The Garden

While the house fell into ruins, both the garden, with its shrine and altar, and the monumental incarnation of room N seem to have continued to be frequented for some time longer. The final surface of the garden seems to have been bare earth or grass, as the Claudian surface probably was before it. Ceramic evidence from this surface (349) and from the upper destruction layers of room N (245) stretches into the second century A.D³¹. During this period, however, few changes seem to have been made. The only stratigraphic features associated with the use of the garden are a series of small holes filled with small stones, bone fragments, and in several cases charcoal or, possibly, ash (cuts 345, 355, 362, 363, 364, 365, and 367, filled by 344, 354, 358, 359, 360, 361, and 366, respectively). By analogy with the interpretation of similar deposits found in the small garden of a taverna at Pompeii, these small pits may be seen as indications of ritual activity, perhaps the burials of the remains of sacrifices associated with the shrine. Interestingly, the Pompeian garden was provided with benches similar to the bench constructed in the garden of the House of Diana³². Only one (344) of the fills of these pits produced datable pottery; it seems to belong to the beginning of the second century, but it is impossible to say whether the other pits were dug around the same time or whether they span the period from the collapse of the house to the mid-2nd c. A.D. Ongoing ritual activity in this space may also be indicated by the presence of a number of lamp fragments in 349. It is difficult to tell the extent to which the structural elements of the garden were maintained during this period, although the presence of 2nd c. pottery in a patch of pebbles and mortar (336) eroded from the podium of the aedicula shows that some of the structures in the garden were already slowly collapsing by that time. Ceramic evidence suggests that the garden, as a whole, was abandoned before the end of the 2nd c. A.D., as the last garden surface contained no pottery later than the middle of the 2nd c. and the material found in the various layers of collapse also bore a 2nd c. date³³. This interpretation is also supported by the coin record, which stops with Hadrian (catalogue n. 69, found in collapse 242).

The first part of the garden to fall apart seems to have been the decorative niche that replaced the earlier fountain. A number of decorative elements cannibalized from various Republican and Augustan structures were clustered around its base, perhaps waiting to be reused. These fragments probably represent some combination of collapse and deliberate stripping of reusable material from the niche, as the apparently deliberate assemblage was complemented by the dispersion of some of the other decorative elements and the presence of tiles and other material associated with collapse rather than spoliation. The deposit was composed of a layer of bright reddish soil (281) under which lay a friable brown deposit (301) that contained many of the heavier objects. In addition, another scatter of marble and terracotta decoration (340) lay under pis» collapse on the final garden surface between the niche and the enigmatic stone block to the south (pl. 8), and a layer of brown soil (282) containing roof-tiles as well as marble fragments had collected within the niche itself. The decorative elements in these various assemblages included opus sectile components of various shapes and sizes, fragments of Campana plaques, and a small, rectangular marble statue base with multicolored stone inlays (see III, p. 191ff). The material is somewhat eclectic, and certain elements may be related to other areas of decoration within the garden. Similarly, in the aedicula, before the collapse of the structure, were assembled a number of pieces of statuary and marble furniture probably deriving from the Claudian sanctuary, including the torso of the cult statue (227; see part III; pl. 60). These pieces were packed together into the cella in a haphazard manner that cannot possibly reflect their original arrangement in the garden. Their placement seems to be the result of an effort to clear out the garden and store the larger marble elements of the decoration in one place, perhaps in preparation for reuse (or even for consumption in a lime kiln). While it is thus difficult to confirm the date of the collection of this material, it is probable that the marble objects were gathered and deposited in the cella of the aedicula at some point after the garden had begun to fall apart.

[ 1 2 next ]

31. In the 1996 excavation season, the uppermost garden surface was designated 278. A winterÌs worth of precipitation substantially muddied this surface, and when excavation resumed in 1997, the layer that we removed as 278 was probably the product of the exposure of the garden surface to the elements. For this reason, 278 should be considered to equal 349, the surface we noted in 1997. It should also be considered to equal 341 and 351, two more products of the exposed surfaceÌs reaction to winter weather.

32. Fulford and Wallace Hadrill 1995-1996, 95-97; 100; 104.

33. The second-century presence in this area is supported by the similar dates of pottery fragments found in the wall collapse lying in the street outside the garden (373).