The Mills-Bakeries of Ancient Ostia and the Distributions of Free Grain

By Dr. Jan Theo Bakker / 05.05.2000
Professor of Archaeology
Leiden University


Horrea for grain – light blue; dolia defossa – dark blue; other store buildings – yellow; bakeries – purple; isolated finds of millstones and kneading-machines – green circles

Each archaeologist studying Ostia is acutely aware that the ruins represent the situation in late antiquity and the early Middle Ages. This situation was in great contrast with the developments in the second century, when Ostia was flourishing like never before or after.

During its heyday Ostia must have had some 40.000 inhabitants (depending on the number of floors one wishes to reconstruct). In the third century – a period of deep crises – this number must have decreased considerably. Commercial buildings along the Tiber and spacious apartments elsewhere in town were abandoned and left to collapse. The commercial emphasis shifted to Portus, which was Ostia’s harbour district, situated a few kilometres to the north. Portus even became an independent city from the reign of Constantine. This however was not the end of Ostia. The city became a pleasant seaside living environment. Several commercial buildings were modified drastically and changed into wealthy habitations. In the early fifth century the final decline did set in. Buildings were looted, marble was taken to lime ovens, and metal objects were melted down. Flimsy walls, found and often removed by the excavators who worked in the first half of the 20th century, testify to impoverished habitations.

Obviously then the number of bakeries that can be recognised today is much smaller than the total during Ostia’s golden age. When the population shrunk, the number of bakeries was reduced accordingly. The discarded bakeries were sometimes reused for other purposes, and sometimes left as they were. It should also be remembered that only some two-thirds of the city have been excavated.

The Appearance of the Mills-Bakeries

Roman bakeries were usually mills-bakeries. Only in later antiquity can we make a distinction between the milling and baking, namely after the introduction of water mills, such as have been found on a steep slope of the Ianiculum in Rome and near Arles (the Barbegal). Until the mid-1980’s two of these mills-bakeries had been identified in Ostia: the House with the Millstones (I,III,1), to the east of the well-known House of Diana, and Mill I,XIII,4, in the south part of town.

A Dutch team studying the bakeries was then able to identify five or six more. One of these establishments, near the museum, has disappeared completely, and is known only from the excavation diaries. The identification of Building II,VIII,9, to the east of the Great Warehouse (II,IX,7), remains uncertain. Only the south part of the building has been unearthed, and further excavation may show whether it was indeed a bakery. No doubts remain about the function of the House with the Ovens (II,VI,7), House I,IX,2, the House with the Wooden Balcony (I,II,2.6), and finally the House with the Cistern (I,XII,4).

All of these bakeries have a lot in common, and the technical achievement was quite considerable. The workshops are quite large, and cover a ground floor area ranging from 640 to 1525 square meters. On average the bakeries contained nine millstones.

Reconstruction drawing of a millstone. / From La construction romaine, matériaux et techniques, by Jean-Pierre Adam 1984, fig. 735.

A millstone consisted of two parts: an immobile, conical base (meta) and on top of that a stone that was shaped like an hour-glass (catillus). Mules or horses were attached to a wooden frame over the catillus. They walked in circles and rotated the catillus over the meta. The grinding took place between the two parts, that were at a very small, fixed distance. If the distance was too small, the grain would have been burnt, and if it was too large, too much bran would have remained. Specialist carpenters maintained the machines. Many of their tools were actually found in the House with the Millstones, where they also had a small workshop, the shop-sign of which can still be seen in the east facade. Inside the millstone dosage cones were used.

Drawing of a dosage cone (left) and of the use (right). / From Bauten und Katapulte des römischen Heeres, by Dietwulf Baatz 1994, fig. 20,12 and fig. 14.

Machines were also used for the kneading. Like the millstones they were made of porous volcanic stone. They are bowls in which the dough was kneaded by a combination of fixed and rotating blades. A few blades were inserted in the side of the bowl, and a few were attached to a vertical bar. Slaves or animals turned the vertical bar.

Cross-section of a kneading-machine.
From Technologie und Terminologie der Gewerbe und Künste bei Griechen und Römern, by Hugo Blümner 1912, fig. 26, p. 65.

Huge, well-ventilated halls were built for the machinery. Often the first floor formed part of the workshop as well. Grain was led from there to the millstones through wooden pipes. The floors suffered a lot, and were therefore covered with basalt blocks, in which imprints of hooves remain. Many basins are found, because water was needed in very large quantities, for the kneading, as drinking water for the animals, for moistening the grain before milling and so on. The bread was baked in huge ovens, consisting of a base with a cupola, that had an inside diameter of 3.5 to 5 meters. Wood was burned inside the cupola, as in modern pizza ovens. The ovens could contain rotating grates on which the bread was placed.

A marble block in which ollae were inserted, found in Ostia, now in the Vatican. It was made by P. Nonius Zethus, Aug(ustalis), for himself, the freedwoman Nonia Hilara and his wife Nonia Pelagia. To the left and right of the inscription the interior of a bakery is depicted. First century AD. CIL XIV, 393. / Photograph by Bill Storage.

Detail of the relief on the block: a millstone. / Museo della Civiltà Romana. Photograph by Jan Theo Bakker.

Detail of the relief on the block: various objects used in a bakery. / Museo della Civiltà Romana. Photograph by Jan Theo Bakker.

The Ostian bakeries may be called factories, especially when they are compared to their counterparts in Pompeii. The latter contained only three to four millstones, very close to each other. The spacious Ostian bakeries must have been operated in a much more efficient way and more hours per day. However, waterpower was not used for the milling in Ostia. Many water wheels have been found in the city, but they all supplied baths. Rough and admittedly tentative estimates can be made of the production capacity of the millstones. A millstone in Pompeii must have supplied some 90 people, if it is true that the city had approximately 10.000 inhabitants. A water-powered millstone could serve between 320 and 690 people. An Ostian millstone may have produced flour for some 150 to 300 people. This means that, in the second century, there must have been approximately 178 millstones in at least 20 bakeries. In the identified bakeries must have been 60 to 65 millstones.

The Sanctuary of Silvanus

For the study of the bakers’ trade in Ostia we possess various other sources, that are quite informative: reliefs, inscriptions, legal texts, and in particular a shrine with wall paintings and graffiti. This shrine is a narrow and dark room in the back part of the House with the Millstones. It could only be entered by passing through the entire bakery.

It was excavated partly in 1870, when fifty bronze and silver statuettes were found. Unfortunately many of these were stolen during the excavation, and we do not know which deities they represented. The bakery and the shrine were excavated completely from 1913 to 1916. It became clear that the bakery had been destroyed by fire and had not been rebuilt (coins and the masonry suggest that the fire occurred at the end of the third century). As a result, many objects were unearthed. Some were related to the workshop, such as parts of horse harnesses, bells that rang when the hopper was empty, and metal parts of dosage cones for the millstones. Other finds were related to surprisingly wealthy habitations on the upper floors – bronze revetment of furniture, sometimes with silver inlay, marble and terracotta friezes, fragments of black-and-white mosaic floors and of painted ceilings, tiny columns, marble revetment, and so on.

After the 1870-excavations only a few objects were found in the shrine, but on the walls many paintings were found of people and deities. Most of the figures were identified correctly by the excavator, Guido Calza, but the relationship with the bakery had to our surprise never been investigated.

The main deity in the shrine was Silvanus, a god of the woods. The room was therefore called “Sanctuary of Silvanus” by the excavators. Silvanus was depicted next to the entrance, on the outer wall. He was also painted in the back part of the room, on the right wall. The latter painting, which is now in the museum of Ostia, once had a painted text, stating that it had been made after a vision: Silvanus had appeared to one of the bakers in a dream.

Next to Silvanus one can still read the following graffito: “Calpurnius, night-watchman from the group of lieutenant Ostiensis, from the seventh cohort, during the reign of Caracalla and the consulate of Laetus and Cerialis [215 AD], X”. Calpurnius, a fire fighter from Rome, stationed in Ostia, patrolled through the city at night with a torch. “X” means that he prayed for “ten more years for Caracalla”, as the Americans would say today. On the opposite wall Calpurnius scratched the exact day on which he visited the shrine: April 25 (probably because on that day the Robigalia were celebrated, in honour of the god Robigus, for the prevention of the rust disease in grain).

Graffiti of night-watchmen have also been found in barracks in Rome, in Trastevere. Their work in the sparsely lit streets was quite dangerous, as can be deduced from texts such as “All was safe”, and the apparent need for support by the divine Emperors.

Portrait of Caracalla / Capitoline Museums. Photograph by Jan Theo Bakker.

A thin wooden partition wall created an anteroom. Here some horses and the Dioscures were depicted. In the shrine proper is an altar, in front of some niches. On the floor is a black-and-white mosaic of a person who is about to kill a sacrificial animal. Parallels for this figure are found in the shrine for the Imperial cult in the Barracks of the Fire Brigade.

On the right wall, next to Silvanus, are the scanty remains of a figure with a lance or sceptre. On the left wall we see a whole row of figures: Augustus, Harpocrates, Isis, Fortuna, Annona, a figure with a cornucopiae (a Genius? Serapis?) and Alexander the Great. From the location of Calpurnius’ second graffito can be deduced that all these figures had been painted before April 25, 215 AD. They are to be understood as follows. Augustus and Alexander the Great are present as illustrious predecessors of the Emperor, Caracalla, son of Septimius Severus. In 214 AD, at the age of 26, he had left for Egypt. During his journey he was overcome by an Alexander-mania, which several ancient historians have described. He ordered that throughout the Empire depictions should be made that presented him as a new Alexander. The Ostian bakers obeyed. They also thanked the Emperor for his distributions of free grain, personified by Annona. Isis (here, as so often, associated with Fortuna) and the child-god Harpocrates refer to the large quantities of grain that were imported from Egypt, which was an Imperial domain. The Dioscures were worshipped in Ostia as protectors of shipping, in this shrine as protectors of the grain fleet that transported the grain from Alexandria to Ostia and Portus.

Distributions of free grain

Does the presence of Annona in the shrine point to distributions of free grain in Ostia, or is it merely a reference to the well-known distributions in Rome? Each month 150.000 to 200.000 inhabitants of Rome received five measures of grain, a gift by the Emperor. This was not a system of relief for the poor, but a way for the Emperor to strengthen his ties with the populace. So far few historians have wondered how bread was made of this grain. It seems out of the question however that the recipients, often living in rented apartments, made their own bread, if only because of the fire hazard. The recipients could of course have made their own arrangements with bakers or people operating ovens, but this does not seem to have been the case. The Dutch legal scholar Sirks, member of the team, has shown that the Emperor entered into contracts with bakers for this purpose. The bread in Juvenalis’ famous sneer “bread and circuses” (Satires X, 81) is to be taken literally, it is not a poetic description of grain: the Emperor did not leave the milling or baking to the recipients of free grain.

Originally the Emperor had to draw up contracts with individual bakers, such as M. Vergilius Eurysaces, whose enigmatic tomb – in my view quite possibly a folly – can still be seen outside the Porta Maggiore in Rome. Trajan simplified the system and devised a body or guild (corpus) of bakers in Rome, with which he did business. From now on a single contract sufficed. Membership was not compulsory. The Emperor just needed enough members for his own purposes: the feeding of his slaves and personnel, and of the fixed number of recipients of free grain.

The grain was taken to warehouses owned by the Emperor – either by the recipients themselves or by porters -, and here the bakers collected it. The recipients paid a baker a fixed amount for the milling and baking, but nothing for the grain. In other words: they received cheap bread, rather than free grain. Arrangements must also have been made for the Imperial slaves, and we know that the Imperial fire fighters received free grain as well.

Ostia is the only city other than Rome in the western half of the empire in which a bakers’ guild is documented. It was probably instituted during Trajan’s reign. During the reign of Commodus the prefect of the grain supply, residing in Rome, bestowed favours upon it. Many Imperial slaves worked in the harbour, and fire fighters from Rome were stationed in the Barracks of the Fire Brigade (our Calpurnius was one of them). It seems most likely that the Ostian guild baked bread for these people, on the basis of contracts with the Emperor. But did Ostia have distributions of free grain as well? The study of the bakeries by the Dutch team provided some answers.

The dates and distribution of the bakeries

Region V: Horrea for grain – light blue; dolia defossa – dark blue; other store buildings – yellow; bakeries – purple; isolated finds of millstones and kneading-machines – green circles

More than half of the buildings in Ostia were erected during the reign of Hadrian, shortly after the completion of Trajan’s harbour. In this period many bakeries must have been built, but of the identified bakeries only one (Mill I,XIII,4) can be assigned to it. The other workshops were installed at various points in time from the reign of Antoninus Pius to the Severan period, that is, during the next 100 years. Often Hadrianic buildings were altered for the purpose. This is a most surprising development. For example, the House with the Ovens is situated to the west of the Barracks of the Fire Brigade and most likely served the fire fighters. The Barracks were built under Hadrian, but the bakery was installed later, during the reign of his successor, in Hadrianic shops. Apparently the bread for the fire fighters was originally prepared elsewhere in town.

Equally surprising is the distribution of the bakeries. There is a concentration in the centre of town, especially to the east of the Forum. Exceptions are the House with the Ovens, the position of which was probably dictated by the Barracks, and Mill I,XIII,4, the only bakery from the early second century. Why were so many bakeries – noisy workshops, where slaves and convicted criminals toiled in flour dust – located and tolerated near the representative and monumental centre of town? Two bakeries even flank the Decumanus Maximus, a clear disfigurement of Ostia’s main street. It causes little surprise that this situation was corrected much later, when these two bakeries had been abandoned: a large, decorative apse and a nymphaeum were then installed in rooms along the Decumanus.

The Great Warehouse

The explanation for the distribution is found quickly. The bakeries in the centre of town are near the Great Warehouse, one of Ostia’s largest warehouses for grain. We seem to witness rationalisation, economic simplification of the bakers’ trade. Apparently it was more practical or cheaper if the bakeries were situated near their main raw material. The building dates back to the reign of Claudius and Nero. It was rebuilt and extended during the reign of Commodus and the Severan Emperors. Eventually it could hold between 5.660 and 6.960 tons of grain, enough to feed at least 14.000 people for one year. The exceptional importance of the building is indicated by the name of the street to the west – the only street in Ostia of which the ancient name is known. In a city full of warehouses it is significant that precisely this street was called Semita Horreorum, i.e. Crossroad of the Warehouse.

This road started near the Tiber, ran between the House with the Millstones and the Great Warehouse, and continued towards the south, along the east side of Mill I,XIII,4. The latter bakery, with its exceptional date and location, is in this way connected with the cluster of bakeries.

After the reign of Hadrian one bakery after the other was installed near the Great Warehouse: the House with the Wooden Balcony, the House with the Cistern, a bakery near the museum, the House with the Millstones, and perhaps a bakery to the east of the warehouse. The warehouse probably also supplied Mill I,XIII,4 and the House with the Ovens (another large is situated near the latter bakery, but was built only after its installation). Some 54 millstones may have been filled with grain from the Great Warehouse. If our suggestion is correct that an Ostian millstone produced flour for approximately 225 people, then the total production would have been sufficient for some 12.000 inhabitants.

The size, quality and complexity of the Great Warehouse suggested to Rickman that it was Imperial property. The investigation of the House with the Millstones confirmed this idea. This bakery – the one with the shrine where Caracalla was thanked – had a direct connection with the warehouse, through a roof over the Semita Horreorum. A further argument is the tremendous size of the original warehouse. In the first century Ostia had few high-rising buildings, and must have been similar to Pompeii. The colossal building must have made quite an impression. It seems to have been built very consciously near the Forum, thus stressing the generosity of the Emperor.

The bakeries that were supplied by the warehouse must have been owned by members of the Ostian bakers’ guild. The date of installation of the bakeries presumably reflects the enrolment of new members over an extended period of time.

Distributions of free grain in Ostia?

There are good reasons to think that the members of the Ostian bakers’ guild did not just bake bread for Imperial slaves and personnel. They seem to have worked for recipients of free grain as well. The location of the Great Warehouse and the dependent bakeries simply cannot be understood if we suppose that these buildings supplied only Imperial slaves and a few hundred fire fighters. The bakeries swept over the area near the Great Warehouse like a dark cloud. We already mentioned the disfigurement of the Decumanus. The nature of the area between the House with the Millstones and the Capitolium changed dramatically. To the north-east of the temple a block with a large mansion, several attractive rented apartments and a large garden had been built (I,IV). The mansion was now converted to a hotel annex brothel. In the north part of the garden a warehouse with large vessels for oil was installed. Bars and a restaurant were built along the road connecting the Great Warehouse and the temple. Many examples can be found in Ostia of the profound influence of workshops and other commercial buildings on the cityscape, but nowhere is this influence as strong as here, the result of the Imperial involvement.

We should also note that Mill I,XIII,4, on the Crossroad of the Warehouse, is situated in the south part of town, far from the harbour district, where the Imperial slaves worked. Finally the supposed number of people served by the bakeries (12.000 to 14.000) implies an extremely high number of Imperial slaves in Ostia. Perhaps one-third of the population of Ostia consisted of slaves, as in Rome, that is some 13.000. However, many of these were not owned by the Emperor.

The presumed distributions may have been started by Claudius, who spent a lot of time in Ostia, was responsible for the first harbour at Portus, and during whose reign the Great Warehouse was built.

The final publication of the Dutch team included a full description and analysis of three Ostian bakeries. The remaining workshops were studied in a concise manner. It is our hope that others will take up our work, compile detailed catalogues of the remaining bakeries, and verify our conclusions.

Small millstones, stored in the Piccolo Mercato. / Photograph by Jan Theo Bakker.

Individual Bakeries

Caseggiato dei Molini (Regio I – Insula III – I,III,1)

Click Image to Enlarge / Red: Hadrianic. Yellow: Antonine. Brown: Severan. Grey: undated and third century. Legend: see below.

The excavation of the House of the Millstones was completed in the years 1913-1916, by Raffaele Finelli and Guido Calza. The building had been excavated partially in 1870 by Carlo Ludovico and Pietro Ercole Visconti. Because of political events the Visconti’s could not finish the excavation. Ostia was property of the pope, but on September 20th 1870 king Victor Emanuel captured Rome. The pope retreated to the Vatican, and Ostia was from now on under the authority of the new Italian state. After the first excavations the building was named Casa di Aquilina, after a stamp on a lead water-pipe, found below the pavement of Via dei Molini. When the excavations were continued in 1913, most of the rooms that had previously been excavated had been filled with earth again. Objects found in the building were stolen twice: in 1870 and 1959.

Phase 1: the Hadrianic period

The building was erected in the early second century AD, during the reign of Hadrian (opus mixtum). It consisted of shops along Via dei Molini, a hall at the intersection of Via dei Molini and Via di Diana (1-3), and two halls (17-18, 19-21) and a few rooms behind the shops. In the east facade two recesses for reliefs have been preserved. One is now empty, in the other is a terracotta relief of a Genius with cornucopiae and patera, and a snake. The snake represents the Genius Loci, the protective deity of the place.

Phase 2: the Antonine period

In the Antonine period the House of Diana was built to the west. There was an alley between the two buildings. In the same period the House of the Millstones was modified (opus latericium). Shops 15 and 16 were built or perhaps rebuilt. Slender brick piers were set against the west and east wall in halls 17-18 and 19-21. They do not seem to have supported a roof. Room 17 seems to have been the only part of the halls that was covered, witness beamholes. Room 8 apparently had a special status. It was not accessible from the street, but received light from it through two windows. Two more recesses for (lost) reliefs were added, in the east and south facade. There are two external staircases: room 4 (already a staircase in the Hadrianic period) and room 9. The steps rested on wooden beams, fastened in small holes in the lateral walls. The first floor in the east part of the building was at a height of a little over four metres. Via di Diana was partly blocked by a short wall set against the east end of the facade. There is a pendant across the street.

Phase 3: the Severan period

Important structural modifications took place in the Severan period (opus latericium). Five sub-phases have been recognized. The modifications began during the reign of Septimius Severus (193-211 AD). The work was finished during the first decades of the third century. The main alteration was the raising of the level of the upper floors. A new first floor was installed at a height of c. 6 metres. The other upper floors must also have been rebuilt. This destabilized the building and counter-measures had to be taken.

The new first floor was supported by arches in rooms 6, 7, 17 and 19-21. The two arches in hall 19-21 are huge. A pier in the centre of hall 17-18 must have carried beams. Staircase 9 was rebuilt. Behind the staircase a small room (11) was created. The alley was closed off at the north end by room 29, which has a barrel vault. Five piers were set against the east facade, and another five opposite these, across the street, against the Grandi Horrea. The piers obviously supported east-west running arches, and stabilized the building.

Phase 4: later third century and undated alterations

Some further alterations have been dated to the later third century (opus latericium and vittatum), other alterations could not be dated. Basins were installed in rooms 26 and 28. Room 27 is spanned by two high arches, that cannot be dated, even though one of the piers is Antonine. The piers further stabilized the building. Another basin was built in room 21. Seven very large windows were hacked out, high up in the west walls of rooms 17-22. The remains of two large ovens were found in rooms 23 and 14. The one in room 14 was given up at an unknown date, and room 14 became an independent shop. In the west wall of room 23 are two holes, low in the wall, apparently draught-holes. In the north wall is the lower part of a niche or, more likely, chimney. Bar-counters were installed in shops 15 and 16. In the east side of the Severan pier between rooms 4 and 6 a relief was inserted with carpenters’ tools: a saw, a plummet, two pairs of compasses, and rulers can be recognized. Water was piped into the building through a lead pipe, from the workshop of Iulia Aquilina, below Via dei Molini and entering the building in room 7. All walls in the building were at one time plastered (a total of seven layers has been recognized). On the east facade the excavators saw traces of red letters on a thin, white background. Between two of the piers on the east side of Via dei Molini two tiny rooms were built, both with a low podium with a few treads, that once supported a ladder.

Drawing of the relief with carpenters’ tools. NSc 1913, p. 206.

The find of millstones, kneading-machines and floors of basalt blocks show that the building had become a bakery. The workshop will be discussed in more detail below.

Reconstruction drawing of the building in the Severan period (Jan Theo Bakker). Seen from the south-east.

The fire and the finds

The building was destroyed by a fire. The excavators saw traces on walls and floors, and found a thick destruction-layer (1.5 to 3 metres high). The fire can be dated fairly accurately. No masonry in the building can be dated to the time of Diocletian or later. Many coins were found. The series ends under Probus (276-282 AD). Apparently the building was destroyed in the last quarter of the third century, perhaps as the result of an earthquake, of which several traces have been found in Ostia. After the fire at least the lower part of the ruins was left undisturbed. On Via dei Molini parts of the walls of the building were found, and on top of these a thick layer of rammed earth: a path had been created at an average height of 2.20 above the Hadrianic street level.

Numerous objects were found in the building, such as keys, locks, revetment of furniture, lamps, candelabra, strigiles, and kitchen utensils. In bar 15 a complete amphora and parts of amphorae were found. Parts of horse-harnesses, a ring for tying up animals to a wall, and bells must be related to the bakery (the bells rang when the hopper was empty). Finelli reports the find of many lance-heads, but these were in fact dosage-cones, that were inserted in the lower part of the millstone (meta). Grain flow-control was achieved by suspending a hopper over the conical end of the cones. Many tools were found: hammers, hatchets, chisels, axes etc. These must have belonged to the carpenters, whose presence in the building is attested by a relief. Presumably they had specialized in the maintenance of the millstones and their wooden frames. The distance between the lower and upper part of the millstones was of great importance: if it was too big, the grain would fall through; if it was too small, the grain was burned.

A number of objects had fallen down from one or more surprisingly wealthy apartments on the upper floors: parts of bronze revetment of furniture were found (sometimes with silver inlay), elaborate candelabra, fragments of marble and terracotta friezes with amorino’s, parts of black-and-white mosaic floors with geometrical patterns and floral motifs, fragments of painted ceilings, parts of opus sectile, and tesserae of glass-paste.

In the south-west part of the building a shrine was found, the Sacello del Silvano (room 25), that is described on a separate page. But from the finds can be deduced that there was at least one more shrine in the building. In 1870 thirteen statuettes and one small bust of deities were found. During the same excavations a gold ring was found with an aureus of the Emperor Decius (249-251 AD), perhaps related to the cult of the Emperors, and lamps (some from the workshop of Annius Serapiodorus) with depictions of the Pastor Bonus. The precise place of discovery of the 1870-finds is not known. In room 14 part of a vessel with the depiction of a bat (the mammal) was found. The bat may have had an apotropaic purpose. In room 15 a small snake, possibly the Genius Loci, and an approximately 0.60 high marble statue of the armed Venus came to light. Perhaps in room 16 a statuette of Mercurius was found. Eight statuettes were encountered in rooms 17-19. One of these, a scorpion, is perhaps to be related to the cult of Mithras. A tiny octagonal column with base and capital, found in the same rooms, may have formed part of an aedicula. Also from these rooms come seven rectangular bronze sheets, with depictions of signs of the Zodiac and the symbol of a planet (average height 3.8 cm.). They were attached to something by means of nails. According to Floriani Squarciapino they are to be related to magical practices or the cult of Mithras (note that in the mithraeum in the adjacent House of Diana a Petronius Felix Marsus is documented, a snake-charmer, who was thought to have occult powers). In the same rooms a gorgon-mask, part of the handle of a vessel, was found.

The bakery

It seems unlikely that the building housed a bakery in the second century, in view of the absence of a large, roofed area for milling. It was installed in the Severan period, when room 29, with a barrel vault, was built. The only barrel vaults in the building are in this room and the adjacent oven-room 23. Barrel vaults slow down the cooling of an oven and do not catch fire as easily as a ceiling resting on wooden beams.

It is probably not a coincidence that opposite the Caseggiato dei Molini, across Via dei Molini, is one of Ostia’s largest store buildings for grain, the Grandi Horrea, dating back to the reign of Claudius. There is however no exit related to the Molini in the west wall of these horrea. The ladders in the two rooms on Via dei Molini cannot have been used for supplying grain. Ladders are not suited for porters carrying loads. Because there are two ladders, many people were expected to use them. Possibly this was a fire escape: after a rebuilding in the Severan period the Grandi Horrea had only one, narrow exit. Presumably the grain was supplied over Via dei Molini, using a floor resting on the arches spanning the road, and on the two tiny rooms, that have surprisingly thick walls. The grain was then fed to the millstones through pipes. This hypothesis is supported by the find in the Molini of parts of masonry covered by opus signinum, which according to Calza had fallen down from the upper floors and possibly belonged to basins.

As to the interior of the mill-bakery, at least the rooms with a basalt floor and a high ceiling were used for the various steps in the bread making: 6, 7, 17 to 22, and 26 to 28. Millstones, kneading-machines and ovens bear witness to these activities. Basalt blocks with imprints of hooves are found in rooms 18, 19, and especially 7. They may indicate that the animals who turned the millstones had horseshoes. The imprints do not form circles, so that apparently blocks were moved, which is understandable, because the ruts created by the animals rotating the millstones will eventually have made them stumble, necessitating the replacement of worn blocks by smooth ones.

At least rooms 19-21 were used for the milling. This large hall has a floor of basalt blocks and a basin. Many millstones were found here (there must have been at least ten). The metae have a square hole in the top, with lead against the sides and on the bottom. On one catillus is the inscription CHRY. Inscriptions have also been found on millstones from Pompeii, Palestrina and Rome, and in the Ostian Molino I,XIII,4. They are initials or the beginnings of names, whose is not known. The volcanic stone the millstones are made of comes from Orvieto.

Five kneading-machines were found in room 22. Blades were inserted in holes in the sides. The grafting of the vertical spindle can still be seen in the bottom of the north-western one (iron and lead with a small, round hole in the centre). The basin in room 28 must have supplied the large quantities of water necessary for the kneading. Sieving took place during and after the grinding and before the kneading, possibly in room 21. The moulding of the bread was the first step after the kneading, and for this room 13 may have been used, where a wooden table may have rested on thin walls.

The remains of two ovens are reported. One was in room 23, the other in the east part of 23 and west part of 14. The oven in room 14 was abandoned after c. 250 AD. Rooms 29 and 23 had barrel-vaults, related to the oven in room 23, which was approached from the west. The missing of large stretches of masonry on the north and south walls of room 23 is probably also related to the oven. In the south wall the stretch is funnel-shaped and possibly the start of a chimney. In the dividing wall of rooms 29 and 23 are two draught-holes.

The oven in room 23 was uncovered once more in 1996. The floor in front of the oven is considerably below that of the other rooms. The oven is made of a cupola of tufa stones on top of a brick podium. In the front of the podium is a deep niche. Most interesting are four horizontal grooves inside the oven, in the tufa blocks: two wide ones, and two narrow ones above. Possibly they testify to the presence of one or more revolving grates on which the bread was baked, an understandable device, because the oven has a depth of almost five metres.

The long basin in room 26 looks like a trough, suggesting that a stable was nearby, for which room 17 is the best candidate. The threshold leading to room 26 from room 18 is very heavily damaged, probably the result of the passage of animals. Perhaps the bakers also used the Caseggiato di Diana as a stable. In the room to the south of its courtyard a trough and a basalt floor were installed, and there are also basalt blocks in a room to the east of the courtyard.

Rooms 17-22 were lit by large windows, that were probably also installed for ventilation. The various basins, also on the first floor, must have been served by Ostia’s aqueduct. Lead water-pipes were found below the pavement of Via dei Molini, branching off to room 7, and below rooms 24 and 25. An important, if not the main entrance was room 12, along staircase 9 and the tiny room 11. On the south wall of room 11 is a fragment of a stucco relief, a more expensive kind of decoration, indicating that the room was of special importance: an office, or, as the excavators have suggested, a shrine.



The large hall with millstones (rooms 19-21), seen from the south (room 18). The red bricks of the huge arches, one of which is intact, is Severan. The arches were set against slender Antonine arches of yellow bricks, one of which can be seen at the left edge of the photograph. Note the large window in the west (left) wall. Photograph: Melissa Sellers.
Room 22 with five kneading-machines, seen from the west. Photograph: Simon Bakker.


The alley to the west of the building (rooms 26-29), seen from the south. Note the two basins that were set against the west (left) wall. Photograph: Jan Theo Bakker.

The oven in room 23, seen from the west. Note the grooves in the tufa cupola. Photograph: Jan Theo Bakker.


Detail of the grooves (for revolving grates?) in the oven in room 23. Above two wide grooves are two narrow grooves. Photograph: Jan Theo Bakker.

The remains of the oven in room 14, seen from the east.

Photograph: Jan Theo Bakker.


Via di Diana to the south of the building, seen from the east.
Photograph: Jan Theo Bakker.

The bar in room 15, seen from the south-east.
Photograph: Jan Theo Bakker.


The bar in room 16, seen from the south-west. Photograph: Jan Theo Bakker.

The relief of a Genius with a snake in the east facade. Photograph: Max Victor David.


The relief with carpenters’ tools in one of the Severan piers that were set against the east facade. Photograph: Jan Theo Bakker.

The east facade between rooms 6 and 7. On the east facade are remains of a thin layer of plaster with red paint. The paint is both on Hadrianic and Severan masonry. Photograph: Jan Theo Bakker.

Detail of the red paint on the east facade. Photograph: Jan Theo Bakker.


Bronze sheets found in rooms 17-19: Sagittarius, Lion, Scorpion. Floriani Squarciapino 1953, Tav. CXIV, 1.

Bronze sheets found in rooms 17-19: Mars, Venus, Gemini, Sol. Floriani Squarciapino 1953, Tav. CXIV, 2.


A bronze statuette of a Lar, found in room 17. The Lar is wearing a tunica and sandals, and has a patera in his right hand, while the left hand is lost. An inlay of silver laurel- branches in the base identify the Lar as Lar Augusti. Photograph: Sopr. Arch. di Ostia B20.

Bronze statuettes from the building: Hercules, Minerva, and one of the Dioscuri. Photograph: Sopr. Arch. di Ostia B22.


Bronze objects from the building: feet of candelabra, a twisted snake, and a vase-handle ending in a female face. Photograph: Sopr. Arch. di Ostia B19.

Bronze objects from the building: a scorpion, a protome with a horse, and Mercurius. Photograph: Sopr. Arch. di Ostia B23.

Three busts from the building: of a young negro, of a winged child, and of Jupiter-Serapis. Photograph: Sopr. Arch. di Ostia B25.

Sacello del Silvano (Regio I – Insula III – I,III,2)

The history of the excavations

The Shrine of Silvanus is situated in the south-west corner of the Caseggiato dei Molini (room 25, to the north of the tiny room 24, that can be entered from Via di Diana). Many rooms of the Caseggiato dei Molini were unearthed for the first time in 1870, under the direction of P.E. Visconti. Room 25, the Sacello del Silvano, must have been investigated in that year as well. In the Giornale di Roma of May 16th the find is reported of a painting of Silvanus. It was a large painting of Silvanus “with the usual attributes”, found “near the entrance” to the building. On it EX VISO was read, written with large purple letters. It had a graffito with a date, “possibly from the time of Maximian”. In 1887 the words EX VISO were published in the CIL by Dessau, who remarks that the painting can no longer be found, and that not even C.L. Visconti and Lanciani, who had been present at the excavations at the time, knew anything about it. Now on a painting of Silvanus found in the Sacello during the First World War a graffito with the consular date 215 AD was found by Wirth in 1928. It is hard to believe that the building had two paintings of Silvanus with a date. The words EX VISO were never seen again, but they can easily have disappeared if they had been applied “dry”. The disappearance of the painting could be due to the filling of the room with earth after the 1870-excavation.

A further problem is that in 1914 a layer was found in the Sacello described as “incendio o materiale incendiato”, of varying height, with an average height of one metre. In this layer a large number of objects was found, mainly fragments of utensils and tools, and a marble statuette of a Lar. The statuette was found between the altar and the large niche in the south wall of the Sacello, the other finds according to Calza near the entrance of the shrine. Now the Visconti’s did not work with the accuracy of modern times, but they will not have missed the statuette. And what is more, the statuette, with a preserved height of 37 centimetres, was found at a distance of only approximately 80 centimetres from the painting of Silvanus, at a height of approximately one metre, that is some 25 centimetres above the bottom of the painting. If the painting was found, how could the statuette have been overlooked? But the story becomes more complicated.

In 1909 Lanciani wrote: “In the excavations of 1858 led by Visconti a house was discovered in the “Strada delle Pistrine”, in the lararium of which some fifty bronze and silver statuettes of domestic gods were lying partly on the steps of the altar, partly on the floor”. In the survey of the 1855-1870 excavations compiled by Paschetto there is nothing that concurs even vaguely with this description. That the Sacello del Silvano is meant here has been suggested by C.C. van Essen. There are two arguments in favour of the identification. First of all the location of the shrine, on the Strada delle Pistrine (Via dei Molini): I know of no other large private shrine along this road. Secondly the nature of the find: apparently the shrine was suddenly destroyed and left as it was, which concurs with the fire which destroyed the Caseggiato dei Molini.

However, the words “on the steps of the altar” do not seem to refer to the Sacello, because there a simple, rectangular altar may be seen. The answer to this problem lies perhaps in Finelli’s remark “tutta l’ara … fu anche rialzata più tardi facendovi sopra una costruzione molto andante”. A second problem is that the find of the statuettes is not mentioned in the reports from 1870. Here 14 statuettes from the building are mentioned, all found before April 5th, whereas the painting is mentioned for the first time on May 16th (we do not learn where and in what context the statuettes were found). Why is Lanciani the only one to report this spectacular find, why does he do so at such a late date and why does he mention the year 1858?

The answer probably partly lies in a theft. It is hinted at by Finelli, who says that the first excavators were “scavatori poco scrupolosi e gli oggetti rinvenuti non tutti raggiunsero i musei papalini”. J. Carcopino goes into more detail: “Une circonstance qui ne dut pas engager C.L. Visconti à donner à ses fouilles dans ses parages une trop bruyante publicité, ce fut l’affaire G….. A 83 m. à l’Est de l’angle Est du temple, et à 21 m. au Sud “fra gli avanzi di edificio privato incominciati a scavare sul principio di quell’anno (1870) … si rinvenne una quantità prodigiosa di sculture figurate ed ornamenti in bronzo quasi che una o più delle officine tornate in luce avessero appartenuto ad un fonditore di metalli. Molti di questi piccoli oggetti d’arte furono involati (G…..) e venduti a Napoli” (Notes Lanciani). Ce surveillant indélicat est nommé en toutes lettres dans les notes de M. Lanciani: il se déroba, par la fuite, aux poursuites judiciaires”. Lanciani clearly refers to the Molini (21 metres to the south of the (north)eastern corner of “the temple” (the Capitolium) must be to the north; the mention of workshops should be noted). The theft furnishes an explanation for a gap in the reports. All finds had been reported on April 5th. Then the painting of Silvanus is mentioned for the first time on May 16th. The end of the investigations is reported on June 3rd. The gap may the result of the avoiding of a “trop bruyante publicité”.

The available data does not allow a reconstruction of what exactly happened in the Sacello del Silvano in 1870, but the arguments for thinking that the Sacello was excavated, and that the painting of Silvanus and a large number of statuettes were found are, in my view, sound.

[Note. In 2006 and 2007 I discovered that Lanciani wrote the following in 1889 (“Les recentes fouilles d’Ostie. La Caserme des Vigiles et l’Augusteum”, MEFRA, 174-179; Notes from Rome, p. 214): “La découverte d’un magasin d’objets d’art, avec plusieurs douzaines de statuettes en bronze, faite en 1870 dans la Via delle Pistrine …”; “In the spring of 1870, the last campaign led by the Viscontis, a room was discovered on the west side of the Via delle Pistrine containing sixty or seventy statuettes of bronze, presumably the property of an old dealer in works of art”; JThB, 1-Sep-2007]

The building periods

Rooms 24 and 25 (the Sacello) were installed in the southern part of an alley or corridor between the Caseggiato di Diana (I,III,3-4) and the Caseggiato dei Molini. Two Antonine phases are found here. The northern part of the east wall is Hadrianic, the southern part Antonine. The southern end of the latter part was rebuilt in the Antonine period. The west wall, containing two brick piers, is Antonine as well. Set against the southern part is another Antonine wall. The latter wall and the rebuilt part of the east wall form an entity and create the southern entrance to the alley.

Plan of the shrine (max. w. 2.06, max. d. 6.50). North is up. Dotted: Hadrianic; hatched: Antonine; double hatched: c. 210-235 AD. Bakker 1994, fig. 19.

In the lateral walls of room 24 and directly to the south of the dividing wall of rooms 24 and 25 are two F-shaped grooves, one in each wall, meant to receive beams which presumably blocked the alley for the general public or for animals. The nature of the obstruction indicates that for some time no rooms were installed in the southern part of the alley on the ground floor. The doorway in the dividing wall between room 24 and room 16 to the east was freely accessible.

Beamholes can be seen in the west and east walls of rooms 24 and 25 (one of these in Severan masonry). Their tops are at the average h. 3.65, the normal height for the Hadrianic and Antonine periods in the rest of the building. The beamholes in the west wall of room 16 are passing through the wall, suggesting that there was a roof over room 24 during the first Antonine phase. The secondary west wall can only have been meant for the insertion of beamholes: apparently no permission was given to insert them in the Caseggiato di Diana. The existence of an upper floor in the second Antonine period is indicated by the presence of a mezzanine-window over the entrance from Via di Diana. There was no door on the first floor in the dividing walls of rooms 16 and 24, and 17 and 25. Apparently then the mezzanine-floor was reached along a mezzanine-staircase or ladder in the alley.

3D-view of rooms 25 (center) and 24 (right). From the northwest. Bakker 1994, fig. 20.

Room 25 in its present form was created during the period c. 210-235 AD. The space between the two Antonine brick piers of the west wall was filled in. Against the northern pier a wall was set at a right angle, leaving an entrance 0.96 wide. The western half of the door in the Hadrianic east wall was filled in. In the filling are two slit-windows, lighting room 17. Later in the Severan period the eastern half of the doorway was filled in as well. The slit-windows were now blocked, but not filled in, so that two niches resulted. The dividing wall between rooms 24 and 25 has been so heavily restored that its masonry cannot be dated. It was set against the west wall and into the east wall of the alley. In the eastern part of the south side was a 0.75 wide floor-niche, which was later filled in. In the west part of the north side are two wall-niches, one 0.41 wide, the other measuring only 0.15 x 0.12 x 0.15. The former was later filled in.

After the Severan period the eastern half of the doorway connecting rooms 24 and 16 was blocked, in two phases. In room 24 a recess or floor-niche resulted, on the back of which slight traces of red paint have been preserved. In the upper right corner of this niche was a second, small wall-niche.

The paintings and the floor-mosaic in room 25

The oldest layers of plaster can be found on the north, east and west walls, one on the north and east walls, two on the west wall, with traces of red paint on the first (layer 1). The layers are very thin.

A superimposed, thick layer is found on all walls (layer 2). Next to the larger niche in the south wall a small figure wearing a tunica was depicted. Calza thinks this may have been a Lar. Not enough of the figure has been preserved for a reliable identification. Furthermore large rectangles were painted, on the west wall containing three “floating” figures: a bird, a gorgoneion, and a dolphin. On the southern part of the west wall Silvanus was painted (now in the museum of Ostia). He is a little over one metre high and standing in a frame. He was partly gilded and is holding a branch and a sickle-shaped knife. He is accompanied by a dog and standing next to an altar. The painting was inserted in the layer, witness grooves. It was not a structural part of the layer, added immediately after the completion of the rest of the paintings by a “master”-painter: one of the grooves cuts off rectangles of layer 2, and there is no band next to this groove, contrary to another groove to the north (see below).

At least four later layers, all very thin, can be found. The oldest of these has several figures, which were originally red (layer 3). Later several colours were added (layer 4). Layer 3-4 is encountered first of all directly behind the entrance. On the west wall three horses and a male figure holding a sceptre or lance were painted. The attribute and the proximity of horses indicate that the figure, which has been very badly preserved, is a Dioscure. A horse and presumably the other Dioscure were painted on the opposite wall. On the southern part of the west wall, to the right of Silvanus, at least one figure was painted. All that has been preserved is the right arm, holding a lance or sceptre.

The remains of the Dioscure on the west wall. Bakker 1994, fig. 21.

On the southern part of the east wall a row of figures was painted. The first of these is holding a sceptre or lance and wearing a mantle around the hips, whereas the upper part of the body is naked (the other arm and the head are missing). Calza, who for the most part saw layer 5 (now largely disappeared), which repeats layer 3-4 with only one exception, calls this a figure in the posture of the Augustus of Primaporta. Moormann however points out that the figure represents a first century Emperor, whose identity will be discussed below. The second, third, and fourth figures were correctly identified by Calza. Next to the Emperor is the small Harpocrates, holding a cornucopiae and touching his lips with one finger of his other hand. He is standing at the feet of Isis, who is holding a sistrum. Her other arm cannot be traced. The fourth figure is Fortuna, holding a cornucopiae and a rudder. Next to Fortuna is a figure with rudder and tessera, regarded by Calza as Liberalitas. Nicolet however has shown that Annona is depicted: the tessera is found as an attribute of Liberalitas, but she has a cornucopiae in the other hand. The next figure is male in view of its short mantle. In one hand is a cornucopiae, the other does not hold an attribute. Around the head was perhaps a nimbus. The gender and the cornucopiae suggest it is a Genius. I know one other depiction of a Genius with only one attribute, and no parallels for the nimbus. The final figure was called by Calza a figure in the posture of the lysippian Alexander the Great. Moormann has confirmed this identification. This figure is naked except for a mantle hanging down from a shoulder, and holds a sceptre or lance. It is not clear whether there is an object in the other hand. Around the head is a nimbus.

The next intervention (layer 5) can be traced only on the southern part of the east wall and the outer north wall. The Emperor was, as far as can be seen, left unchanged. Harpocrates, Fortuna, Annona, and the Genius were repainted. Isis was replaced by a male figure, holding a patera in its right hand (a Genius?). Between the four central figures torches were painted. A painting of Silvanus was seen by Finelli on the outer north wall, next to the entrance. According to him it belongs to layer 5. Finally a thin, white layer was carelessly applied (layer 6).

On the floor of the room is a black-and-white mosaic, with the depiction of a victimarius, an assistant at sacrifices, holding an axe above his head. The figure was identified by Calza. At an unknown point in time a masonry altar was set on top of some object depicted in front of the victimarius.

The dates of the plaster and the floor-mosaic

Layers 1 and 2 are on masonry which according to Heres is from the period c. 210-235 AD. For layer 2 stylistic datings of c. 200 and 210 AD have been suggested. The painting of Silvanus, inserted in layer 2, cannot be later than 215 AD, witness a graffito, discussed below. For this painting stylistic datings in the period of Commodus and of c. 210 AD have been suggested. The analysis of the paintings by Moormann has confirmed the dating of layer 2 and Silvanus in the Severan period. The masonry and the graffito narrow down the dating of layers 1 and 2 and of Silvanus to the years c. 210-215 AD. The partial blocking with slit-windows of the door between rooms 17 and 25, dated by Heres to the later Severan period, can then also be pinpointed to these years, whereas the south wall of the room cannot be later than c. 210-215 AD.

The later phases are difficult to date. A parallel for layers 3-4 is offered by a fragment of a painting from Caseggiato IV,II,5, on which the lower part of three figures has been preserved. The figures bear a great resemblance to figures on layers 3-4, as to their shape, posture, clothes and colours. What remains of the first and third figure is almost identical to the lower parts of the Genius and Alexander the Great. The lower part of the central figure may be compared to the lower parts of Isis, Fortuna and Annona. For this painting dates in the second century and in the Severan period have been suggested. For layers 3-4 or layer 5 the later third or fourth century and a date of c. 240 AD have been suggested. For layer 5 Moormann suggests a date in the period Caracalla – c. 250 AD.

Layers 1 and 2

From the thin layer 1 no conclusions can be drawn. When layer 2 was applied the Sacello was equipped with an interesting ante-room. At 2.34 from the north wall is a vertical groove in the plaster of the west wall (1.5 cm. wide), next to which is a red band belonging to layer 2. At 2.13 from the floor it runs horizontally towards the north wall over a distance of 0.16. Then it continues vertically again. I suggest that a wooden partition was fastened in the groove. Rectangles and “floating” figures were painted to the south of the groove. To the north a single band is visible, running upwards in the corner between the west and north walls, and, at a height of 2.36, continuing horizontally towards the south.

The inserted painting of Silvanus replaced another painting of special importance, most likely of a deity, because the composition of layer 2 to the north of Silvanus takes the presence of a painting breaking the pattern to the south into account: three large compartments are found, the middle one with an accolade and a curved band below; the vertical bands separating these compartments are accompanied by lines to the left, with the exception of the band bordering the left end of the first one, next to Silvanus, which has a line to the right. The reason for the insertion of Silvanus into layer 2 can be deduced from the words EX VISO, read on the painting in 1870. Apparently a deity (Silvanus himself or another deity) had appeared to someone in a dream and ordered the dedication of the painting.

The grooves around Silvanus presumably held a wooden frame: to the left of the righthand groove, which runs diagonally downwards towards the floor, the pattern of 2 was not restored; the presence of a frame in the groove above Silvanus is suggested by two painted beam-ends on either side of Silvanus’ head, which may have “supported” it. The plaster above Silvanus was painted yellow.

On the east wall layer 2 is partly covered by superimposed layers, and one or more deities might be concealed there. Nevertheless it is clear that the shrine was regarded as a shrine of Silvanus, in view of the painting of this deity belonging to phase 5 on the outer north wall, next to the entrance: as a religious “shop-sign” it explains to which deity the room was dedicated. The back part of the room was the focus of the cult-room, as is indicated by the presence of Silvanus and by the large niche in the back wall.

Silvanus was one of the most popular deities of the Imperial period, but has never had state-temples and -feasts. He was a god of the woods and gardens, of live-stock, of the villa, and of boundaries. The knife and branch he is holding refer to the woods and garden, the dog sitting next to him is the guardian of the plot of ground. In Italy he was worshipped intensively by slaves and freed slaves, especially those of the Emperor. Unfortunately the reason for the erection of the dedicatory inscriptions is not often mentioned explicitly. Documented are, in Italy: pro salute domini, or dominae, or imperatoris; pro sua salute; ob libertatem; huic loco tutela; Silvano sancto, cui magnas gratias ago conductor aucupiorum; ex viso, ex iussu. In Ostian inscriptions he is mentioned together with Hercules, Isis, Magna Mater, Numen Domus Aug., and Numen Sarapis. A relief of and a dedication to Silvanus were found in mithraea. Mention is made of a collegium Silvani Aug. maioris quod est Hilarionis functus sacomari. Two dedications were made pro salute imperatoris. Most dedicators were slaves and freedmen.

Layers 3 and 4

Of layers 3-4 three groups remain. First of all the four horses and the two Dioscures. Secondly the row of figures from the Emperor to Alexander. Thirdly one or more figures next to Silvanus.

The Dioscures and at least four horses were painted in the ante-room. The most southern horse on the west wall does not continue to the south of the groove for the presumed wooden partition, where very scanty remains of a superimposed layer are found. It appears that the continuation of the horse was on the wooden partition which was attached in the groove.

The Dioscures, the twins Castor and Pollux, were usually depicted with horses, who played an important role in their official worship. They were patrons of the equestrian order, and a sacrifice to Castor was the focus of an annual parade of the Equites. They were furthermore patrons of the chariot-races in the circus. Ostia was the only place where the Romans continued the ancient relation between the Dioscures and sea-faring. The twins were here officially worshipped in their capacity of protectors of navigation. Each year on January 27th Ludi Castorum were held at Ostia – led by the Praetor Urbanus, later the Praefectus Urbi -, quite possibly involving horse-races. The Ostian temple, where the Dioscures gave oracles, has recently been identified: it is to the west of the so-called Palazzo Imperiale, on top of the so-called Navalia. Few dedicatory inscriptions mentioning the Dioscures have been found in the western half of the Empire. In Rome depictions of the Castores were bequeathed to the mensores machinarii frumenti publici in 198 AD. In Ostian dedications they are found together with Neptunus and Jupiter-Serapis. The relation with the sea is obvious in the former inscription. Serapis was sometimes worshipped as protector of shipping.

Group two was opposite and most likely supplemented by group three. Of group three hardly anything remains, and we can only guess at the identity of the figure with lance or sceptre next to Silvanus. Group two can nevertheless be regarded as a separate unit, because the scene is framed by two rulers, an Emperor and the lysippian Alexander the Great.

The figures of group two were originally framed by bands, later covered by layer 4. The bands frequently intersect the figures and in that case presumably pass behind them. Some of the bands, both horizontal and vertical ones, have a slight inclination. Such bands and lines are found frequently on Ostian paintings. Possible explanations are carelessness, and economic or aesthetic motives. As to the total height of the figures, only that of Alexander is known (0.83), but from the size of the bodies of the remaining figures can be deduced that all had the same height, with the exception of course of Harpocrates.

As to the identity of the Emperor, Moormann points out that only during the first century AD, and especially during the Julio-Claudian period, were Emperors represented nude apart from a drape around the hips and sometimes over the shoulder, i.e. in heroic form, especially after death. What is seen therefore is an outmoded depiction, not of a living, but of a deceased Emperor. Here Augustus, the most praised and famous Julio-Claudian Emperor, immediately comes to mind.

The Egyptian goddess Isis was worshipped by the Romans as deity of plenty and nature, steering fate, helping men in various ways, as healer of the sick, and as protectress of shipping (Isis Pelagia, Isis Pharia). It has been thought that she protected the Emperor as Isis Aug., but because her husband Serapis is called Serapis Aug. only once in inscriptions, whereas his relation to Emperors was much stronger than that of Isis, Aug. perhaps rather refers to her “reign”. In Ostia and Portus Isis was associated particularly with the large quantities of fiscal grain which were drawn from Egypt, and with the Alexandrian grain-fleet. One may compare an aureus from 215 AD, on which Isis is seen giving two ears of grain to Caracalla, who is wearing military dress, holding a spear, and subduing a crocodile. The Ostian temple, probably situated in the area around the Palazzo Imperiale, has not yet been found. The Isis-feast Navigium Isidis, celebrated on March 5th in relation to the opening of the shipping season, may have been held in Ostia. From the way in which Isis is depicted in the shrine it cannot be deduced in which capacity she is present. She was often depicted with sistrum and situla.

The aureus from 215 AD, showing Isis giving ears of grain to Caracalla, who is subduing a crocodile. Obverse: ANTONINVS PIVS AVG GERM; reverse: P M TR P XVIII COS IIII P P.

At the feet of Isis is Harpocrates, the young Horus, the child of Isis and Serapis. As usual he holds the horn of plenty, a reference to fertility and prosperity, whereas one finger of his other hand touches his lips (according to the Romans to ask for silence and attention, originally a child’s gesture). In the Roman world he was a sunnaos theos only, and in inscriptions he is always found together with Isis.

To the right of Isis is Fortuna. She was primarily regarded as a benevolent power, as a goddess of luck and good fortune, rarely as the personification of blind fate or even a malicious power. She protected human beings and localities. She was often called Fortuna Aug., guardian of the Emperor, whereas Fortuna Redux watched over the safe return of the Emperor from foreign journeys. In the shrine she is depicted in the way she is found most often: with rudder, indicating her control of fate, and cornucopiae, referring to the blessings she could bestow. Perhaps the rudder also refers to sea-faring, although other parts of ships, such as the prow or an anchor, could highlight this aspect.

The fourth deity is Annona, the personification of the food supply, especially the grain supply, of Rome. In her right hand is the tessera frumentaria, on presentation of which a restricted number of recipients collected free grain each month in the Porticus Minucia Frumentaria in Rome. In her other hand she holds a rudder, referring to the overseas transport of the grain.

The last figures are a Genius and Alexander the Great. The latter and perhaps the former are provided with a nimbus, a symbol of heroic status and deification.

Augustus and Alexander the Great frame Isis, Harpocrates, Fortuna, Annona and the Genius. Thus the central and most important part of the scene is literally to be understood within the framework of the two rulers.

Augustus and Alexander as two exemplary rulers from the past in whose footsteps the contemporaneous ruler should follow. Depictions of illustrious men from the past could be found in household shrines. Denique hodieque in multis domibus Marci Antonini statuae consistunt inter deos penates, says a biographer of Marcus Aurelius. Suetonius presented a statuette of Augustus to Hadrian, who gave it a place amongst his household gods. The two lararia of Alexander Severus contained depictions of Alexander the Great, deified Emperors, Abraham, Christ, Orpheus, Achilles, Apollonius of Tyana, Cicero, and Virgil. It was a very small step to regard such men as exempla virtutis, which could even lead to association, imitation and identification. Thus we know that in the third century AD two Emperors emphatically evoked Alexander the Great as predecessor – Caracalla (se Alexandro Magno Macedoni aequandum putabat) and Alexander Severus (Alexandri habitu nummos plurimos figuravit; Alexandrum praecipue imitatus est) -, whereas three Emperors did so to a lesser extent – Gordian III, Philip the Arab and Gallienus -.

With the help of the graffito mentioned above it can be shown that Caracalla is referred to. It is found next to Silvanus, in a tabula ansata: “Calpurnius, night-watchman from the centuria of Ostiensis, from the seventh (or sixth) cohors, during the reign of Caracalla, in the year of consuls Laetus and Cerialis, X”. The year is 215 AD. The graffito was written by one of the vigiles (fire-fighters), who were stationed in the Caserma dei Vigili. Calpurnius was, when he wrote the graffito, sebaciarius, night-watchman. This duty was assigned to the vigiles in rotation for a period of one month. The sebaciarii patrolled through the city at night with tallow-candles, on the look-out for fire, and probably acting as police. The X is in an ansa of the tabula and means votis decennalibus. Calpurnius therefore asked Silvanus to give Caracalla ten more years as Emperor.

The graffito of Calpurnius next to Silvanus (above) (0.11 x 0.04) and the  graffito below that of Marius and Anna on the east wall (below) (0.08 x 0.015). Bakker 1994, fig. 22.

Parallels for this graffito have been found in Rome. In 1866 a fire-station (excubitorium, not the barracks, statio) of the seventh cohort was excavated in Trastevere. In and near its shrine 97 graffiti were found, most of which are from sebaciarii. The earliest dated graffito is from February 215, the last from March 245 AD. Not everything can be understood, and not all graffiti have been completely preserved. Some short ones mention only the name of a miles, frumentarius, or sebaciarius, while others simply record that the sebaciaria have been performed in some month (sebaciaria fecit mense …). The longer graffiti of the sebaciarii are similar to that of Calpurnius. These contain the names of the sebaciarius, the centurio, the cohors, and the eponymous consuls, and end with sebaciaria fecit mense …, or a similar statement. Apparently the patrols were not without danger, for sometimes it is stated that all had gone well (e.g. omnia tuta, salvis commanipulis). One graffito refers to the Genius of the watch-house. In a number of cases there is a plea for the welfare of the Emperor (e.g. salvo domino nostro) and the continuation of his government, sometimes expressed by writing X or XX in an ansa of a tabula ansata. It would not be surprising if the shrine in or near which the graffiti were found was related to the cult of the Emperors.

It can be established that Calpurnius must have seen the added figures of layers 3-4 when he wrote his graffito. On the east wall, between the two former slit-windows, is the graffito “Marius, Anna, April 25th”, followed by a leaf. A little below is written CERIALE, the name of one of the two eponymous consuls of 215 AD, which is also found in the graffito next to Silvanus. The handwriting shows clear similarities with that of Calpurnius’ graffito, and it may be assumed that it was he who added the year to the day. The two graffiti on the east wall are now read on layer 2, but must have been written on and through layers 3-4, which have here disappeared: they are very hard to find and read, because they are very superficial; the graffito of Marius and Anna was written with large, regular and accurate letters, which presupposes legibility; the letters of CERIALE are much thinner and shallower than those of the graffito next to Silvanus.

The graffito of Marius and Anna on the east wall (text 0.11 x 0.22). Bakker 1994, fig. 23.

Layers 3-4 therefore have a terminus post quem non of April 25th 215 AD, and the Emperor referred to must be Caracalla. He had left for Egypt in 214 AD, at the age of 26. After arriving in Thrace an Alexander-mania seized him. Herodianus and Dio (Herodianus 4, 8, 1.2; Dio 78, 7, 2) inform us that all of a sudden he became Alexander. He ordered the erection of statues that should show the likeness between himself and Alexander. According to the ancient historians such statues could also be seen in Rome, sometimes strange hybrid depictions of Caracalla and Alexander. He also referred to himself as “Augustus of the East”. Some people in the Caseggiato dei Molini had apparently acted upon Caracalla’s orders. Links between Caracalla and Augustus are also documented. The expression “Augustus of the East” is clearly ambiguous. In acclamations of the Fratres Arvales from 213 AD the phrase maior Augusto is found.

Caracalla was perhaps depicted next to Silvanus (group three). The Genius may well be the Genius Augusti, especially if there really is a nimbus around his head. The Genius Augusti and imagines of the Emperor frequently formed part of the contents of house-chapels, from Augustus onwards. Thus the Genius of Octavianus is presumably painted in a niche in building IX 9, 13 in Pompeii. Below the niche the graffito EX SC was read, which must refer to a senatusconsultum of 30 BC, ordering wine-offerings to (presumably the Genius of) Octavianus.

We then come to the deities in between Augustus and Alexander. They may refer to one or more donations from the Emperor. There is no central deity on the east wall, so that the figures in between Augustus and Alexander are to be interpreted either as a “story”, from left to right, or as juxtaposed figures, which may or may not be interrelated. If it is a “story”, Annona suggests that Isis and Harpocrates refer to the Egyptian grain which was distributed by the Emperor (for the protection of the transport of the grain the Dioscures are better candidates). The presence of Fortuna however presents a problem, because the reference to good fortune in between Isis and Annona does not contribute anything to the “story”. Apparently then we see a juxtaposition, and a reference to either different donations, or aspects of a single generous deed. As to Fortuna, the reference to good luck is so general that she cannot symbolize a specific donation. The frequent association of Isis and Fortuna comes to mind, a pair balanced nicely by Annona and the Genius Augusti.

The most obvious explanation for the deities is that they refer to the supply of grain by the Emperor. Isis and Harpocrates refer to the grain imported from Egypt, a province under the jurisdiction of the Emperor, and Annona to the fiscal involvement with the grain supply of Rome. The Dioscures are epicted in their Ostian capacity, as protectors of shipping.

Layer 5

On layer 5 the importance of the Emperor was stressed. Alexander was made wider, and the two rulers were now towering above the other figures, who were placed at a lower level or made smaller. The theme of the paintings was, as far as can be seen, left unaltered, with one exception: Isis was replaced by an unidentified figure holding a patera, perhaps a Genius. Strangely enough Harpocrates was repainted. It is not clear why layer 5 was applied. Perhaps the thin layers 3-4 had deteriorated so much, that layer 2 was shining through. The linking of Alexander the Great and Alexander Severus also comes to mind, but this would not have prompted a repainting, because Alexander the Great was already present. Not much money was spent on the new layer: for the new Fortuna one arm of the older Fortuna was reused. Nor was this layer a qualitative improvement. The figures of layers 3-4 had been painted in a fairly detailed and naturalistic way. The figures on Isis and Annona are little more than red blurs, and Alexander was changed into a disproportionate creature.

Layer 6

It is not clear whether the final, white layer covered all walls completely. It may have been the beginning of repainting, but it is also possible that the walls were to be left white.

The floor-mosaic

On the floor of the cult-room is a depiction in mosaic of a victimarius, a man about to kill a sacrificial animal. It has not been dated. Victimarii are also found in the Severan mosaic of a shrine for the cult of the Emperors in the Caserma dei Vigili, here killing bulls. In the cult of the Emperors bulls were sacrificed to the Genius Augusti. It is an attractive thought that our victimarius too is related to the bull and the Genius Augusti.

The fire and the finds

In room 24 Calza and Finelli found a destruction-layer resulting from fire, c. 3.00 high, with fifty-six unidentified coins. In the Sacello fifty bronze and silver statuettes of deities were presumably found in 1870, on the floor and near the altar. Calza and Finelli found a marble Lar-statuette behind the altar and a fairly large number of objects near the entrance. Some of the latter clearly do not belong in a shrine, such as an ear-pick, a pick-axe, fragments of a saw, a palette-knife to smooth paint, and two knives belonging to a painter or pharmacist. The latter finds, together with layer 6, suggest that the room was being repainted when it caught fire in the late third century. It must be remembered however that it is not at all clear what happened in the shrine during the excavations of 1870.

The functions of room 24 and the Sacello

The ground floor of room 24 was originally part of the alley. Its small and isolated first floor may have been a bedroom for slaves. The later history of the room is a problem, because the masonry blocking the door between rooms 16 and 24, and the dividing wall of rooms 24 and the Sacello cannot be dated accurately. When the building was destroyed the ground floor of the room was independent and accessible from Via di Diana only. A niche in the north wall had been blocked. The room had not, however, been abandoned: the niche left after the blocking of the door in the east wall had been painted, and in it another small niche had been made. The isolation of the room and the niches suggest that the room was a shrine. Cult-rooms which could only be entered from the street have been found in Pompeii as well.

As room 24, room 25 was originally part of the alley. It still was for a very short period of time in the later Severan period: the two slit-windows lighting room 17 indicate that there was no roof over this part of the alley. In the years c. 210-215 AD the cult-room was created. The statuettes of deities present in the room at the time of its destruction indicate that it was a cult-room until the end.

The initiators of and the people using the shrine

The building housed a bakery at the time of the application of the figurative wall-paintings. The paintings emphasize tge grain supply, but the reference to grain is only half the story of the new paintings. The bakers showed devotion towards the Emperor, and the grain was fiscal grain. The hypothesis may be forwarded that a direct relation existed between the bakery and the Emperor, and that the bakery was related to the fiscal component of the grain supply. The bakery was then owned by a member of the corpus pistorum or by the corpus itself, and an exemption from munera was earned by baking bread from fiscal grain, for the vigiles, for Imperial slaves, or perhaps even for the beneficiaries of Ostian distributions of free grain (frumentationes).

It cannot be established who was the initiator of the shrine: perhaps a baker from the Caseggiato dei Molini, perhaps the corpus pistorum, perhaps even a religious guild. It may be noted that Silvanus was popular especially amongst slaves and freed slaves, and that many slaves must have been working in the bakery. The importance of the Emperor in the shrine opens up the possibility that Silvanus was regarded as Silvanus Aug., a frequent association. Silvanus perhaps replaced Vesta, the protective deity of the bakers.

Although Calpurnius left a highly informative graffito it is not clear why he chose this shrine for his act of devotion, instead of the shrine for the cult of the Emperors in the Caserma dei Vigili. He cannot have been a passer-by, because the shrine could only be reached through the bakery. The new paintings must of course have been important to him. Perhaps the fire-fighter regularly visited the building because it was a fire hazard.

As to the visit of Marius and Anna to the shrine on April 25th, two religious feasts taking place on this day are documented. First of all the Sarapia, in honour of Serapis. Hardly anything is known about it. Secondly the Robigalia, in honour of Robigus, the purpose of which was the prevention of rust disease in grain. Caracalla’s fondness of Serapis could explain the former feast, whereas the latter fits very well into the context of a bakery. If Marius and Anna visited the shrine at the occasion of the Robigalia they nevertheless presumably worshipped Silvanus or the Emperor, because no dedications to Robigus are known and, with one possible exception, no depictions of him.



The shrine seen from the north-west. To the left is a horse, behind that two niches (blocked windows), and on the back part the row of figures. On the back wall is the lower part of a figure, to the left of two niches and behind a masonry altar. Photograph: ICCD.

The shrine seen from the north-east. To the right are horses, behind that a groove for a wooden partition, and on the back part paintings of layer 2. At the far end was the painting of Silvanus. Photograph: ICCD.


The back part of the west wall, with paintings belonging to layer 2. The painting of Silvanus is now in the museum, and was inserted with the help of the computer. At the far right are the remains of a figure belonging to layers 3-4 (the Emperor?). Photograph: ICCD.

The painting of Silvanus. Bakker 2001, fig. 6.


The graffito next to Silvanus. Photograph: Jan Theo Bakker.  COH(orte) VII (or VI) (centuria) OST(iensis) IMP(erante) AN(tonino) CO(n)S(ulibus) L[a]ETO ET CE RIALE SEBARIVS CALPVRNIVS, X

The figures as they appeared during the First World War (layer 5). To the right of a niche (a blocked window) are Augustus, Harpocrates, and an unidentified figure. Note the right arm of Isis (layers 3-4) behind the last figure. Photograph: Sopr. Arch. di Ostia.


The figures as they appeared during the First World War (layer 5). This is the central part, with Fortuna between two torches and Annona. Photograph: Sopr. Arch. di Ostia.

The figures as they appeared during the First World War (layer 5). This is the right part, with an unidentified figure (Genius?) and Alexander the Great. Photograph: Sopr. Arch. di Ostia.


The figures as they appeared after the almost complete disappearance of layer 5: Augustus, Harpocrates, Isis and Fortuna. The head of the Fortuna of layer 5 can still be seen on the right shoulder of the last figure. Photograph: ICCD.

The figures as they appeared after the almost complete disappearance of layer 5: Annona. Half of the Annona of layer 5 can still be seen. Photograph: ICCD.


The figures as they appeared after the almost complete disappearance of layer 5: a Genius (?) and Alexander the Great. Photograph: ICCD.

The parallel for the row of figures from Caseggiato IV,II,5. Photograph: ICCD.


The remains of a horse on the east wall. To the right was probably a Dioscure. Photograph: ICCD.

Remains on the front part of the west wall. From right to left: a horse (head lost), a horse, the lower part of a Dioscure, the front part of a horse (apparently continued on the wooden partition). Photograph: ICCD.


The mosaic of the victimarius in front of the altar. Photograph: ICCD.

The marble statuette of a Lar. Photograph: Sopr. Arch. di Ostia, B376.


The paintings as they appear today: layer 2 on the west wall. Photograph: Jan Theo Bakker.

The paintings as they appear today: the left part of the row of figures. Photograph: Jan Theo Bakker.


The paintings as they appear today: the right part of the row of figures. Photograph: Jan Theo Bakker.

The paintings as they appear today: the figure on the back wall. Photograph: Jan Theo Bakker.

Caseggiato del Pantomimo Apolausto and Caseggiato del Balcone Ligneo (Regio I – Insula II – I,II,2.6)

Plan of the building. After SO I.

The House of the Pantomime Apolaustus and the House of the Wooden Balcony were originally one complex, dated to c. 120 AD. In the south-east part a nymphaeum was installed in the first quarter of the fourth century. In the nymphaeum an inscription was reused mentioning the pantomime Apolaustus, a freedman of Marcus Aurelius and Lucius Verus. He gave his name to the south part of the building, a separate unit from an unknown point in time. The north part of the building was named after a wooden balcony that adorned the facade. The complex was excavated by Paribeni during the First World War, but had already been searched in the 18th or 19th century.

As to the original masonry, all outer walls and those of courtyard 27 are of opus latericium, for the remainder opus mixtum was used. A sidewalk ran only along the north facade. On the facade are traces of thin plaster with traces of red paint. Remains of thin and thick plaster are found inside. Beamholes for a ceiling are found at 2.75/3.00 (bottom and top respectively) from the present floor level. The total area is roughly 730 square metres.

Along the Decumanus Maximus is porticus 20. To the east is Nymphaeum I,II,1 (room 21). Rooms 22 and 23 may have been shops. In the north-east corner of 22 are remains of a staircase, accessible from Via dell’Ara dei Lari. To the west is corridor 24, with a threshold for one door in the south entrance. Room 25 has a secondary north wall in latericium, and secondary masonry in the central part (latericium, rubble, vittatum of very bad quality). It has a floor of basalt blocks. Room 26 may have been another shop. There is a well in the north-east corner.

The building is arranged around the central courtyard 27. There are remains of a staircase with travertine steps set against the east wall. There is a threshold for two doors in the passage to Piazza dei Lari. A travertine staircase and a basin were set against the west wall. In the south-west corner is a brick well with a tufa rim. Next to the well are three fragments of the upper part of millstones (catilli), one with part of the contraption into which the horizontal beam for rotating the millstone was inserted. The courtyard had a floor of basalt blocks.

Rooms 28, 29, 45, 38, 41, and 42 were shops. The shop-entrances were blocked or narrowed with vittatum, small tufa blocks and latericium. The rooms have a floor of basalt blocks. In the south-west corner of 28 lie the remains of a latrine: next to a hole in the wall is a square travertine seat of a latrine, with an opening shaped like a keyhole. In the north-west corner was a small basin. In the south-east part of shop 38 is a staircase, accessible from Piazza dei Lari. In room 41 are fragments of various objects, some of which may have been part of machines. The wall between rooms 41 and 42 was rebuilt. Paribeni has noted that the level of the first floor was now changed, because there are no beamholes in the new wall at the height of the old beamholes. At the level of the old holes a painting was placed in room 42, depicting a cantharus and vegetative ornaments. Corridor 39 was reached from Via di Diana. In the doorway is a threshold for two doors.

Wall-painting with cantharus from the east wall of room 42. From NSc 1916, p. 420 fig. 8.

Room 40 is a second courtyard. In the north-west corner lies a damaged, round, low object of volcanic stone, perhaps part of the lower part of a millstone (meta). On the floor are basalt blocks. In room 44 a large oven was installed. The low podium (h. c. 0.50?) was made of latericium, set against the mixtum. During the excavation the lower part of the superstructure was found, a ring made of large tufa blocks (diam. of cupola, outside meas., c. 5.00 x 5.20). The floor of the oven was made of two layers of brick, with sand in between. The oven opened towards the east, and part of the brick arch of the opening was found. In a collapsed part of this arch brick stamps were found from the period of Hadrian and Marcus Aurelius (one from the years 125-134 AD, 26 from 161-176 AD). The north wall of this room (modern mixtum) is very thick (0.70 – 0.75; the av. th. of the primary walls is 0.51). In the south-east corner the wall is thick as well. High up in this corner the lower part of a chimney may be preserved (a deep hole, starting at h. 2.55, w. 0.40).

The floors of basalt blocks, basins, remains of milling machinery and the oven leave no doubt that a bakery was installed in the complex. It was built however as a group of shops around a courtyard (room 27 and the south part of room 40). From the width of the doorways can be deduced that customers did not access the shops from the courtyard. The characteristic shop-entrances, with mezzanine-windows above, were along the north, east and perhaps south facade. The bakery was installed later, possibly during the reign of Marcus, as is suggested by the brick stamps found amongst the ruins of the oven.


All photos by Jan Theo Bakker


Wall of the Via Diana (top of plan)


Room 42 (left) and 41 (right) on plan


Room 38 (left) and Piazza dei Liari wall (right) on plan


Room 40 overview and entrance


Room 44


Area from Room 27 to Room 41


Room 27 (left, center) and Room 28 (right)

Caseggiato della Cisterna (Regio I – Insula XII – I,XII,4)

Building plan

In the House of the Cistern four units can be recognized:

  • The original Hadrianic building.
  • A bakery in the north and central part.
  • Water-wheels and cisterns in the south part, related to nearby baths.
  • A late antique exedra in the north part (the exedra, I,XII,3, is described separately).

The Hadrianic remains

Plan of the octagonal room. From Bakker 1999, fig. 30.

In the north part of the building are a few remains of a very large octagonal room from the Hadrianic period. The room contained four large semicircular niches, starting at floor level. The north-east one was obliterated by Exedra I,XII,3, the north-west one is badly preserved and overgrown. The south-west niche – with the exception of the west part, the reconstruction of which is hypothetical – and especially the south-east one – in which an oven was installed – are preserved quite well, with a maximum height of a few meters.

The brick walls have a thickness of 0.60. The lower part of the walls shows the core of the wall over a height of 0.40. In the surrounding rooms the Hadrianic facing continues up to the present floor level. The sides of the octagon measure a little over 20 feet of 0.296 (average length 6.02), created by the openings of the niches, wide passages between the niches, and walls 1.07 long on either end. The latter sections partially block the openings of the niches. The diameter of the room, from one corner between the short sections to another, must have been approximately 14.50. In the south part of the south-west niche is a doorway. The jambs show the core of the wall over a height of 0.92. Apparently this was originally a window, later changed into a doorway. In the east part of the south-east niche is an opening with a lintel arch. This opening was blocked, on the inside with brick masonry that is identical to the masonry of the niche, on the outside (in a shop) at a later date (bricks and small tufa blocks). This was either a doorway or a window. In the corner between the two short sections on either end of the wide passage between the south-west and south-east niche is some rubble masonry, probably foundations for columns.

Parallels for this kind of octagonal room can be found in Imperial buildings from the first to the third century AD. They are in the Domus Flavia on the Palatine, in the Piccole Terme and Piazza d’Oro in the Villa Hadriana, and again in Rome in the Baths of Caracalla and of Diocletian.

For the function of the room in the Caseggiato della Cisterna we have two clues. The missing facing of the lower end of the walls suggests a hypocaust. To the west, below the Foro della Statua Eroica, the remains of a Hadrianic bath have been found. Was the octagonal room perhaps the vestibule of the bath, slightly heated by the air that had passed through the actual bathing rooms?

In the south part of the buidling the remains of a Hadrianic arcaded courtyard can be recognized. To the east, along Semita dei Cippi, is a row of shops.

The bakery

In the northern half of the building a bakery was installed, to which some opus latericium seems to be related. On the floor of room 1 are some basalt blocks. In the west part of the tiny room 2 is a basin, widening towards the north. In the east part is a structure looking like a tiny oven. On top of a podium (h. 0.35) is a small cupola (inside diam. 0.60), made of pieces of brick and tufa, opening towards the north (w. of opening 0.38). The area to the south might be called room 3. In the east part is the best-preserved Hadrianic floor-niche. In the niche a large oven was installed. When the oven was built, two walls in the opening of the niche were partly removed. The brick podium of the oven (h.0.98) is protruding from the niche. In the front of the podium is a deep floor-niche or hole. The opening of the oven was towards the west. The first ring of the cupola, made of large tufa blocks, has been preserved (inside diam. west-east 4.00, north-south 4.20). There are some basalt blocks on the floor of this room, and on the floor of rooms 4, 5, and 6. Many basalt blocks and parts of other objects have been piled up in room 6. In the north-west corner of the shop to the east of room 3 is part of a kneading-machine, including part of a hole for the insertion of a blade. In the late 1990’s further excavations were carried out in Exedra I,XII,3. Several rooms with floors of basalt blocks were found, some with the imprint of hooves.

A basin, floors of basalt blocks, a large oven, and part of a kneading-machine indicate that a bakery was installed here. Its size and history are obviously problematic. Heres suggests a date in the last quarter of the second century for secondary piers in room 5. Recent excavations have shown that it was a large bakery, that evidently stopped functioning after the building of Exedra I,XII,3, in the middle of the fourth century. The northern part of the building has a total area of roughly 640 square metres. Four pairs of piers on Semita dei Cippi suggest a connection with building V,V,3, that looks like a commercial building.

The water-wheels and cisterns

In the southern half of the Hadrianic courtyard a large cistern (b) was installed, at two levels. This may have taken place between the end of the second and the fourth century (opus latericium and vittatum). The lower part could contain 80.000 litres. The cistern was partly supplied by the aqueduct, and partly by two water-wheels (a). It seems to have served quite a large area of the city.



The oven in room 3 seen from the west. Photograph: Jan Theo Bakker.

The interior of the oven in room 3. Photograph: Jan Theo Bakker.


A tufa block belonging to the base of the oven (fallen over) with grooves, possibly testifying to a rotating grate. Photograph: Jan Theo Bakker.

The tiny oven (?) next to the large oven. Photograph: Jan Theo Bakker.

Exedra (Regio I – Insula XII – I,XII,3)

Plan of the exedra. After SO I.

The northern end of the Caseggiato della Cisterna (I,XII,4) was in late antiquity changed into an exedra. It was erected on top of a bakery. Rooms with floors of basalt blocks, belonging to the bakery, were found in the late 1990’s. The exedra consists of a large semicircular wall (opus vittatum) with a rectangular wall-niche in the centre. The wall is followed by a colonnade.

The exedra was built in the middle of the fourth century AD. It was meant as decoration of the Decumanus Maximus, but also blocked the Semita dei Cippi, indicating that in this period fewer or no goods were taken from the Tiber to the south part of the city.


All photos by Jan Theo Bakker.

The exedra seen from the north-east, from Semita dei Cippi. Photograph: Jan Theo Bakker.

The floors belonging to the bakery, shortly after the excavation. Photograph: Jan Theo Bakker.

A bakery near the Caseggiato dei Dolii (I,IV,5)

Plan of the building. After SO I.

The remains of a bakery near the Caseggiato dei Dolii (I,IV,5; to the south of the museum) seem to be mentioned in an excavation report from 1878. The excavation is described of a room containing the lower part of a dolium. A liquid had left rings inside. It was connected with a basin to the south by means of a terracotta pipe. “Nel vano seguente sono stati trovati alcuni catini, e frammenti di catini di macine da grano, restaurati ab antico con grappe di bronzo impiombate”. The room with the dolium was apparently to the south of the museum, and is perhaps a room to the south-west of the large room with dolia defossa in the Caseggiato dei Dolii (I,IV,5). The presence of millstones nearby is mentioned by Carcopino: “Je laisse naturellement de côté les matériaux rapportés, comme les restes d’une meule en lave qu’on voit au pied de la terrasse du casone [the museum]”; “A l’entrée de B13 [the northernmost excavated room of I,V,2, to the north-west of the museum], un beau fragment de meule”. These remains are perhaps to be identified as the three catilli in the garden in front of the museum. The room to the south of the dolia had a pavement of basalt blocks.


All photos by Jan Theo Bakker


Wall of Via Delle (left) and Room 5 (right)


Room 1 (left) and Room 9 (right)


Room 12


Room 12

Molino (Regio I – Insula XIII – I,XIII,4)

Mill-bakery I,XIII,4 was excavated between 1938 and 1942. A few remains can be dated to c. 80-100 AD. The main building period (opus mixtum) can be dated to c. 100-125 AD. The building was now a mill-bakery. Various modifications took place in the later second and in the third century (opus latericium and vittatum). The latest modifications (opus vittatum), in the north-west part, have been dated to c. 370-440 AD. The width of the inner walls (0.44-0.46) indicates that the building was not very high.

Along the Cardo Maximus, to the west, are four shops (rooms 1-2, 3, 5, and 6-7). Four more shops are along the Semita dei Cippi, to the east (rooms 12, 13, 15, and 16). The workshop could be entered from corridors 4 and 14, and from a narrow alley to the north (17-18-19). Wagons may have entered the building through corridor 14. At some point in time a staircase (D) was added here. Corridor 4 could be blocked by a beam that was inserted in L-shaped holes.

Building plan

In the centre of the building is a large hall, with floors of basalt blocks. Only the northern part (8a) seems to have had a roof, carried by two long rows of brick piers. In the north part of 8a are five kneading-machines, made of volcanic stone. In the south part are eight mills of volcanic stone, surronded by mortar-beds. On three mills are inscriptions, perhaps the initials of the makers: S P C Y R, P A R, and M I V. Letters can also be read on one of the kneading machines: G I A H. The grain may have been poured into the mills from an attic, that was reached via staircase B.

The southern part of the hall (8b) did not have a roof. Here, in the south-east part, two cisterns were found, below the floor level. In the wall behind the cisterns the lower part of a curved wall-niche can be seen, which once contained a statuette of a deity. Two staircases (A and C) were set against the south wall.

The southern part of the workshop consists of rooms 9, 10 and 11. Grain and firewood may have been stored on the attic over these rooms. In room 9 was a large oven (inside diam. c. 4.50). Three layers of large tufa blocks have been preserved. They do not rest on a podium (unless the podium is still to be found below the present floor level). Four piers in the corners of the room carried arches and a concrete ceiling, necessary in view of the risk of fire and to contain the heat. In room 10 are the remains of another kneading machine. It is made of travertine. But this room was probably primarily used for the shaping of the dough into loaves of bread, over tables.

There must have been a stable for the animals who operated the machines, but its position is not clear. The bread was probably sold in the building, in the shops.

In area 8 a large number of bronze objects was found, that do not seem to belong to the building: simple vessels, lamps, two candelabra, a table support with a female herm (with some silver) and a bacchic mask that decorated the rim of a large vessel (situla) (Floriani Squarciapino 1949; the museum guide seems to suggest building I,XIII,2). Photographs are available in the Virtual Museum.

Photographs and Drawings


The row of mills in hall 8a, seen from the west. Photograph: Eric Taylor.

Detail of a mill in a mortar bed. Photograph: Eric Taylor.


Detail of a mill with an inscription. Photograph: Eric Taylor.

The kneading machines in hall 8a, seen from the west. Photograph: Eric Taylor.


Hall 8b, seen from the north-east. Photograph: Jan Theo Bakker.

The oven (9), seen from the east. Photograph: Jan Theo Bakker.

Proposed reconstruction of an attic in hall 8a, seen from the west. Drawing: Bernard Meijlink (Bakker 1999, fig. 20).

Caseggiato (Regio II – Insula VIII – II,VIII,9)

Plan of the building.  From Paschetto 1912, fig. 88.

This building, partially excavated at the end of the 19th century, is directly to the east of the Grandi Horrea (II,IX,7), across the road. The walls are of Hadrianic opus mixtum. In the centre of the building is a large round structure. The scanty remains consist of a tufa circle, without podium. It has been suggested that it was a basin, and that the building was a fulling mill.

The round structure may also have been an oven. On an old plan a basin, a floor of basalt blocks and a round object (a millstone or kneading-machine?) seem to be indicated, two rooms to the north. Therefore the building may also have been a bakery. Excavation of the north part of the building will be helpful.

A number of rooms across the street (part of a row of rooms along the east side of the Grandi Horrea) also had floors of basalt blocks.


The remains of the oven (?), seen from the west. Photograph: Jan Theo Bakker.

The substructure of the staircase in the central part of the building, seen from the south-east. Photograph: Eric Taylor.

Caseggiato delle Fornaci (Regio II – Insula VI – II,VI,7)

Plan of the building. After SO I.

The House of the Ovens was excavated in 1907 and 1912-1913. It is to the north of the House of the Painted Ceiling, from which it is separated by alley 11. It was built in the Hadrianic period (opus latericium). The original building was characterized by a large hall on the west side (hall 14-15), with wide doorways leading to the street. The hall was surrounded by smaller rooms. In the south-east part were an internal and an external staircase (rooms 8-9). The building seems to have had a commercial function.

In the period Antoninus Pius-Marcus Aurelius a bakery was installed in the building. Along the side walls of rooms 2-6 piers sustaining arches were added. They supported a ceiling at a height of c. 3.75 metres. A partition wall created halls 14 and 15. A staircase was added in the northwest part (room 17).

Floors of basalt blocks were present in the entire western half, and in rooms 3, 4, and 7. Large basins were built in rooms 5 and 15, a smaller one in room 7. Fragments of a kneading-machine and a millstone are still present in the building. The milling presumably took place in the large halls 14 and 15. In room 1 are two ovens. Therefore rooms 3 and 4 are ideally placed for the kneading of the dough. A dolium defossum in room 3 may have contained ingredients for the dough. Room 2, to the south of the ovens, is the obvious place for sieving and moulding the bread. The strengthening of the ceiling of rooms 2-6 is probably related to the storage of grain and wood on the first floor. Presumably the grain was poured into the mills from the first floor. Rooms 12, 13 and 16 may have been stables for the animals that operated the mills. The basin in room 7 may have been a small trough.

There must have been approximately ten mills in this bakery. A few of these were probably operated for the fire-fighters, whose barracks are directly to the east of the bakery. The latest masonry in the building has been dated to the late third or early fourth century (opus latericium and vittatum). In alley 11 a mediaeval lime-kiln was found.


Hall 14 seen from the south-east. Photograph: Jan Theo Bakker.

Room 6 seen from the west. Note the piers set against the lateral walls. Photograph: Jan Theo Bakker.

Fragments of machines for the milling and kneading. Photograph: Jan Theo Bakker.

The lime-kiln in the south-east corner of alley 11. From the north-west. Photograph: Jan Theo Bakker.