The Small Bath

A modest bathing complex sits along the ancient coastal road approaching Antiochia from the east, approximately 90 m outside the city’s main gate. Named the Small Bath (in contrast to the much larger Great Bath, which lies ca. 250 m west and inside the city walls), its hall-type design comprises a central rectangular hall flanked on either side by a series of smaller rooms (Fig. 4).Footnote ¹³ The building is oriented southeast–northwest and was entered from the eastern side via a vaulted entry corridor running northeast–southwest. Along the northern flank of the hall are four rooms (Rooms 1–4), Rooms 1–3 having been identified as cold rooms due to their northerly aspect and the presence of mosaics, and Room 4 as a latrine. Sitting at a slightly higher level, on the exterior north side of the complex is a rectangular, barrel-vaulted room lined with waterproof plaster (Room 5). A circular opening in its southern wall, abutting Rooms 2 and 3, along with a heavy clay-like deposit at the same end and the room’s location adjacent to a water source, collectively suggest its function as a cistern. This supply of water to Room 3, with its circular pool, is especially relevant when considering how the complex was repurposed as a wine production facility (below). The southern flank of the complex is comprised of rooms that would have been heated, and excavations have thus far brought to light a glass furnace and a kiln, whose precise functions and dating are yet to be determined. Their installation postdates the use of the structure as a bathing complex (see below).Footnote ¹⁴

Fig. 4.

Plan of the Small Bath, including mosaic design and later architectural additions highlighted in red. Room numbers correspond to those given in the text. H = Hall. (Plan by G. Demet.)

Local stone and a white lime mortar characterize the architecture of the Small Bath, but remains of white marble revetment, painted wall plaster, niches for statuary, and in situ mosaics hint at what was once a well-appointed complex. Indeed, mosaics have so far been excavated and conserved in the Hall and Rooms 1, 2, and 3. Their designs suggest a date from the late 2nd to early 3rd c. CE, and include star and lozenge patterns, bands of guilloche and waves, and individual shapes (e.g., flowers, cubes, knot motifs) within octagons, placed inside an arrangement of triangles and squares. Along with a fairly homogeneous color palette of red, black, and white, these mosaics present some degree of similarity to those from within the courtyard of the Great Bath.Footnote ¹⁵

Baths of this type are documented at Seleucia in Pamphylia and at least three cities in Rough Cilicia: Syedra, Antiochia, and Anemurium.Footnote ¹⁶ They all date to the 2nd to 3rd c. CE, based on the design of their mosaics,Footnote ¹⁷ and share a similar plan: a central hall accessed via a corridor at right angles flanked by a row of cold and hot rooms on either side, the latter of which have a projecting apse. These similarities have led to suggestions that the baths were designed by the same architect or that there was strong cross-influence.Footnote ¹⁸ While we do not know the identity of the patron or commissioner of the Small Bath, it is conceivable that it was a local of notable standing and that the construction of the bath was an act of euergetism that demonstrated their engagement with local traditions and connections to the imperial family.Footnote ¹⁹ Its peri-urban position alongside the main coastal road, with an entrance designated by large pavers, made it an obvious marker when entering the city.

Like other structures at Antiochia, the Small Bath experienced several phases of secondary use. Shapur I’s invasion of coastal Cilician cities in the 260s CE may have disrupted the Bath’s original purpose, its fate made worse by its extramural position, and it seems never to have been reinstated as a bathing complex. At least one phase as a makeshift burial ground has been identified, with human remains revealed in Rooms 1 and 3, the Hall, and the Corridor. Carbon-14 analysis of bone samples places these individuals between the 10th and 11th c. CE, with the individuals in Room 3 deposited after the large agricultural production basins had been filled in and smaller rectangular basins subsequently installed (below).Footnote ²⁰ More recently, two separate hoards of over 3,000 mostly silver, but also bronze and gold, coins were deposited with a terminus post quem of the 1620s.Footnote ²¹ Numerous square and rectangular basins in several areas reflect a phase of industrial production, perhaps in the Middle Byzantine period. Some of these were filled with a heavy clay layer, possibly indicating their use for clay processing, or as pools in leather-tanning processes. Comparanda can be found in bathhouses at Amorium and Hierapolis.Footnote ²² Of most interest here, however, is the first transformation of the Small Bath – a moment which saw at least its entire northern half converted into a facility for agricultural production at substantial scale, most likely to process grapes and olives into wine and olive oil.

Repurposing for agricultural production

At least four areas within the Small Bath were converted, through varying degrees of intervention, to outfit them for the purpose of agricultural production. These architectural modifications – which restructured the building and changed the way in which people moved through and interacted with its spaces – can be divided into two groups: 1) retrofitting of existing architectural elements in Rooms 2 and 3; and 2) additions and alterations in the Hall for the installation of processing equipment. For the former, a neat, rectangular channel (0.17 × 0.31 m) was cut through the base of the wall that divides Rooms 2 and 3. As part of the latter and more extreme modifications, a new wall was constructed in the Hall, approximately 3.75 m from its eastern entrance, dividing what was once a continuous space outside of Rooms 1–3.

Considering this addition to the Hall, comparatively little change took place in Room 2. The western section of its mosaic floor shows evidence of wear, suggesting that some sort of repetitive action took place in this part of the room. Indications appear to get stronger closer to the western wall, with the border of white tesserae immediately against the wall especially worn down, close to where the abovementioned channel leads into Room 3 (Fig. 5). The channel is rectangular in profile and was formed using mortar and stones much smaller than those of the pre-existing wall. Its interior is coated in opus signinum, suggesting that it conveyed liquid between the two spaces.

Fig. 5.

Detail of the worn mosaic floor along the northwestern edge of Room 2 and plastered channel leading from Room 2 to Room 3. (Photo by L. Rodriguez.)

In Room 3, at the base of the eastern wall, the channel branches into two: to the north, it leads into a small cylindrical basin, while to the west, it steps down into a neatly built semi-circular, recessed compartment before leading into a large vat (Fig. 6). This vat, with marble veneer preserved on a small set of three stairs at its southeastern edge and imprints of marble veneer tiling across its base, was an original component within the Small Bath’s cold room, perhaps a plunge pool.Footnote ²³ While the stairs were retained for its new function, a small circular depression was cut roughly at the center of its base, and the interior walls of the vat were lined with a ca. 2 cm-thick smooth coating of opus signinum. This waterproof coating appears to overlay the veneer impressions, suggesting that it was a later addition, while it does not seem to extend down into the small central depression. It is possible that a stone bowl was once inserted here: reused marble basins were utilized in a similar manner elsewhere, for example, in Late Antique Delos.Footnote ²⁴ It remains unclear whether the smaller basin in the eastern corner of Room 3 and the semi-circular compartment between the channel and the large vat were original components of the room in its primary phase as a bathhouse. The compartment might have been used as a seat for those in the plunge pool, while the small basin could have held water. The former, however, is neatly aligned with the channel, which might suggest that it, too, is a later addition, while the small basin cuts the room’s original wall facing and employs an identical construction technique to the secondary agricultural features. There is also distinct degradation of the signinum plaster coating directly under the compartment on the larger vat wall, perhaps an indication of more regular or consistent contact with a liquid flowing down this wall from the channel. Opus signinum of identical composition to the interior walls of the vats coats the floor of Room 3, forming a smooth waterproof working surface for the entire room. A small, raised wall of roughly cut stone with marble veneer on its southern face runs along the southern side of the room, where its wide entrance opens onto the Hall. Combined with Room 3’s threshold mosaic, which seamlessly abuts the marble facing of the southern side, this element probably belongs to the original phase of the room. Its function would have been as a splash guard, preventing water from spilling into the Hall, a purpose it continued to serve in the room’s wine-producing phase.

Fig. 6.

Room 3 with branching channel leading into large and small vats, of which the larger is accessed by a set of stairs. Imprints of veneer tiling are visible at the base of the large vat. (Photo by L. Rodriguez.)

The Hall presents an altogether different set of modifications. It was divided into two with the construction of a northeast–southwest wall spanning its width, joining the eastern side of the entrance to Room 2 to the long southern wall of the Hall. Excavations in the space created to the east of this wall revealed a stone rotary mill of the trapetum type usually identified with olive processing (Fig. 7). Its millstone (orbis) was found completely intact, along with half of the mill basin or mortar. A repurposed column, positioned upright and elevated by a square stone plinth, was also found in this space. Directly to the northwest of this column fragment, a rectangular opening (0.58 × 0.39 m) was constructed through the new wall spanning the Hall, reusing three large, worked architectural blocks, the biggest of which formed a lintel, with two smaller blocks laid vertically on each side (Fig. 8). The interior lower half of the northern vertical side block appears to have been worn down considerably more than its upper section, while a small, dovetailed recess is carved onto its western face. There are also signs of chipping, perhaps from repetitive wear, on the blocks that form the bottom edge of the opening on its western side. The aligned placement of the column and wall opening suggests that they functioned together.

Fig. 7.

Rotary trapetum mill stone (a) and base (b) for olive crushing found in the eastern Hall. (Photos by E. Dodd.)

Fig. 8.

Rectangular opening to hold a press beam constructed in a new wall spanning the Hall, positioned directly over the stone press bed, shown here in situ. (Photo by L. Rodriguez.)

The remaining space of the former Hall, to the west of the new dividing wall, revealed further components related to the processing of agricultural products. A large stone press bed was located immediately beneath the opening constructed in the dividing wall (above), mirroring the placement of the column on the other side (Figs. 8–9). Originally an architrave with fascia, the press bed repurposes this marble architectural block and has a circular or slightly pear-shaped channel (canalis) carved into its upper surface. Five additional straight channels ensured liquid was conveyed efficiently off the press bed: four evenly spaced across the front, and a central channel drawing from directly under where the press baskets (fisci) would have been placed. Remarkably little space separates the canalis from the limits of the recycled block, illustrating the extent to which available materials were being repurposed at this time, or perhaps showing that producers were hoping to utilize press baskets with the widest possible dimensions and therefore maximize output. Approximately 1.75 m to the west of the press bed, another large reused architectural block was located, sitting in alignment with the column, opening, and press bed. Originally used in an architrave or frieze course, the block had been re-carved into a monolithic press counterweight (Fig. 10). On its upper face, a dovetail mortise with central socket betrays its use in a lever and screw press system. The counterweight was found in situ, standing upright in a large cut through the original mosaic floor of the hall, which had been created to place it at a lower level, thereby facilitating the action of its lever and exerting greater downward force.

Fig. 9.

Marble stone press bed, with pear-shaped channel, carved into a repurposed architrave with fascia. (Photo by E. Dodd.)

Fig. 10.

Marble counterweight with dovetail mortise and central socket, used in a lever and screw press system and reusing an architectural block. Shown in situ, placed within a large cut through the earlier mosaic floor of the bath, allowing it to sit below ground level and exert greater force. (Photo by E. Dodd.)

Wine and olive oil production in the Small Bath

The architectural modifications and equipment inserted into these rooms find comparanda across the Mediterranean in production installations for wine and olive oil. Before presenting possible interpretations of the function and chaîne opératoire of this facility, it is worth noting several details which help to refine our understanding of its use.

The discovery of the trapetum mill at the eastern end of the Hall is a strong indication that olive processing took place in the repurposed Small Bath.Footnote ²⁵ Olive crushing was seen by Roman agronomists as a distinct step in the production of oil, and mills are often found in separate rooms, usually adjacent to pressing areas.Footnote ²⁶ The mill’s location on the eastern side of the newly inserted dividing wall conforms to this setup. The mill basin (or mortar) is fractured, with just over half remaining; however, it preserves its curved milling surface and the raised central socket with a square inset mortise where a vertical wooden beam would have been secured and around which the millstone(s) rotated.Footnote ²⁷ The basin is neatly worked, with a flat upper surface suggesting that the full height of its internal grinding area has been preserved (ca. 0.15 m). This appears quite shallow, more so than other examples at Antiochia,Footnote ²⁸ indicating that it was not designed to mill large quantities of produce simultaneously and perhaps suggesting a smaller scale of production. The dimensions and neatly squared nature of the mill basin might mean that, like the counterweight and press base, it too was carved from a recycled architectural block. Only one millstone was recovered with the base, and its relatively large size relative to the basin implies that only one stone was used in this system.Footnote ²⁹ It was carved from a single block, rounded on all edges, and is flat on one side, convex on the other, with a large, almost square mortise carved through its center. In a standard trapetum setup, this was where a horizontal wooden beam was affixed to the vertical beam rising from the mill basin, extending beyond the millstone and acting as a handle for people or animals to power its rotary operation. Presumably, the flat side faced inwards and the convex side outwards, which encouraged more contact with the curving basin interior and helped stop olives being pushed up and out of the mill.Footnote ³⁰ While trapeta are typically associated with olive processing for oil production, designed to separate the olive flesh from its pit and begin the oil extraction processFootnote ³¹ – indeed, Cato (Agr. 20–22) provides a detailed description in this context – it is possible that they were also used to crush other produce (see below).

In the western section of the Hall, the press bed and counterweight also indicate the production of olive oil or wine. Despite arguments that the presence of a stone press bed is definitive proof of olive oil production, this is not a reliable method of distinguishing between oil and wine production. Even in very early examples, organic residue analysis (ORA) of samples from stone press beds reveals indicators for the presence of wine and, therefore, their likely use to process grapes as well as olives.Footnote ³² The enormous quantity of stone press beds found across the Mediterranean also suggests that some balance should be sought between their use for olive oil and wine, or that, perhaps, they were used for both commodities.Footnote ³³ More reliable might be attempts to identify indicators of oleic acid erosion, to which some attribute the radial markings often found on stone press beds;Footnote ³⁴ however, this approach, too, requires further work, including comparative experimentation on different types of stone using not just oleic acid, but also acids found in grapes and wine. There is relatively little trace of erosion on the Small Bath press bed, aside from faint signs on the interior side of the canalis. Helped by the durable nature of its reused marble architrave block (above), it is completely intact, measuring 1.28 ×1.03 m and 0.3 m in height. Despite substantial recent work focusing on press components,Footnote ³⁵ there remain few identifiable regional or chronological patterns for press beds (counterweights are slightly more revealing, below). The five parallel outlet channels that characterize this piece find few parallels elsewhere; the slightly pear-shaped canalis is more common, however, finding roughly similar comparanda elsewhere in Anatolia, Greece, and Lebanon.Footnote ³⁶ The Small Bath press bed is fairly rudimentary in its carving compared to finely worked examples elsewhere (e.g., Delos, Histria, Dalmatia, Italy), which, it has been suggested, were used in perfumeries. This might provide an indication of the level of craft expertise and knowledge available in Antiochia at the time.

The counterweight (0.51 × 0.55 m; 1.24 m in height) has a half-dovetail mortise, which extends across the entire width of its upper face, with a central screw socket. This precise counterweight type is not overly common across the Mediterranean; however, a dense cluster of them can be found in Lebanon and Palestine (Waliszewski’s type G.3, or Frankel’s Din’ila weight), with other comparable weight stones in North Africa (e.g., at Volubilis), leading to suggestions that the type might be linked to a regional or cultural preference associated with the Phoenicians.Footnote ³⁷ It was used in lever and screw press systems, for either wine or olive oil, which were increasingly popular from the 1st c. CE onwards and came to dominate the Late Antique eastern Mediterranean production landscape. A large wooden beam was attached to a vertical screw thread, which was in turn anchored into the counterweight by a wooden tenon secured in the mortise (Fig. 11). It is unknown whether the counterweights were designed to remain on the ground or were instead lifted up as the screw was turned, thereby placing even greater pressure on the grapes or olives. Regardless of its exact operation, the designers of this facility were aware of the need to place the counterweight at a lower level than the pressing area, making a large cut in the bath’s existing mosaic floor to accommodate this.Footnote ³⁸

Fig. 11.

Visualization of a lever and screw press that operates using a similar counterweight type. (Drawing by S. Brazeal, after Frankel Reference Frankel, Frankel, Avitsur and Ayalon1994, fig. 46.)

Further evidence for the presence of a lever and screw press system can be drawn from the window in the northeast–southwest wall built across the hall and the upright column section found to its east. The neat construction of the window from three large, worked stone blocks suggests that it needed to be durable, and we suggest that it once housed the fulcrum end of the lever (prelum) used in the screw press. Windows or niches to hold press beams in this manner can be found in Greece, Spain, Anatolia, and Syria-Palestine.Footnote ³⁹ Moreover, the reused upright column immediately behind the window may have served as a “backstop” for the lever beam, and the signs of wear on several of the window blocks further illustrate the repetitive motions of a lever.Footnote ⁴⁰ The press system as we reconstruct it here therefore runs northwest to southeast along the length of the existing hall space, using the aligned counterweight, press bed, window, and column.

Pressing and crushing operations appear to be restricted to the two sections of the hall space, while in Rooms 2 and 3, the arrangement of floors and vats indicates other production activities. The combination of a worn mosaic floor in Room 2 connected by a channel to several vats in Room 3 resembles arrangements at wineries elsewhere in the eastern Mediterranean. Particularly heavy areas of wear towards the channel opening, however, might suggest that, along with treading, rolling actions were also used. This has been suggested for rock-cut facilities, especially in North Syria and the Levant,Footnote ⁴¹ due to the production of particular regional wine types from sun-dried or raisined grapes, which require a harder stone crushing or rolling surface to extract their juice. Similar arguments have been made for Rough Cilicia, a region known for its production of a particular type of raisin wine called passum, and it has previously been suggested that this wine was produced at Antiochia and exported in LRA 1 or Zemer 41 amphorae.Footnote ⁴² The rooms of the Small Bath also contained numerous niches, perhaps for small statues in its original phase. On the west wall of Room 2, one of these sits directly above the channel that conveyed liquid into Room 3 (Fig. 12). It is tempting to think that even in the building’s secondary phase, as a production facility, this niche continued to be used for the image of a deity or other ritual object, blessing the freshly extracted grape must and calling for bountiful and safe production. Embedded aspects of ritual can be seen through Bacchic frescoes in Vesuvian production spaces or early Christian iconography and inscriptions on press components on Delos.Footnote ⁴³

Fig. 12.

Room 2 facing west, showing the niche, perhaps to hold a statue or lamp, sitting directly over the channel that conveys must from Room 2 to 3. (Photo by E. Dodd.)

No traces of roof collapse were found to suggest the bathhouse was in a state of architectural disrepair by the time the agricultural facilities were inserted. It was probably still largely standing and roofed. Indeed, a section of intact vaulted roof with tegulae remarkably in situ was excavated over the northern end of the building’s entrance corridor (Fig. 13). The cistern just above the complex that originally provided water for the bath also continued to serve a crucial hydrological function, storing and conveying water essential for oil production and for cleaning the facility as a whole. Reusing standing and intact structures was a deliberate choice by those who constructed this facility, and it provided specific benefits in the context of agricultural production. The bathing complex was near a natural water source and had existing storage infrastructure; the roofed rooms would also have encouraged shaded production spaces and better ambient temperature control for the hotter months of the vintage and colder season of olive oil production.Footnote ⁴⁴ The particular characteristics of a bathhouse might have made it ideal to repurpose for wine or oil production. Further east along the Cilician coastline, the Harbour Baths at Elaiussa Sebaste were also transformed into an oilery.Footnote ⁴⁵

Fig. 13.

Excavated stretch of intact tiled and vaulted roof from the bathing complex, over the northern end of the entrance corridor. (Photo by L. Rodriguez.)

It is worthwhile considering several possible interpretations for the use of these spaces in agricultural processing: 1) wine and olive oil were both produced, in distinct areas; 2) the entire facility was devoted to grape processing and wine production; or 3) the spaces were dedicated to olive processing and oil production.Footnote ⁴⁶

The first conclusion might seem the most obvious. In this scenario, Rooms 2 and 3 were used for wine production – the former for treading and crushing grapes, and the latter for collecting and fermenting must – while the Hall served to produce olive oil using the lever and screw press, press bed and trapetum. Previous scholarship argues that vats of this large size were more often used for wine production than for oil, owing to the greater quantity of liquid expressed from grapes, though this is by no means a perfect criterion and large vats are known from oileries elsewhere.Footnote ⁴⁷ The large and small vats in Room 3 may have collected different treadings or qualities of grape must; perhaps the comparatively small vat was used for the initial juice (either prototropum or mustum lixivium), which created the highest quality wine in Roman antiquity.Footnote ⁴⁸ The two vats were served by the split channel, and must could easily have been redirected simply by blocking one side or the other. The large mosaic floor and waterproofed edges of Room 2 were ideal for treading the substantial quantities of grapes that would have been necessary to fill the large vat, while increasing traces of wear nearer to the channel might indicate the use of stone rollers to crush raisins in passum production (above), or simply repetitive movements associated with consolidating processed produce and liquids into the channel for its collection in the adjacent room.

The trapetum is a strong indicator for the processing of olives in the eastern section of the Hall, with the pulp then brought around the dividing wall and directly onto the nearby press bed. The mosaic area immediately west of the press bed was cut in a circular shape, presumably to accommodate a ceramic or stone collection receptacle, such as those found nearby at Syedra in an olive oilery (Figs. 14–15). A circular stone installation located southwest of the press bed may have served as the masonry base of an olive mill (Fig. 16), also found at the facilities of Syedra and at Pirano in Istria in Late Antiquity, or a platform for a stone-built storage container, of the kind seen at Burgaz in western Türkiye.Footnote ⁴⁹ If this was the base for a mill, it appears too large to have been used in conjunction with the trapetum that was recovered from the eastern space of the Hall. Instead, one of the much larger mill basins found in fragmentary form around the Small Bath may instead have been placed on its top: for example, the relatively intact example documented just uphill from the Small Bath, or smaller fragments found both on the south slope and incorporated into the apse of Room 8. A later wall that separated this stone installation from the press bed complicates interpretations of production flow, requiring more refined ceramic dating to understand whether these structural features belong to the same phase.

Fig. 14.

Large round cut through the earlier mosaic floor of the bath, where a portable ceramic collection receptacle was placed adjacent to the press bed. (Photo by L. Rodriguez.)

Fig. 15.

Inset ceramic jar acting as a collection receptable below the press bed at one of the olive oil processing facilities at Syedra, Türkiye. (Photo by L. Rodriguez with permission of Ertuğ Ergürer, Syedra Ancient City Archaeological Excavations.)

Fig. 16.

Circular stone masonry feature found constructed on the earlier mosaic floor of the bath, perhaps acting as the base of an olive mill. (Photo by L. Rodriguez.)

More problematic is a lack of evidence of decantation systems for the separation of the olive oil from its vegetable water, which was an essential part of the process, but portable containers may have been used, then removed once the facility was abandoned. It remains possible that the press was multi-purpose, for both wine and olive oil production. After all, the vintage and the olive harvest fall at different times of the year, and farmers could, and still do, manage the cultivation and processing of both commodities.

The second and third possibilities are less likely, though they merit consideration given persistent difficulties in ascribing production installations to a specific commodity. It is possible that Rooms 2 and 3 were used in the manner described above, but that the trapetum and press were also used in wine production. If more durable sun-dried or raisined grapes were being processed, perhaps a mill was required to crush grapes before treading or pressing, facilitating the release of juices. The press could well have been used in wine production even if olive oil was also produced, but it is equally conceivable that it was inserted even if the entire facility was devoted to vinicultural enterprise alone. Indeed, the other known wine production installation at Antiochia also used a lever press.Footnote ⁵⁰

Conversely, signs of crushing or rolling on the mosaic floor of Room 2 could perhaps support an interpretation of olive processing. The use of stone rolling devices to crush olives is known ethnographically and has been suggested for archaeological contexts in Anatolia and Syria-Palestine.Footnote ⁵¹ Olive oil and vegetable water (amurca) would then have been captured in one of the vats in Room 3. The dual vat system could have allowed smaller quantities of free-run oil and amurca to be collected in the small vat, and, following this, wastewater that resulted from washing the olives to be collected in the large vat before milling and pressing. It is also possible that Room 2 was used simply as a holding space for harvested olives before processing. Such facilities, whether specific rooms, floors, or vats, are noted in ancient sources, with recommendations for the collection of the free-run amurca, which could be used for certain secondary purposes.Footnote ⁵² Spaces in several Italian complexes have been interpreted in this manner.Footnote ⁵³ We might even consider the signs of wear on the mosaic tesserae as a product of oleic acid erosion or, as suggested above in the case of wine, the result of pushing fruit and juices towards the channel. Other storage spaces like this might have been used to wash olives before milling and pressing.Footnote ⁵⁴ Alternatively, the system of two differently sized vats could also have been used for the collection of oil of varying qualities from subsequent crushings or pressings. Lower-quality oil was typically released in second or third pressings, after heated water was poured over the already pressed pomace to encourage the release of more oils, resulting in the collection of a greater volume of liquid. However, this would rely on a connection from the press bed in the Hall to the vats in Room 3, for which there is currently no evidence. Perhaps more plausible is the use of the dual vats in decantation or separation of oil: the mixture of oil and amurca could be collected and left to settle in the large vat, after which the capulatore (a specialized oil separator) skimmed the oil into the adjacent smaller vat for further clarification. Vats of different sizes and shapes, and with no interconnection via over- or underflow, have previously been interpreted in this manner, with it even being suggested that an adjacent seat at the San Rocco villa in Francolise (Italy) was used by the capulatore.Footnote ⁵⁵