The general description of the Garden Houses is mainly a summary of two articles. Most of the text is based on:
- Axel Gering, 'Die Case a Giardino als unerfüllter Architektentraum. Planung und gewandelte Nutzung einer Luxuswohnanlage im antiken Ostia' ('The Garden Houses as unfulfilled architect's dream. Planning and changed usage of a luxurious residential complex in ancient Ostia'), Römische Mitteilungen 109 (2002), 109-140.
The description of the water supply is based on:
- Saskia Stevens, 'Reconstructing the Garden Houses at Ostia. Exploring Water Supply and Building Height', BABesch 80 (2005), 113-123.
We also consulted:
- M. Floriani Squarciapino, 'Ornamento per fontana da Ostia', BdA 37 (1952), 29-32.
- Scavi di Ostia IV.
- Thea Heres, Paries, Amsterdam 1982.
- V. Scrinari - M.A. Ricciardi, La Civilta dell'Acqua in Ostia Antica, Roma 1996.
- Rina Cervi, 'Evoluzione architettonica delle cosidette Case a Giardino ad Ostia', in L. Quilici - S. Quilici-Gigli (edd.), Città e monumenti nell'Italia antica, Atlante tematico di topografia antica 7 (1998), 141-156).
We are not able to read the following book, because it is written in Basque:
- E. Subias Pascual, La domus dels Dioscurs d'Ostia Antica, Tarragona 1993.
Click here to open a plan of the Garden Houses in a new window (Cervi 1998, fig. 2).
The Casa delle Ierodule (III,IX,6) (also called Casa di Lucceia Primitiva) is discussed on a separate page.
The Garden Houses (III,IX) were excavated in the years 1938-1942. Since then they have often been mentioned as an example of rational planning and of progress in residential architecture in the second century AD. Gering maintains, that the original concept needed modification already in the second century. In the third century there was a shift from residential to commercial usage. At the end of the century many buildings were destroyed by an earthquake and fire, an event followed by some repairs and rebuilding.
Unfortunately no stratigraphic information was recorded during the excavation in the 1930's and 1940's, although some stratigraphic evidence was added by further excavation in 1965 and 1967 (Casa delle Pareti Gialle), and 1969-1970 (buildings III,IX,3-7). It has not been possible to establish a chronology for all of the additional walls. Therefore Gering focuses on the functional relations between the modifications. Gering has written a synthetic study. In this study he re-examines absolute datings of masonry proposed by Thea Heres, and aims at complementing the study of the complex by Cervi.
The date of construction of the Garden Houses can be established fairly accurately. Brick stamps have been found from the years 123-125 AD. The oldest wall paintings and (geometric black-and-white) mosaics have been dated to the years 130-140 AD. A little is known about earlier buildings in the area. Buildings III,IX,25-26 were built during the reign of Trajan. In front is part of a porticus of thick travertine columns, perhaps from the first half of the first century. In 2006 a trench was open to the south-west of building III,IX,15, containing travertine drums of columns and a wall in opus mixtum (Trajanic?), with a different orientation than the Hadrianic buildings. In the walls of the trench was much charcoal, suggesting that the earlier constructions have been destroyed by a fire (the Fasti mention a large fire in Ostia in 115 AD). In the Hadrianic period the level was raised considerably.
The Garden Houses were conceived primarily as a luxurious residential complex, with expensive apartments. It is located near the ancient shore-line. The ground floor consisted of 16 medianum-apartments, a domus, and many shops. The apartments were shielded from busy streets by open spaces and rows of shops. They have the usual representative rooms with a height of two storeys. They have inner staircases, leading to the first upper floor. Outer staircases led to further floors. In the centre of the complex is a pair of two times four apartments (III,IX,13-16 and 17-20), back-to-back. The apartments have a ground floor area of c. 220 square metres, but the use of two apartments by one family was planned: there are doorways between the apartments.
The central apartments were surrounded by an unpaved court, presumably a garden. In the garden are six large water-basins. Presumably they were originally covered. Water came from two spouts. Around the basins were gutters with depressions, in which buckets may have been placed. On all four sides of the garden are further apartments (III,IX,1 (the later Domus dei Dioscuri) / 3 / 4 / 6 (Casa delle Ierodule) / 8 / 10 / 12 (Casa delle Pareti Gialle) / 21 (Casa del Graffito), shops (III,IX,2 / 5 / 7 / 9 / 11) and the Domus delle Muse (III,IX,22). The area to the north-west and south-west has not been excavated. To the south-east is a street (Via delle Volte Dipinte), to the north-east is an unpaved strip, then a row of shops flanking the Cardo degli Aurighi (III,IX,23-26). The main, ornamental entrance (vestibule) was in the south-east side. Above an entrance in the north-east side is a brick relief of a stick, a parallel for which is found in the Caseggiato dei Misuratori del Grano (I,VII,1-2). The thickness of the walls is 60 cm., suggesting a total of four storeys and perhaps 1200 inhabitants. The quality of the floors of the storeys varied: fragments were found of opus spicatum and mosaics.
Water supply and building height
Stevens has confirmed the suggestion that the central apartments were four storeys high (c. 17.70 metres or 60 Roman feet). She concludes that the upper floors of these apartments were supplied with water. According to her rectangular recesses contained terracotta pipes for drainage and lead pipes for the supply of water. The latter recesses are 30 cm. deep and contain a sediment resulting from leaking water. The recesses for drainage are smaller, and contain no sediment. Buildings 17-20 have significantly more recesses than 13-16. In the former complex lead pipes were installed in the corridor between apartments 17-18 and 19-20, and in the east wall of the staircases to the west of the corridor. Stevens suggests that toilets were a standard facility on the third floor. Buildings 17-20 may have had fountains on the upper floors. In most of the buildings of the perimeter there was no private water supply. The inhabitants of these buildings must have used the six basins in the garden.
The first group of modifications
The first modifications in the complex ante-date the Hadrianic mosaics. The doors between the central apartments, creating apartments of double size, were blocked. Shops became part of houses (Domus delle Muse, house III,IX,4 combined with 5).
On Gering's general plan, fig. 19, a niche is indicated in the north-east wall of the southernmost shop in the Domus delle Muse. It is interpreted by Gering as a lamp-niche. In reality the niche is in the shop to the north. It has been interpreted by me as a cult-niche (Bakker 1994, 86, 224, pls. 38-39). In the vault are the remains of a painted figure, perhaps a Lar. I have argued that lamp-niches did not exist, with an exception in the Pompeian Terme del Foro (Bakker 1994, 14).
The second group of modifications
In a second phase, after the first decoration had been made, many doors were added in the apartments. The doors are of a similar type and have similar measurements. In this way annexes were created of representative rooms, or representative suites. Of a completely different nature are reinforcements (supporting piers, blockings of doors and windows), necessitated by problems with the foundations, an earthquake or a fire. They are found in the east part of the complex (III,IX,10 / 11 / 12 / 22). The new masonry is covered by new paintings, dated to the second half of the second century.
The third group of modifications
In a third phase, shops were connected with each other in several units (III,IX,2 / 7 / 9). It is possible that many bases for ladders in the shops, leading to the living quarters of the shop-owners, were now removed. It is not clear whether this is an indication of increased wealth, or of influence of the guilds.
The fourth group of modifications
In the Severan period new paintings were made in the Casa delle Pareti Gialle and Domus delle Muse. In the medianum of III,IX,3 a fountain-niche was added in an axial position. The floors of the medianum and of the room opposite the niche were decorated with marble.
In the garden a free-standing structure was built, between III,IX,1 and 13. On the floor is an irregularly shaped mosaic (preserved measurements 2.90 x 3 m.) from the Severan period. The central panel is black-and-white, the border is polychrome. A Nilotic scene is depicted, with crocodiles, pygmies, and flying amorini, the latter at a slightly higher level. The mosaic was surrounded by a wall of opus latericium. According to Becatti and Ricciardi this was a basin, possibly with fountain niches. According to Floriani Squarciapino and Ricciardi a sculptural group, found in a nearby room, may belong to it. The group is made up of a seated, naked fisherman and of fish in the sea. It has two large holes, through which water may have passed. It is not clear why Gering calls the structure a summer-triclinium (dining-room).
The fifth group of modifications
Some modifications could be traced only with difficulty, because many doors of varying width and height, that were made at a later stage, were blocked again when the excavators restored the complex. These secondary doors can for the most part only be seen on old photographs. The doors indicate, that shops and workshops were installed in many apartments, especially the central apartments, in the later third century. So, for example, a representative room in the central apartment III,IX,14 was changed into a shop, witness a shop-threshold and a door 2.60 wide (originally a window). Wooden shelves were inserted in L-shaped grooves, for example in the central apartment III,IX,17.
So far nobody has studied the grooves and holes in the walls of the Ostian shops, although they may contain a clue to the function. A good start could be a book by Stephan Mols, mentioned by Gering, Wooden Furniture in Herculaneum, Amsterdam 1999. See p. 55ff. and pls. 148-166 for cupboards and racks.
Damaged mosaics were 'repaired' with opus signinum, paintings were covered with a thin white layer. A basin in one of the rooms of III,IX,4 points to commercial usage. This changed usage explains why many graffiti of numbers and drawings can be found. Most of the Severan marble floor in III,IX,3 had been removed before the building collapsed in the late third century.
The Domus delle Muse, Casa delle Pareti Gialle and Casa del Graffito remained luxurious dwellings. The mosaics were still repaired with polychrome marble, and paintings imitating marble were made.
The Garden Houses were largely destroyed by an earthquake, triggering also a fire. Traces of fire were found in the destruction layer, and can still be seen in re-used fragments of upper floors. Torsion-cracks testify to the earthquake. The masonry was lifted by the earthquake and then fell. Often one part of a wall was lifted in a different way (higher, lower, etc.) than another, so that after falling it would for example protrude from an adjoining stretch. Such cracks can be found throughout the Garden Houses. The latest coins found in the destruction layer belong to the reign of Aurelian (270-275 AD). They were found below collapsed ceilings during further excavations in buildings 3-7, in 1969 and 1970. The find of statuettes and pottery proves, that the buildings were not deserted at the time of the collapse.
It is clear that the earthquake affected the entire complex, but detailed evidence is available only for the buildings that were excavated properly, in 1969-1970. It could then be established, that buildings III,IX,4-8 were never rebuilt at the same level. The upper floors had collapsed, or had become instable and were torn down. The rubble was then used as a new foundation, 90-200 cm. higher (explaining why paintings up to a height of 2 m. have often been preserved quite well). Walls that were built later start at the new level.
The sixth group of modifications
According to Gering many thin walls were built at the higher level directly after the earthquake. Many doors and windows were blocked. A row of brick piers was built in the north-east part of the courtyard. Gering maintains that the House of the Dioscures was also built in the late third century, not in the first quarter of the fifth century, as suggested by Heres. He refers to the repair of a crack that must in his view have taken place shortly after the earthquake. Unfortunately he does not provide enough details to be convincing (the arguments for a date in the fifth century: Heres 1982, 136). From now on most of the buildings in the complex had only one upper floor. It may have been struck by another earthquake in 346 AD.