Patterns of Reconstruction at Pompeii

The following discussion examines aspects of the Macellum that indicate the chronology of construction for some of the damage areas. After reviewing these areas in detail, chapter 7 develops reconstruction scenarios for all the areas.

6.1 Area D

Figure 6.1: The juncture of W49.50 with the south wall.

Figure 6.2: The juncture of W49.50 with the south wall, with the repair area noted

On close inspection, it appears that the masonry of the shop wall bonds with the material of the repair, with the stones of the shop wall penetrating into the perimeter wall; figures 6.3 and 6.4 show detailed views from the west and east faces respectively.

Figure 6.3: Detail of the juncture of W49.50 with the south wall (view of the west face).

Figure 6.4: Detail of the juncture of W50.49 with the south wall (view of the east face).

The bonding of the shop wall with the repair means that the repair and that portion of the shop wall are contemporary construction: either both are modern, or both are ancient. The shop wall appears to be ancient construction. The material in the upper portion of the shop wall is consistent with the visible portions of material below it, and it bonds with the quoined brick at the end of the shop wall, which includes patches of apparently ancient plaster. If the shop wall is indeed ancient, then so is the repair that it bonds with. The damage corresponding to the repair was therefore not due to events in 79. The most likely scenario is that the damage in this area was due to the earthquake in 62, with the resulting conclusion that the opus reticulatum construction on the south face pre-dates the 62 earthquake.

This interpretation leaves unanswered the question concerning the location in the shop wall of the seam between pre-62 and post-62 material. There must be such a seam in this interpretation since it concludes that the upper portion of the wall is post-62, bonding with the repair material, while some lower portion of the wall is pre-62 construction. The plaster covering much of the lower portion of the wall makes it difficult to determine this seam. It is possible that much of the wall was damaged in 62 and largely rebuilt.

6.2 Area A

Moving to Area A, there are two items of evidence that indicate that the repair visible today is modern; the first is provided by the Pompeii model at the National Archaeological Museum of Naples, and the second concerns the blockage of a drain pipe a W6.7.

Looking first at the model at the Naples museum, figure 6.5 includes a photo of the model and a photo of Area A, revealing that the model delineates a distinct region corresponding to Area A without painted plaster. The photo of area A includes a yellow highlight indicating the seam between the repair above and primary material below.

The model at the Naples Museum (photo courtesy of Larry Ball)

Figure 6.5: The model at the Naples Museum and a photograph of Area A, highlighting the seam of the repair. Note the plasterless patch in the museum model corresponding to the repair area.

Figure 6.6 shows a clearer view of the south face of area A where the undergrowth is removed. Note that there are large patches of plaster along the bottom of the repair area, but no plaster on the repair area itself.

Figure 6.6: The south face of Area A.

A similar condition can be seen on the north face of the wall, figure 6.7 shows images generated from the photogrammetric model. The lower image has been enhanced to show detail in shadow, and annotated to highlight the seam between the primary material and the repair.

Figure 6.7: The south face of Area A.

Plaster is visible immediately below the seam in three of the four shops, but no plaster is visible on the repair area. The condition of plaster occurring directly adjacent the repair but not on it is consistent with the Naples model, and suggests either that the repair is modern (i.e. post-excavation), or that the repair is ancient but was never plastered.

Evidence pointing to a similar conclusion can also be found at the juncture of wall W6.7 and the north perimeter wall. Figure 6.8 below shows this juncture, with W6.7 on the left and the perimeter wall on the right. Note the red clay pipe embedded in W6.7.

Figure 6.8: The juncture of W6.7 and the north perimeter wall.

For the perimeter wall on the right, there is a clearly visible seam between the primary construction of black lava below and the large grey blocks of repair material above. Note that this seam is close to the top of the visible portion of the clay pipe. Figure 6.9 below shows a close-up view of the top of the pipe.

Figure 6.9: A close-up view of the pipe embedded in W6.7. Note that the pipe is blocked.

The figure clearly shows that the pipe in W6.7 is blocked by repair material, figure 6.8 shows that the repair of the shop wall appears to bond with the repair material of the perimeter wall, indicating that the repair of W6.7 and the perimeter wall are contemporary. The conclusion, then, is that the repair of the perimeter wall and the shop wall did not bring the building back to its original condition; plaster was omitted from the perimeter wall and plumbing was not restored in the shop wall. The restoration was incomplete.

There are two fundamental interpretations of the incomplete restoration: the repair is ancient, and is incomplete because there were other priorities at the time, or the repair is modern, and is incomplete because modern repair has no need for plumbing or plaster. Of these, the scenario of modern repair is more credible. There is little doubt that modern repair would have omitted plaster and plumbing, in contrast to ancient repair, where such omissions would be possible, but much less likely.

6.3 Areas A and E: Comparing corner treatment

Areas A and E both include a condition where heterogeneous repair material meets brick construction at a corner. Differences in the treatment of this juncture indicate chronology.

Figures 6.10 and 6.11 below show the south and north faces of Area E at the corner of the south gate. Note the quoining of the heterogeneous material and the brick.

Figure 6.10: The south face of Area E, at the south gate.

Figure 6.11: The north face of Area E, at the south gate.

Figure 6.12: The south face of Area A near the north gate.

Figure 6.13: The north face of area A near the north gate.

south face

north face

Figure 6.14: Details of the Area A corner juncture, from figures

At area A, the heterogenous repair material clearly does not quoin with the brick and block material at the corner. On the south face, the brick abutting the heterogenous material meets the heterogeneous fill in a rough seam, and the brickwork shows a significantly different size and color than the primary brickwork below. On the north face, the primary construction includes limestone block at the corner, but in the repair area above, the construction is completely heterogeneous, without blocks or quoining.

One simple explanation for the different treatment at the two areas is that the repair at Area E is ancient, while the repair at Area A is modern. At Area A, the brickwork in the repair area at the corner is very similar, though not identical, to the primary material below. Figures 6.15 and 6.16 compare the repair brickwork with the primary brickwork.

Figure 6.15: A close-up of brickwork on the north face of Area A, highlighting areas shown in figure 6.16

lower part

upper part

Figure 6.16: A close-up of brickwork on the north face of Area A. The left photo is of primary material in the lower portion of the wall, the right photo is in the repair area that quoins with heterogeneous material.

In the brickwork in the upper part of the wall, the construction is less regular, with more variation in the thickness of the mortar beds. The appearance of the mortar is also different, with the mortar in the upper brickwork containing black sand. The differences between the upper and lower brickwork at Area E suggest that they are not contemporary, but the similarities of the brick material and quoining pattern suggest that they are both ancient. In contrast, the sharp difference between the upper and lower brickwork at the corner at Area A suggest that the repair is modern: work intended to level the wall using highly different materials and methods.

6.4 Area B

Despite their apparent similarity, Areas A and B show very different patterns of damage. To compare these patterns, it is first necessary to look closely at Area B, since the region of damage is less clear than the other areas. Figure 6.17 shows two photos of the south face of area B, one is annotated to highlight the a seam in the materials.

Figure 6.17: The south face of Area B

Dobbins [1994, fig. 46] interpreted the highlighted seam as the boundary between pre-62 construction below and post-62 repair above, since the profile of the seam resembles the scooping profile of other damage areas; however, inspection of the north face of the wall and comparison with other areas shows that the profile of the repair probably lies elsewhere.

Figure 6.18 below shows two photographs of the north face of Area B, corresponding to the back wall of shop 9; the lower photo includes annotations highlighting two seams in the wall fabric.

Figure 6.18: The north face of Area B

The north face shows three strata of construction, rather than the two visible on the south face. In order to compare the north and south faces, Figure 6.19 shows the annotated photographs, with the photograph of the south face reversed, so that the seams in the two photos more closely correspond.

Note: image reversed for purpose of comparison.

Figure 6.19: Comparison of the north and south faces

The primary contrast of the north and south faces of area B is that the south face shows two layers of construction, black lava opus incertum below and mixed opus incertum above, while the north face shows three layers, black lava opus incertum at the bottom, mixed opus incertum above that, and a mixed construction including large stone blocks above. It is possible that portions of this upper region are a thin coating applied in modern times to stabilize the surface, however the portion including the large blocks appears to be repair; such blocks are clearly visible in the repair at Areas A, C, and E.

Appendix F includes a digital video clip derived from the photogrammetric model, allowing the reader to view the north face of the wall and interactively fade it away to see the corresponding features of the south face. Although the dimensional accuracy of the model is not high, the video treatment does confirm that the large blocks on the north face correspond to the area of plaster on the south face: this plaster masks the repair when viewed from the south, and also confirms that the repair is ancient.

The region of out-of-plane failure at area B is limited to the uppermost portion of the wall noted on the north face. The seam noted on the south face probably does not mark the profile of out-of-plane failure. The mixed opus incertum construction above the marked seam is not found at any other repair in the Macellum; instead, this material is consistent with the mixed opus incertum found below the repair at Area C, shown in figure 6.20 below.

Figure 6.20: The south face of Area C

Concluding that the repair at Area B is limited to the uppermost part of the wall raises the question of the meaning of the scooping profile of the marked seam on the south face and the distinctive patch of opus reticulatum just below it. The answer is that these features represent repair and remodelling operations on the building that pre-date the 62 earthquake; they are significant with respect to the overall chronology and development of the building, but have little bearing on the question of its post-earthquake reconstruction.

Another question raised by this conclusion concerns the significant difference in the pattern of damage between Areas A and B. At area A, the damage region is much lower, approximately one meter above interior grade at its lowest point, while the damage region at Area B is about three meters above interior grade. The key difference between the two areas is the presence of internal walls parallel to the perimeter wall at area B. These walls, W9.10 and W11.12, serve to buttress the shop walls against out-of-plane failure, enabling the shop walls to remain as effective buttresses for the perimeter wall. In particular, if there were stairs along the back of the shops, as there were at Area A, then these interior walls would have helped prevent the progressive collapse mechanism described in section 5.5.1; although the stair opening would have interrupted the connection between the second floor framing and the perimeter wall, the framing would still have connected with the interior walls parallel to the perimeter wall, and these walls offer much better resistance than the shop walls alone.

6.5 The Western Wall

The western portion of the perimeter wall does not show signs of out-of-plane failure, however the unusual double-wall construction of this wall and the modern rebuilding of portions of its exterior make it difficult to interpret the masonry clearly. One of the questions raised by this area is the reason for the double wall. One interpretation is that the double-wall construction reflects two priorities and stages in the post-62 reconstruction: first, to restore the building to operation; and second, to expand the building and create a stronger architectural presence on the Forum.

The interior layer of the west wall corresponds to the first priority, returning to operation. The nature of the repair at Areas D and E indicates that return to operation was more important than restoration of architectural quality, and this priority probably prevailed in the initial reconstruction of the west wall. Any damage to this wall was probably repaired quickly to restore the market building to operation after the earthquake.

The exterior layer was completed some time after the inner layer, as concluded by Dobbins [1994, pp. 680]. The outer facade may have been done years after the inner layer was completed, and was not so much a repair of damage as an improvement and expansion to an already functioning building.

6.6 Summary

The conclusion that the repair at D and E is ancient leads to another important conclusion, since it means the wall failed during shaking in 62 AD, but the currenlty visible ancient fabric survived shaking in 79. This result suggests that shaking in 62 was more severe than that of 79.