The House

Though the moment of abandonment of the house is fairly well-established, the time frame of the collapse of the building itself is not entirely clear. It is possible that the house could have come down as a result of a seismic event or a serious fire; an abrupt collapse may be indicated by the discovery of both the masonry jambs (64, 173) of room G fallen face-down and intact, complete with partial plaster covering, on the floor of the atrium. In addition, a layer of charcoal 1-2cm. thick on the floor of room F (170) may represent wooden furniture burned during a violent destruction. On the other hand, there is no overall trace of burning, nor is there a literary record of an earthquake in this area in the later 1st c. A.D. In addition, the fact that the garden walls did not fall for another hundred years might argue for a more gradual process, and the charcoal in F could indicate nothing more than a fire lit by squatters in the abandoned building.

In any case, parts of the house seem to have stood for long enough for the better building material to be despoiled. The massive doorsill between the vestibule and the fauces seems to have been removed during this period, leaving a compact brown earth backfill (196). Similarly, a doorsill seems to have been removed from the entrance to room G. Finally, a worked stone sill belonging to the door between corridor M and room Q was also removed. At the same time, material began to filter into the disused cistern, beginning with the bucket used to draw water: its metal handle was found lying on the cistern floor at the bottom of the draw shaft.

The subsequent ruin of the house took place according to the classic model for the collapse of buildings constructed primarily in pis» or rammed earth²⁹. First, the wooden beams of the roof rotted and the roof fell in. While many of the destruction layers were disturbed by later activity, an excellent example of this stage was preserved in the tablinum, where directly over the mosaic floor was a layer of broken tiles and fragments of ceiling plaster (76). The ceiling plaster bore polychrome painted decoration on one side, while on the other was visible the impression of the reeds to which the plaster had been affixed (pl. 74). Similar deposits were present in rooms Q (326; pl. 11), H (also 76) and K (144), although in the latter the collapse had been somewhat disturbed by the roots of a large olive tree just to the northwest of the house.

Next, with the roof gone, the tops of the walls were exposed to precipitation and began to deteriorate. In several places, it is clear that the first step in this deterioration was the separation and collapse of the wall plaster coating the pis» core. Intact sections of plaster collapse were found in several rooms: one was located in the front part of room C, somewhat mixed with stucco ceiling decoration; another was found in the southeast corner of the atrium (14, covering assemblage 80; pl. 59); and a third was excavated in ala H (77). After the plaster had peeled off, the walls themselves would have slowly melted, slumping down to cover the floors and socles in what would eventually become a fairly even layer. By the modern period, most of the in situ collapse had been churned up by later agricultural activity, but the remnants of the original pis» covered pavements or roof collapse in several rooms. Where the archaeological deposit was shallower, in the northeastern part of the house, intact pis» collapse was located mainly along the walls, with some spots left just over the pavements. In the rooms to the southwest, where the ground rose and deposits were deeper, some layers of plaster or roof collapse were still covered by a thin layer of intact wall material. Within the main part of the house, in situ pis» collapse was excavated in the following areas: along the walls of the vestibulum and fauces (146, in which was also found a coin of Nero, catalogue n. 61); in patches along the walls and near the floors of the atrium and room E (8); in room D, along the southeast side of the room (167); in room F (168); along the walls of rooms G (74) and H (75); over the roof collapse in J (135); in corridor M (221); in room P (163)³⁰; in rooms P, M, and R (295); and in room Q (276). Roof and pis» collapse were also present in the cistern, having fallen down the draw shaft, but the cistern seems to have remained open for some time and the action of the water that accumulated in it made different layers of collapse difficult to individuate. In much of the house, then, the years around the end of the first century seem to have marked a fairly rapid period of collapse. For reasons that will become apparent below, it is likely that the lower stretches of most of the walls of the house remained at least partially visible above the ruins of their former pis» elevations.

29. For the formation of this sort of deposit see Fentress et al. 1981.

30. In room P, 163 covered a thin deposit (159) that was initially interpreted as soil accumulated during the period of abandonment, but the lack of similar layers in the other rooms of the house suggests that 159 may be another layer of pis» collapse or collapse associated with the cooking platform.