The House of Diana on the Forum

The house is entered through a narrow vestibule between two shops which opened directly onto the forum plaza. Traces on the signinum pavement suggest benches on either side. From this standard beginning, the rest of the structure could be drawn straight from Vitruvius. The vestibule is separated from the fauces by a gap that probably held a threshold block. The fauces slope up steeply and then open into a rectangular impluviate atrium. Substantial falls of ceiling plaster and the rain damage within the impluvium leave no doubt that the atrium was roofed in the Tuscanic style, with the main beams running across the atrium. On the right are two cubicula of equal size. On the left the single, large room seems to throw off the symmetry, but its blocked doorway indicates the earlier division of this space into two corresponding cubicula. At the back of the atrium, to both left and right, extend alae, one of which is shortened to insert a room floored in beaten earth which was probably intended for storage -- or, perhaps, for the wooden cupboard which would have contained the imagines of the ancestors of the owner. The line of sight continues through the atrium to the tablinum, a wide room whose dimensions seemed to have served as a module for the rest of the house. To the right of the tablinum is found the blocked doorway of a triclinium, which probably ended along the line of the rear wall of the atrium. To the left of the tablinum is the opening of an andron, the corridor that connects the atrium and garden in many second-century houses. A suite of service rooms lies to the left of the andron, paved with opus signinum. To the rear of the structure was found a building still imperfectly understood. A raised cistern in the rear corner of a functional floor in opus spicatum, with a drain into the cesspit below the floor, suggest that the structure constituted private baths.

The rest of the space between the back of the house and the street to the south was occupied by a garden. This was cut down into the bedrock of the hill almost two meters. The lowest levels above bedrock show much ash, probably from the fires used to crack the rock. These earliest garden layers contain only black glaze pottery, demonstrating that the earliest phase of the house can be ascribed to the Republican period. However, until the black-glazed pottery associated with the first building phase has been examined, we must rely on Brown's dating for all of these buildings on the forum, which suggests that they were constructed in the 170's B.C. I myself would be inclined to push them back to the beginning of the century, when a new deduction of colonists arrived at Cosa.

The Plan of the Colony of 197 B.C.

It is fair to propose that this plan was reproduced in the buildings on all three sides of the forum. What we can see of 'Atrium Building II,' to the southwest of the entrance to the forum, conforms exactly to this scheme, but for the gardens and service areas. However, there is space for these in the block beside the building, which separates it from the large cistern. The same is true for Atrium Building I, where a plot of the same size separates the building from street 7. All of the other buildings conform exactly to the plan recovered in our excavation of Atrium Building V.

The presence of these large houses on three sides of the forum is, however, an anomaly, at least if it is compared to what Russell Scott's excavations have shown about Republican housing elsewhere in Cosa. Near the museum, the long insulae were divided into house plots 8.5 meters wide, just under half the width of those on the forum plaza. The basic layout of these is similar to what we have seen on the forum, but on a smaller scale. The houses measure half the width of those on the forum. We are dealing, almost certainly, with houses for two classes of colonists, some of whom received plots exactly twice as large as the others.

The Plan of the Colony of 197 B.C.

Now, if we examine the plan of the city as elucidated by Brown it becomes clear that this hierarchy is inscribed in the streets as well. The ordinary streets are 6 meters wide. However, the street leading from the forum to the arx, and the street which leads from the southwest gate to the Eastern Height, are 9 meters wide. Brown called these processional streets, and certainly by the middle of the second century they both led to temples. However, their extra width may also indicate that housing of a higher status lay along them. If we pursue this hypothesis the following plan in fig. 2 results, with the areas occupied by the larger houses colored a darker grey, while those occupied by the small houses are a lighter color. The best estimate of their numbers is 20 larger houses, with between 200 and 205 smaller ones. The proportion is very close to 1 to 10.

The Development of the House and its Garden

Reconstruction: The Garden in the Early First Century B.C.

Atrium Building V, like its cousins at Pompeii and elsewhere, did not retain its early second century plan for long. Sometime in the Republican period a fountain niche was built against the back wall of the garden, canceling out one or two of the large tree or shrubbery pits. There seems no doubt, however, that the garden in this phase served principally as a kitchen-garden. The space occupied by the shadows from the high walls was used for a path, lined with trees, while a compost or manure heap was piled up in one corner. None of the back rooms opened onto it, and access to it and the little bath building was through a door in the andron. There may have been a stair leading out to the street, as there was in a later stage -- although we have no trace of it, the position of the fountain is curiously asymmetrical if we are to assume that it was alone on the wall.

Some time after 71 B.C. a major disaster appears to have struck Cosa, and houses and public buildings all over the site seem to have been abandoned. Brown attributed this to pirate raids, while the aftermath of the Cataline conspiracy is another possibility. Our building may have been abandoned at the same time. Certainly it was rebuilt during the Augustan period, although most of the wall-lines were taken up by the new building. The most important changes were to the rear of the house.

Reconstruction: The Garden in the Late First Century B.C.

The triclinium was opened up to the back, and a loggia constructed along the back wall of the house. Along its edge ran a fine gutter, leading into the cess pit under the bathing wing. The service wing was joined to the small bath building, creating a continuous suite of rooms linked by a corridor along the line of the andron. Access between the atrium and the triclinium was blocked, making the triclinium into a far more private space, closely related to the garden rather than to the public areas of the house.

The whole of the house was newly decorated in this period: the tablinum and triclinium were given simple black and white mosaics, while the rest of the house was paved with decorated signinum floors. In the garden, a new layer of soil replaced the old bed -- again, it was carefully prepared, sieved and manured. Somewhat later, a u-shaped trench, was dug into this bed. We have interpreted this as bedding for a hedge, which seems to have framed an area intended to be viewed from the triclinium. Possibly at the same time a lot of garden furniture was acquired: a capricorn table leg, another table leg in the form of the leg of a fawn, a whole series of herms, and oscilla to hang between the columns of the loggia: this one, decorated with a satyr mask, and this one here, in the form of a mask. The garden was now most emphatically a decorative space, rather than a working kitchen-garden. The rather fine statuary must have left it almost crowded, more like a would-be villa garden than the restricted back garden of a town house. The last major changes occurred towards the end of the Julio-Claudian period, perhaps at mid-century.

Reconstruction: The Garden in the Julio-Claudian Period

Even more striking is the transformation of the garden into a sanctuary. The loggia was now partially blocked off, and a new door opened in the service wing, where the old bath-suite seems to have been transformed into a large, square room. Directly opposite this entrance was constructed an aedicula, or small shrine. A flight of stone steps was constructed in the back corner of the garden, while a bench, perhaps for those assisting at a sacrifice, was built along the wall facing the little shrine. A broken column seems to have served as an altar. The painting on the walls was retained, the shrine itself being decorated with a trellis pattern. The fountain was redecorated, however, its basin blocked and its niche carefully lined with pasta vitrea and shells. Directly in front of the fountain were found three fragmentary campana plaques each of which shows a dancing maenad. This seems to suggest that they were reused in this context to decorate the surrounds of the fountain itself. In front of it a basin on two marble legs, decorated with patera and bucrania, seems to have replaced the earlier stone basin.

Inside the shrine was found a statue base, and, lying on it, a statue of Diana. This is a fairly crude bit of work, the head and hands far too small for the body. The head fits in to a generic type for Diana, and it is probable that the right arm is raised to reach back into her quiver, whose strap is evident across her breast. With the statue was found one half of an inscription: a month of digging passed trying to complete the hexameter, but luckily the other piece showed up in the last week and gives us the full text. It reads, "Cornibus His Augusta Sacris Diana Recepta Ad sua testatur templa patere viam." This translates as: By these sacred horns Diana Augusta has been welcomed, she testifies that the way to her temple is open.

The construction of the shrine was the last action to take place during the life of the house. Destruction deposits cover it, and in these the absence of any pottery later than the 80's suggests that this was the terminus ante quem for its occupation. This agrees with evidence from the rest of the town, where only a few buildings seem to have remained in occupation after the end of the century.