of soil (fig. 20): Hannah Cavers, Lalla Ciccone-Coulson, Daniel Dompierre-Outridge, Judith Fleerackers, Samuel Gagnon-Smith, Taisei Inove, Jenna Kolotyluk, Kyra Kielo, Kylie Lloyd, Katlin Long-Wright, Olivia Quon, Madeleine Seed, Jem Tari, Becca Williams and Dani Tsimbaliouk. Daniel also served most efficiently and with unruffled calm as Comptroller of the Pot-shed, ably assisted in the post-excavation analysis by Dani, Katlin and Taisei. Siena Wood Hutton, who also joined us in the post-excavation phase, most nobly and cheerfully took on the gargantuan task of cataloguing all the 862 smashed mosaic fragments from tepidaria 1 and 2. Without the extraordinary efforts of the entire team, Gerace would not have yielded up in 2017 such a superlatively exciting and important new body of information, which has once more greatly expanded our knowledge of this site. I am truly grateful to them all.

R. J. A. Wilson

22 September 2017

Fig. 20 The Gerace 2017 team


Further reading

  • R. J. A. Wilson, ‘‘Tile-stamps of Philippianus in late Roman Sicily: a talking signum or evidence for horse-raising?’, Journal of Roman Archaeology 27 (2014), 472–486
  • R. J. A. Wilson, ‘UBC excavations of the Roman villa at Gerace, Sicily: results of the 2013 season,’ Mouseion3 12.2 (2012) [published 2015], 175–230
  • R. J. A. Wilson, ‘Scavi alla villa romana di Gerace (EN): risultati della campagna 2013,’ Sicilia Antiqua 12 (2015), 115–148
  • R. J. A. Wilson, Caddeddi on the Tellaro: a late Roman villa in Sicily and its mosaics [Bulletin Antieke Beschaving Supplement 28], Leuven 2016, 15–21
  • R. J. A. Wilson, ‘UBC excavations of the Roman villa at Gerace, Sicily: results of the 2015 season,’ Mouseion3 14 (2017), 253–316
  • R. J. A. Wilson, ‘Scavi alla villa romana di Gerace (EN): risultati della campagna 2015,’ Sicilia Antiqua 14 (2017), in press (due early 2018)
  • R. J. A. Wilson, ‘UBC excavations of the Roman villa at Gerace, Sicily: results of the 2016 season,’ Mouseion3 15 (2018), in press (due April/May 2018)
  • R. J. A. Wilson, ‘Philippianus and his late Roman estate in central Sicily: the Gerace project’, Current World Archaeology, in press (issue for June/July 2018)
  • R. J. A. Wilson, ‘Philippianus and his rural estate in late Roman Sicily: recent excavations at Gerace’, in P. Higgs and D. Booms (eds), Sicily: culture and conquest. Papers given to a conference in the British Museum, 24th–25th June  2016, London: British Museum, forthcoming (due late 2018?)
  • R. J. A. Wilson, ‘Philippianus: a late Roman Sicilian landowner and the history of the monogram’, in M. Trundell (ed.), Studies in memory of Garrett George Fagan, Ann Arbor: University of Michigan Press, forthcoming (due December 2018/early 2019)
  • R. J. A. Wilson, ‘UBC excavations of the Roman villa at Gerace, Sicily: results of the 2017 season,’ Mouseion3 16 (2019), forthcoming (due spring 2019)