The site of a Roman villa at Gerace was discovered by accident in 1994 when a torrent burst its banks and cut through one corner of an ancient structure, exposing a mosaic. Subsequent limited excavation discovered the ground plan on the surface of a small structure with five rooms and an irregular L-shaped corridor. Trial trenching descending to floor level suggested that there were geometric mosaic pavements in a corridor and in an apsed room. This building was further partially investigated in 2007, but has not been completely excavated. In May 2012 UBC conducted its own first investigations at Gerace, involving a team from the British School at Rome, which conducted geophysical survey over a wide area of the 3-ha site. This identified a 50-m long building to the east of the structure with mosaics, as well as several outbuildings and the location of five kilns. From mid-May to mid-June 2013 the first of four planned seasons of excavation was conducted at Gerace. Two rooms in the ‘villa-like building’ were excavated, and proved to be service rooms, one with a bench and a stone ‘workstation’ (to waist height) as well as an earth floor (perhaps a kitchen), and the other with white plaster on the walls and a white mortar floor. The building, for which a late second century date had been proposed by one previous excavator, and an early fourth century date by another, was dated to not earlier than AD 360 on the basis of African red slip pottery which formed part of the white mortar floor in the latter. Part of the mosaic-paved corridor outside these rooms was also investigated, and the edge of what was clearly the hot pool of a small bath-suite, with white mortar floor still in situ and its hypocaust stoke-hole preserved, was also discovered. The building was destroyed by fire: pottery and two intact African red slip lamps of the second half of the fifth century show that this occurred not earlier than c. AD 450. Adjacent to this structure, the 50m-building first identified by geophysics proved to have an intact stone paved floor but very few finds; it is likely to have had only a short life, and possibly might not even have been quite completed, when it suddenly collapsed, probably in an earthquake. It clearly predates the bath-suite and its stoke hole which demolished part of the long building’s west wall in order to provide room to fire the hypocaust. Pottery in the make-up for the long building’s floor suggest that it is not earlier than the second quarter of the fourth century (and part of an earlier building was identified beneath); it may have been under construction in AD 361/3 when it was flattened by an earthquake which _Libanius_ reports as having destroyed most of the cities of Sicily at that time. The building’s function is enigmatic on the evidence so far available, but it was probably the estate’s granary or storeroom. The finds included 99 tile stamps using 10 different dies, with some tiles receiving as many as three stamps. All seem to have been part of a single production, by a landowner called _Philippianus_ whose name recurs on many of them, and were made for the roof of the villa built after c. AD 370 (the granary-like building’s collapsed roof has no tile stamps). That he might have raised prize racehorses at Gerace is suggested by some of the stamps which feature horses with head plumes, associated also with victory crowns and palm branches. _Vegetius_ and others report that Sicilian circus ponies were highly rated in the Roman world, and _Philippianus_ might have been raising them in this well-watered central area of Sicily in late Roman times. Indeed horses are still kept on the Gerace estate to this day. The alternative is to see this imagery merely as a play on his name, which is cognate with Philippos, ‘lover of horses’. No excavation was able to take place in 2014 but it is hoped that work will resume in 2015.

A second season of excavation was conducted in 2015 at Gerace (EN). In the small late Roman villa of c. 370/375 AD identified by earlier work, the largest room, in the west wing, was found to have a floor of opus signinum, but the west corridor, in contrast to the mosaic-paved south corridor, had a floor of beaten earth throughout its life. The villa may never have been completely finished, even though occupation debris suggests it was occupied for much of the fifth century. Of the small freestanding bath-house at a higher level alongside it, only a single heated pool had survived later demolition. In its place two storerooms were inserted deep into the earth for efficient insulation. In the fire that destroyed the villa c. 450/530 AD, a date supported by C14 analysis, the contents of these stores were carbonized: a huge cache of thousands of seeds, principally barley but also including wheat, grapes and legumes, was uncovered there. The much larger, earlier store-building to the east, provisionally dated to c. 325/350 and destroyed c. 360, is now shown to have been a basilican structure. Three earlier phases of building have been detected below it; there is also a sherd of Attic Black Figure pottery, attesting activity at Gerace c. 500 BC. A further 37 examples of stamps on roof tiles were discovered, including a single example from a hitherto unrecorded die; they were all part of the production, probably on site, of Philippianus, the likely owner of the Gerace estate in the second half of the fourth century. After the fire of 450/530 AD, early Byzantine structures (seventh century?), much disturbed by the plough, occupied part of the site of the late Roman villa