Field Season 2017
During the summer of 2017 with the kind permission of the Ministero dei Beni e delle Attività Culturali e del Turismo e la Soprintendenza Soprintendenza Speciale per i Beni Archeologici di Napoli e Pompei and with vital assistance from Prof. Osanna, Dott.ssa Stefani, Dott. Galeandro, and Assistente Sabini, members of the Via Consolare Project conducted archaeological excavations, survey, and documentation in the area of the Villa delle Colonne a mosaico as part of a multi-year study of the chronology, development, and nature of urban growth of the properties along the Via Consolare, stretching from Pompeii’s surburbium to its forum. Research was undertaken in 2017 by completing those trenches initiated in 2016 (AA012 and AA013), situated respectively in a corridor or cryptoporticus on the western side of the Villa core and within the triangular space at the back of tombs 6, 7, 8, and 9 on the northern side of the Via dei Sepolcri. This summer also saw the opening of two new trenches, located in the viridarium (AA014) and southern' decorated fauces of the Villa (AA015). (Figs. 1 and 2). Topographical survey focused primarily on recording these trenches and the features found within them. Study of the pottery recovered during this and previous seasons also progressed in tandem with excavation. The results of this work have augmented appreciably the narrative of the Villa and its evolution as discussed in the report of last year, revealing a longevity and continuity of use in some areas of the Villa that seems to have survived the transformations that modified the function of space elsewhere in the area of the Villa delle Colonne a mosaico. Furthermore, the chronological development of the Villa is now much clearer, especially in the areas of the viridarium and the area behind the tombs, which have now been confirmed as relatively late additions, extending across the Samnite cemetery that underlie them only in the early Imperial period. Indeed, the periodic expansion of the Villa through the acquisition of nearby space is also documented by the relationship between the Villa and its dependent shops. The following brief report presents the preliminary results of this research and current interpretations at the end of the 2017 field season.

The main goals of this season’s archaeological research were the following:
1. Completion of sub-surface excavation between the Villa delle Colonne a mosaico and the shops that flank the Via dei Sepolcri begun in 2016 (AA012).
2. Completion of excavation in the triangular area behind northern tombs 6, 7, and 8 and documentation of the surviving stratigraphy in this area between the trenches executed by A. Maiuri in 1935. Expansion of the trench of 2016 (AA013), in order to permit descent to a greater depth in the area of the large drain found underlying tomb 7, which was a such a depth that it could not be fully investigated.
3. Exploration of the central area of the viridarium in search of traces of the impluvium and drains mentioned in the previous explorations of Spano published in 1910.
4. Cleaning to any surviving pavement in the decorated southern fauces in order to record it and explore the stratigraphy below it in any areas where it might not survive.
5. Correction and filling in details of the 3D model of the Villa and its dependencies produced in 2015 and 2016, combined with complete survey of excavated deposits using Total Station Survey. Use of open-source Structure from Motion (SfM) technologies to provide a complete, volumetric 3D record of on-going excavations.
6. Recording, processing, and analysis of small finds and ecofacts recovered from the current trenches in concert with excavation in order to provide immediate chronological feedback and to speed the process of publication.
7. Recording, analysis, and processing of pottery recovered in previous seasons remaining to be studied systematically.

The great majority of these goals were achieved successfully during the summer of 2017. While no trench reached natural soils, and two were rendered more difficult by the prevalence of previous modern interventions, sufficient information was recovered from each to establish much of the latter history of the Villa, especially regarding its expansion to include the areas of the shops, and the later inclusion of the area of the viridarium and the area behind the tombs. This expansion involved not only the acquisition of a large amount of space, but also involved the transformation of the area from a Samnite cemetery into a garden and Roman funerary garden. The 3D model of the Villa and its surroundings is now virtually complete, and includes all excavated areas this year. 3D Structure from Motion scanning has continued to open new possibilities for excavation recording, including areas that could never be scanned by traditional time-of-flight scanners. Analysis of materials continued to move forward at pace this year, with the processing of large amounts of pottery both from Insula VII 6 and from the area of the Villa delle Colonne a mosaico, continuing to finalise our study in some of these areas in preparation for final publication.

Archaeological Area AA012
The 2017 field season saw the reopening of some of the area investigated in the 2016 season, including the western room (room 19) of the two-room space (rooms 19 and 20) between the shops along the east side of the Via dei sepolcri and the western defining wall of the Villa della colonne a Mosaico. The western room of the two was specifically chosen because of the final deposit exposed in 2016, which demonstrated a large quantity of ceramics. This report will amend the conclusions presented in the 2016 report and align the present investigations with the phasing of the Villa as established in the VCP 2015 report.

Phase 1 – Earliest Traces – No Evidence
As was the case in 2016 and 2015, the high elevation of Plot 4 room 19, roughly equal to the second story of the shops, prohibited reaching natural deposits. As a result, no information is available about the phases of this area prior to the arrangement of space established by the construction of the shop walls in phase 4.

Phase 2 – Early Villa Construction, the “Rustic” Core - No Evidence
The initial construction of the Villa core, the wall analysis and phasing of which was discussed in detail in the 2015 report, was not directly investigated through the excavation of AA012 in 2017.

Phase 3 – Augustan Adornment, No evidence
This phase of adornment to the Villa, which impacted the area of AA005, appears not to have involved the area of AA012.

Phase 4a – Villa Expansion and Shops
As discussed in the 2015 and 2016 reports, phase four of the area saw a radical change in the scope and scale of the Villa and its surroundings. Likely seeking to exploit the extra-urban thoroughfare that proceeded to the city via the Porta Ercolano, the Villa underwent a significant reorientation in the fourth phase of its construction to incorporate the street front. During this time, a series of double-height vaults were constructed around the northern and western sides of the Villa, presumably to support the vertical extension of the Villa above the level of the Villa’s original core. At the north side of the Villa, two sets of vaulted rooms were added, while on the west side, a single hallway was created by vaulting the westernmost wall of the Villa precinct (MC.04.133/MC.03.004) to the west. The Villa precinct wall on this west side was constructed on an angle that matched the alignment of the street and not that of the Villa. As discussed in the reports from 2015 and 2016, the difference in the alignments of the Villa core and the western range of rooms was taken up by a series of irregular spaces – a triangular room immediately to the east of Plot 4 room 20, and a cistern to its north (Fig. 6).

The shops that define the eastern side of the Via dei sepolcri (Plots 4 and 5) are believed also to have been constructed at this time, also using heavy black lava in their construction that is characteristic of this phase of additions to the Villa. The shops had the vertical height for the allowance of two stories, but whether the vertical height of the spaces was actually divided into two stories, and if they were, the manner in which the upper story was created is not yet clear. Most of the shops’ back rooms preserve barrel vaults that are believed to have been added in phase 5, obscuring the lateral wall constructions and evidence for earlier methods of spanning the spaces for an upper story. The back room of shop 18 (Plot 4 room 12), however, may be an exception. The barrel vault that spanned this space does not survive beyond a vault-shaped wall scar and the remains of the eastern vault spring. With the vault’s absence, a series of beam holes in the northern wall that were contemporary with the primary construction of the wall are visible (Fig. 7). That is, the mortar of the wall was poured around the beams to create the holes when the wall was constructed, not chipped out of an existing wall in order to add beams. One large beam hole in the southern wall suggests the receipt of the beams on the opposite side of the room, but the rest of the wall does not survive at a corresponding elevation. The presence of these beam holes, constructed at the same time as the general build of the wall, would suggest that the first phase saw a second storey of supported by beams, later replaced with vaulting - precisely the reverse of that suggested by Kockel's 1983 research, which saw the beam holes as the replacement for vaulting that had collapsed in the earthquake of AD 62/3. Perhaps there was a similar arrangement for all of the shops along the Via dei Sepolcri, as suggested by the VCP’s work in previous seasons.

During this first phase of the Villa expansion, the spaces comprising AA012 were most likely intended to serve the Villa, rather than the shops. The back walls of the shops’ ground floor rooms were uniformly lacking doorways to the east, so access to the intervening space was not intended from these properties. Instead, access was achieved through the then-open doorway between the Villa courtyard and the long hallway (Plot 4 room 20/21) that ran down the length of the space between the Villa core’s western wall and the intermediary spaces created by the disjunction of the street and Villa’s alignment (Plot 3 rooms 13 and 14). The western room (Plot 4 room 19) also served the Villa, as demonstrated by a doorway through the wall that divided the rooms (MC.04.127/129, wall SU 12.004), providing a connection between the western room and the hallway. This doorway, as far as can be discerned from the depth reached by 2017’s excavations, was a very tall portal (Fig. 8). Neither threshold nor lintel was found for the portion of the doorway exposed, suggesting that its base may have reached the lowest point of the room’s ground floor below the elevation exposed, and that its lintel was above the surviving height of the walls. Perhaps the western room was, like many of the vaulted spaces built in this time period, a double-height room. In support of such a spatial arrangement is the presence of the second portal through this same wall, seemingly a window just north of the doorway to allow light into the western room from the hallway (Fig. 8).

Another potential consideration is that the northern half of Plot 4 room 19 was never meant to be an accessible space, perhaps leaving only the south side of the room as an open room on the ground floor. A well-compacted soil that was resilient enough to be cut into in a later phase for the construction of at least one small wall was present in the northern half of the western room, and although it was not removed in its entirety, had an excavated depth of at least 0.60 m. It is likely to have filled the entire ground floor’s elevation to bring this half of the room to a second story height. Perhaps only the southern portion of the room possessed an accessible ground floor, while the northern half of the room did not have a lower story.

Phase 4b
Subsequent to the establishment and laying out of the initial arrangement of the shops, a few changes were carried out to alter slightly the arrangement of space. A brown mortar held together a motley assortment of building materials and was used to fill in a portion of the southern doorway (see Fig. 8). It may be that this construction was related to the installation of installation of beams, the holes for which were chipped into the north face of the southern wall of the west room (MC.04.124) (Fig. 9). These beams must have been supported on their north end by a wall or construction that does not survive or is no longer evident. It may also be that the northern half of the room, if it were also open in the first phase of the space, was filled with the hard-packed material in this phase, retained by a construction that the 2017 excavations were not able to reach, and supported the northern ends of these beams. The purpose of this further division of the space remains unclear.

Phase 5 – Reconfiguration of Space
Phase 5 in both the Villa and the shops is identifiable by the use of a very brightly-coloured, but generally poor quality yellow mortar used in many constructions. In AA012, the use of this mortar is particularly identifiable in two constructions: a wall built out of fragments of opus signinum floor that sealed the southern end of the hallway (Plot 4 room 20, MC.04.121/122), and a second filling construction to close the doorway in the dividing wall above that in the previous phase(see Fig. 8 and Fig. 9). In addition, the window just to the north of the doorway was also closed in the same mortar type. The purpose of the doorway closure is clear for the hallway on the east side of the dividing wall where a large drainage platform and square plinths were installed for some nearly industrial scale activity that was carried out in this space (excavated in 2016). The platform was constructed to cause whatever liquid was used with it to be directed toward a rectangular drain that exited through the southern wall, constructed with pebblepesto as its capping (visible in the roof of the drain construction).

On the west side of the doorway fill, the closures present more of a problem for sequencing and understanding their purpose. The lower filling of the doorway is stepped back from the surface of the wall, while the upper filling of the doorway is flush with the wall, creating a clear division between the two. The upper fill terminates in an uneven lip of mortar that ought to have been caused by it being poured onto a surface or deposit, leaving this lip rough and slightly bowed, but no such level or surface survived. Indeed, the various elevations of potential floors, such as the beams added into the south wall, also do not correspond with the elevation of this mortar lip, confounding explanation. It is clear that some other activity must have taken place in the west room to explain this lip, but there is no surviving evidence to explain it. The west room, no longer accessible through the doorway in the dividing wall, must have become a part of the shops to the west, access for which must have been via the second story of the shops.

Perhaps also during this phase, two small walls of a fairly high-quality grey mortar binding together cruma and black lava stones were constructed to create a small cubicle of space in the northeast corner of the room (Fig. 10). One running north/south and abutting the northern wall of the room, had its foundations cut into the hard-packed earth of the northern part of the west room. This cubicle was then floored with a thick, fairly chunky opus signinum that was poured up against plasters on both walls. The opus signinum, in fact, was poured through the very narrow doorway created by these two small walls, but whether it continued as a surface beyond this point is unknown since its southern edge terminates immediately upon exiting the doorway.

Phase 6 - post-earthquake(?)
As mentioned above, Kockel proposes that the earthquake of AD 62/3 may have caused a collapse of the vault over Plot 4 room 12 of the shop, which was then replaced by the installation of beams to carry the second floor’s level in the absence of this vault. Also mentioned above, this vault seems more likely to be secondary to the installation of the beams, particularly thanks to one beam hole that does seem to be filled with later masonry. Nevertheless a reorganization of the spaces of Plot 4 room 12, 19, and 20 certainly seems to have occurred in this phase, whether the result of the collapse of the vault in the room 12, or because of the addition of a vault at this point. The downgrading of the rooms for a new purpose seems to have been the case across all of the spaces. It seems that this phase saw room 20, already a hallway, become a cryptaporticus, or perhaps this space was already serving as such in the previous phase. Either way, room 20 was now used to for the deposition of a large volume of material to bring the level of the space up over the drainage platform, creating a sloped ramp from the doorway through the Villa’s west wall descending as the hall continued to the north. The fill was strongly reminiscent of kitchen waste, with a large variety of pottery forms, often cooking courseware, and a remarkable volume of animal bone that had very clearly been butchered. On this material was placed a dolium for the storage of water, as excavated in 2016 (see the 2016 report ).

A desire to change access to and provision of water in these spaces seems to be the driving force behind these transformations. On the west side of the wall in room 19, a cistern with a cylindrical masonry shaft topping a rectilinear base was installed that gave water access to the second story of this space (Fig. 11). Contemporary with the construction of the cistern head was a stone and mortar build, the purpose of which is unclear, but may have served as a manner of cistern overflow. The two small walls that had defined the small cubicle of space in the north-east corner of the room are likely to have been knocked down in this phase, casualties of the re-purposing of the space in general, and allowing for easier access to the cistern. One of these two walls was used to define the northern side of a drainage sluice, built in a poor-quality brown mortar and using a wide variety of materials. This sluice gave access to a hole punched through the dividing wall, as found in 2016, allowing for the drainage of some liquid into a room that exists below the floor of the cryptoporticus on the east side of the wall. The construction of the sluice was seemingly carried out in the process of dumping a huge volume of the pottery waste into the southern portion of room 20. Indeed, the construction often was indiscernible from the material of the dump around it, suggesting that as the space was being filled up, those dumping the material realized the need for access to the hole through the wall for drainage, and so built a structure that brought that access point to a useful elevation as they worked.

The deposit of pottery waste was notable for the volume of largely amphorae, and in many cases, complete amphorae (Fig. 13) that it contained. Several forms enable the dating of the dump to a post-Neroian period. Perhaps the vault of the back room of the shop did collapse at this time, causing further collapses in room 19 and necessitating the decommissioning of this space and its use as a pottery dump.

Phase 7 – no evidence
No direct evidence survives for any notable changes in the area of AA012 during the Villa’s phase 7

Phase 8 – The Eruption of AD 79
No direct evidence for the eruption remains in the western room of AA012.

Phase 9 – Modern Alterations and Interventions
Room 19 was certainly excavated in the modern period to expose both the opus signinum floor in the north-east corner, the sluice, and the top of the pottery deposit in the southern part of the room. Overlying these layers was a heavily disturbed deposit of mixed lapilli, mortar, and soil, over which was a thick layer of black sand. This black sand, which produced a Bourbon clay pipe bowl, was thicker to the northern end, but spread out across the whole area of the room. These deposits clearly relate to the early excavation of the space, but might also be residues of the early sub-surface investigations in the viridarium, which would have produced large amounts of black sand.

Archaeological Area AA013
The 2017 season saw the continued excavation of AA013 in Plot 3, Room 34 of the Villa delle Colonne a mosaico. The objective remained to explore the time-line for the construction of Tombs 6, 7, 8, and 9, as well as their relationship to the phasing of the Villa complex as a whole. The space to the east of Tomb 8 (wall MC.03.189) generated such an abundance of stratigraphy in 2016 that it was necessary to revisit the area to ensure that a comprehensive evaluation of all the deposits identified was completed. Moreover, the 2016 trench boundary was expanded to the north in the western section of the trench to include the back wall of Tomb 9 to allow for an examination of the sequential relationship between the two tombs. (Fig. 14). AA013 South was bounded on the south side by wall MC.03.185 and on the west by the tuff opus quadratum wall which serves as the back wall of Tomb 6 (Fig. 15). Evidence of previous excavations in Room 34 by Maiuri are known both from publications (NdS 1943) and from the 2016 exploration of the area by the VCP. The excavations of 2017 revealed further sondages, and extensive intervention was evident in AA013 North. Even so, the compromised areas were quickly identified and avoided, such that the ample surviving ancient stratigraphy was not jeopardised. While natural soils were not recovered, AA013 was excavated to considerable depth in both the North and South trenches, and six phases were identified in the stratigraphic evidence, spanning Phases 1 through 7 for those defined for the Villa complex as a whole.

Phases 1 through 3 – The Earliest Activities in the Area
The architectural analysis and phasing of the Villa della Colonne a mosaico conducted during the 2015 season concluded that the addition of the viridarium and the communication between the Villa core and Room 34 occurred in the penultimate stage of the Villa’s expansion. Consequently, the chronological relationship between the earliest deposits detected in AA013 and the early phases of construction in the Villa core is unclear. While no direct temporal relationship can as yet be established between the two areas, a preliminary analysis of the earliest deposits identified in AA013 suggests that a prolonged period of activity in the area preceded the construction of the Villa core. Maiuri's excavations uncovered several Samnite graves in the area that would later become the viridarium and Room 34, suggesting a funerary use of the space. The earliest stratigraphic phase identified in both AA013 South and AA013 North consists of a deposit of dark brown soil with very few inorganic contents. The colour, texture, and virtual sterility of the deposits suggests the re-deposition of natural soils and likely reflects early levelling activity. These probably took place in the area prior to the construction of the original Villa core. Overlying this stratum in AA013 South, a deposit of fine, silty, medium grey soil produced scant finds relative to overlying deposits. A Greek oil lamp was recovered from the succeeding deposit, as was charcoal, black gloss pottery, and a relatively large amount of reed impressed finished plaster. The ensuing deposit produced a high concentration of black gloss pottery, which included a nearly complete black gloss lamp, waterproof plaster, and a sea urchin shell midden near the north-east trench boundary. While this evidence is by no means conclusive without further study, an archaic date for these deposits seems likely (Fig. 16).Interestingly early deposits corresponding those in AA013 South were generally absent from AA013 North. It may be that evidence of these layers has been obscured by later construction in the area in ancient times. A subsequent series of deposits in AA013 South have been interpreted as redeposited fill layers intended to raise the floor level in the area. As with the earlier deposits discussed above, their precise correlation to the early phases of development in the Villa core is uncertain. Several layers of differential fill were found overlying the earliest strata. One of which produced a higher concentration of construction materials including mortar, plaster, and chips of Nocera tuff.

Phase 4- Expansion of the Villa. Construction on the Via dei Sepolcri
Phase 4 in the development of the Villa complex coincides with the deposition of an extremely thick, hard-packed levelling deposit associated with the construction of wall MC.03.186 (Fig. 17). The deposit extends the length of the trench from north to south, though it becomes degraded in the southern area just north of wall MC.03.185 and just to the south of the northern trench boundary. The east side of the deposit is truncated by a linear cut that runs from the northern to the southern trench extent. It laps onto the black lava opus incertum foundations of the opus quadratum tuff wall on the west side of the trench, indicating that it was deposited after the construction of Tomb 6. Characterised by exceptionally compact greyish brown soil, it seems likely that formation of the deposit was concomitant with the construction of the tomb, and tuff chips recovered during excavation are likely residue of its construction.

Phase 5-Expansion of Villa. Further Tomb Construction
Sometime following the construction of Tomb 6, the hard-packed surface was cut for the foundation of the back wall of Tomb 7. Architectural analysis of the surrounding opus incertum walls MC.03.187 and MC.03.188 indicates a single construction event. Further examination revealed the use of a distinctive yellowish brown mortar with white limestone inclusions which has been identified elsewhere in the Villa complex in Phases 5 and 6. A number of other cuts were detected in the hard-packed surface which must have occurred sometime between Phase 5 or possibly early in Phase 6. Unfortunately their purpose remains generally opaque. A series of levelling deposits related to construction in the northern area of Room 34 occurred during this period as well. A shell midden was seen in section in the lower layers of the deposits, and limestone and tuff chips seemingly relate to the shaping of Tomb 8. From the midden, three types of marine shells were identified, some of which were unopened, indicating that they may have been used for consumption. The yellowish brown mortar and the heterogeneous assortment of construction materials used in the creation of the drain that runs under the tomb indicate that it was also a Phase 5 addition. The presence of limestone and tuff chips in the drain foundation and in the surrounding deposits suggests that these additions were contemporary with the construction of Tomb 8. The northern expansion of AA013 in 2017 allowed for the excavation of the deposits north of a lapilli filled cut running east-west from the eastern trench boundary to the mouth of the drain. A large, irregularly shaped Maiuri trench was also identified in the area, with the result that surviving ancient stratigraphy was pinpointed in a long, narrow spine between the two. Strata were found to be generally consistent with corresponding deposits on the south side of the trench.The north-west corner of AA013 in the area immediately east of Tomb 9 (SU 013.042) was found to be undisturbed by Maiuri's sondage. It was determined that the stratigraphic sequence in this area also coincided with that found in the other areas of the trench.

Phase 6- Construction of Viridarium and Final Enclosure of Room 34
Phase 6 was the second dramatic period of change in Room 34. Two final cuts detected in the underlying hard-packed layers date to this phase, and precede another major levelling deposit which was deposited throughout the area. The first of these was the foundation cut for the opus incertum wall MC.03.185, which could be seen in section in the south-west corner of the trench. While this wall employed a similar easily distinguishable yellowish brown mortar to that used in the construction the earlier walls, the use of more uniform construction materials is indicative of Phase 6. This is also the period during which the viridarium became incorporated into Villa complex, and it seems appropriate that the wall which contributes to defining the interior space of Room 34, should have been built at this time. The second was an oval cut in the south-west corner of the trench filled with a nearly complete vessel, missing only the handles, which must have been deliberately placed on its side in the cut. The vessel was originally thought to be a cremation urn, however meticulous excavation revealed that it contained only sterile soil. In the absence of the possibility of the vessel serving as an urn, it seems likely that it was deposited as a libation offering, perhaps in connection with Tomb 6. The overlying deposit appears to have been laid immediately following deposition of the vessel. This was an immensely thick, solid, greyish brown deposit that extended over the majority of the area of AA013 South sloping from west to east. A modern Maiuri trench truncated the deposit in the eastern area of the trench. It produced an extremely rich and diverse assemblage of bone, pottery, metal, coins, and construction materials. A large quantity of tuff chips recovered from the deposit have been interpreted as discarded materials associated with the dressing of the tuff blocks tomb 6. Although the construction of the tomb itself dates to Phase 4, it is clear that the shaping of the of the face of back wall of the tomb occurred long after its initial construction.

The construction of the viridarium and subsequent communication between the Villa core and the space behind the tombs, as well as the construction of Tomb 9, resulted in the final enclosure of Room 34. Excavation in AA013 North revealed that the foundations of Tomb 9 rest upon a bed of the limestone chips associated with the dressing of Tomb 8. A circular cut located immediately southeast of Tomb 9 may be a posthole related to the tomb’s construction. The construction of Tomb 9, which abuts Tomb 8 to the south and wall MC.03.191 to the north, effectively separated the area behind the tombs from the Via dei Sepolcri. An evaluation of the construction techniques employed in the creation of Tomb 9, characterised by the uniformity of construction materials and the use of yellowish brown mortar, definitively assign its creation to Phase 6. Analysis of the mortars and construction materials used in the erection of walls MC.03.181, MC.03.182, MC.03.183, MC.03.192, MC.03.193, and MC.03.194 decisively indicate that they also belong to this phase.

Phase 7- The Final Phase
Final phase changes in the area of Room 34 include the removal of a southern portion of SU 013.020 which appears to have originally continued to the west, where it presumably abutted the north end of Tomb 6. Evidence for the continuation of this wall comes from a bit of plaster which survives at the base on the southern face of the wall and laps out over an empty space. Two cuts have been identified which relate to the removal of this wall. The first runs north-south and aligns with the point where the west side of the surviving wall ends, seemingly in an attempt to remove enough of the overlying deposit to reach the wall’s foundations. At the south end, the cut becomes flush with the surrounding level of the underlying hard-packed soils. The north end of the cut terminates in a second cut filled with clean eruption lapilli, dating the removal of this portion of wall to the final phase.

Phase 8- Modern Phase Interventions
Both publications (NdS 1943) and excavations undertaken in 2016 by the VCP attest modern excavations under the direction of Maiuri took place in areas of Room 34. Further evidence of these investigations were identified in the 2017 season. A rectangular trench running north-south in the far east area of AA013 South was quickly detected upon removal of the modern overburden. AA013 North also produced evidence of explorations by Maiuri in the modern period. An irregularly shaped cut in the central northern area of the trench was discerned. Partial removal of the fill of the cut was undertaken in order to allow better access to the areas in the trench at depth. Further analysis of the cut over the drain that runs under Tomb 8, interpreted in 2016 as an ancient trench cut that was open at the time of the 79 AD eruption to facilitate construction on the drain, indicates that it may in fact be the result of excavations in the area which preceded Maiuri. A 1910 plan published in (NdS 1910) displaying a drain which runs east-west from below Tomb 8, may suggest that prior investigation of the area was undertaken. Striations of soil running through the lapilli and a deposit of mixed rubble near the bottom of the deposit were originally identified as the result of section collapses related to seismic activities caused by the eruption. However, it is also possible that they reflect redeposited materials sourced from the original clearing of the area behind the tombs. If this is in fact the case, it may explain why no trace of the base of the drain could be detected, even though the area was excavated to depth.

Archaeological Area AA014
Although the phases of the walls in this area have been confidently sequenced to Phase 5 within the overall sequence of the Villa itself, other aspects of the area remained unclear. Previous excavations published in NdS 1910 documented the presence of an apparent impluvium between the columns and an extensive drainage system during the final phase of the area, but had not been recorded in sufficient detail.In an effort to resolve these questions, an Archaeological Area (AA014) was defined and excavated within the viridarium, encompassing the eastern reconstructed columns and extending to the low north-south wall that divides the at the eastern nymphaeum. Within this area, three sondages were established in areas of the AA to focus the excavation and establish a more complete stratigraphic sequence. Overall, five distinct local phases were identified.

Phase 1 – Truncated Natural and Possible Traces of Pre-Roman phases
Previous excavations in this area produced evidence for inhumation burials in the viridarium, indicating its long-term use for funerary purposes long before its Roman period transformation. While the extensive truncation of deposited caused by earlier excavations in the area meant that few traces of these activities were recovered in AA014 it is nevertheless possible to say something about these early layers. It seems possible that the natural deposits in the area witnessed considerable terracing, removing soil down to the Mercato eruption layer, which was found at depth of over 1.5m below the modern surface in AA014. It was clearly characterised by a light brownish yellow colouration with light grey or yellow small inclusions. This layer was irregularly cut on a north-west to south-east alignment and then was filled with a medium greyish brown sandy silt rubble deposit visible in both the south-west corner and in the western extent of the area. Notably present within this deposit, especially in the higher elevation, was a substantial amount of reed-impressed mortar. Sarno stones and finished plaster were also recovered, however since safety concerns limited excavation, it is not known to what depth this rubble fill continued. The colour and texture of this deposit indicates that it is a mixture of the natural soils that are known to have been present elsewhere in the city. In this, it is likely that a similar natural sequence existed in the immediate area of the villa and that in the Roman phase of the viridarium it was excavated and redeposited with rubble mixed into the matrix, or that it was brought in from somewhere outside of the villa. Precisely what the purpose of these activities was, remains unclear. It is possible that in general they may relate to the on-going use of the area for funerary purposes, with the churning of natural deposits the result of repeated cutting and filling.

Phase 2 (Villa Phase 5) – Ancient Cuts, Fills, and Builds
Overlying the earliest activities in both the south-western corner and in the west of the AA trench extension area was a thin light grey ‘ashy’ layer. Both these stratigraphic units contained substantial charcoal fragments, however SU 014.036 in the western section of the AA trench extension area contained an inverted, mostly intact coarseware bowl with numerous associated carbonised seeds and with traces of blue pigment identified on the base of the bowl (Fig.18). Whether the presence of this bowl is indicative of any intentional or ritual activity is unknown, however the unusual occurrence of seed in within this context may be an important consideration. Activities related to the change from funerary to domestic use might be a plausible explanation of these deposits. Thereafter, another layer characterised by a medium greyish brown silty sand was deposited over this, likely to raise the level of the viridarium to an appropriate elevation for the construction of the impluvium, drains, and column bases. This second fill layer is also characterised as a mixture of the natural sequence, and is similar in colour to the deposit below the ashy layer.

The construction of the impluvium, the southern and eastern drains, and column bases all occurred contemporaneously with the precise sequence of construction clearly visible. The lower course of the opus incertum impluvium as well as the north-west to south-east drain in opus incertum and pan-tiles to the south are mortared together where the drain extends from the southern impluvium face. In the southern trench extension area, it is also clear that the perpendicular intersection of this southern north-west to south-east drain and the eastern north-east to south-west drain were constructed in a single build. Built directly on the surface of the north-west to south-east drain, about 1 m to the south of the south-eastern corner of the impluvium, is one of the two ancient bases, the other mirroring this on the north-eastern corner of the impluvium. These column bases are larger and slightly further removed from the impluvium than the modern reconstructions would suggest. Although little of these ancient bases survive, enough is extant to show that they are comparable in size to the bases of the original mosaic columns now held in the collection of the Museo Archaeologico Nazionale di Napoli.

The identification of these ancient bases this season has been important in the reconstructions of the ancient viridarium as a whole. Attempts to imagine how the ancient space worked have proved difficult with the height, diameter, and location of the modern reconstructed columns. However, the more substantial size of the ancient bases as well as their wider placement, is more in keeping with typical Roman architectural desire for symmetry, especially when viewed as a single vista with the mosaic nymphaeum centred on the eastern wall of the viridarium. The excavation of the ancient impluvium this year also resolved an important question concerning the original flooring level of the ancient viridarium. The walls defining this viridarium area, and the fauces adjoining to the west have to this point provided no indication of the ancient surface elevation, which is problematic – especially with regards to the relationship of the area and the adjoining room to the south. As is now known, extensive excavations in the area served to remove any traces of the original level other than the impluvium and drains, even in the fauces and at the surrounding walls. It is hoped that the identification of this ancient level via these features will prove beneficial to the overall final sequencing of the Villa delle Colonne a mosaico.

The alignment of the drains themselves has proven to be of interest to the reconstruction of the function and form of the Roman viridarium (Fig. 19). The course of an east-west aligned drain of substantial size found in AA013 indicates considerable drainage capabilities of this southern section of the viridarium. However, excavations this year, in association with those published in 1910, indicate that the southern north-west to south-east drain continues to the south-east corner of the viridarium, and has no known association with the drain found in AA013. The excavation of the fauces itself this season (AA015) similarly revealed no identifiable drainage system linking the impluvium and viridarium drainage system to the Via dei Sepolcri to the west. The construction of the upper course of the impluvium leads to an unusual, bent mortar-and-brick outlet that runs under the south-eastern corner of the impluvium itself before linking to the north-west to south-east aligned drain just to its south. The construction of this opus incertum and pan-tile overflow drain has been thickened on top with smaller broken pieces of tile, as has the top of the impluvium. This, in addition to the single mortar type used in the construction of all known features in the viridarium, supports the conclusion that these features were constructed together in a single event. Clearly a very different system of drainage prevailed in the area than would be imagined from the standing architecture.

Intermediate Ancient/Modern Phase – Probable Earliest Modern Investigations
Notably, as the 1910 excavation publication mentioned that the impluvium was found already despoiled of it marble opus sectile decoration, it is clear that some phase of destruction had already damaged elements of this feature. It is as yet unclear whether this destruction event occurred during the final phase of the viridarium prior to the eruption of AD79, or whether it was a result of the primary excavations of the space from the eruption debris carried out in the 1800s, which also appear to have resulted in the removal of the decorated mosaic columns themselves.

The excavations of 1910 also clearly left their mark on the features they exposed. Indeed, they appear to have involved considerable removal of soil across most of the area, possibly in the search for early graves. Both the north-west to south-east drain located in the south of the AA, and associated eastern north-east to south-west drain found in the eastern extent, show signs of damage. The pan-tiles forming the top of the eastern north-east to south-west drain have been broken through with the bed of the tile broken back to the flange on both sides. The broken pieces of these tiles were not found associated with the drain, indicating that they had been removed with whatever contents were exposed within the drain itself. The north-west to south-east drain located just to the south of the impluvium displays considerably more damage, although of a different nature. Excavation of this part of the southern drain extent revealed that it was considerably undercut – to at least half the width of the drain construction (Fig. 18). When the exposed sections of these excavations were backfilled, it is clear that the first backfill deposit did not completely fill the underside of the drain’s total width. After some time, the drain settled into this void, causing the southern half of the drain’s width to collapse, as may be seen by the crack along the length of the exposed base of the drain, and the downward angle to which this broken half is now resting. Similar undercutting is also clearly evident below all ancient features found within the bounds of the AA. Due to concerns of safety and in order to preserve the drains themselves, no attempts were made this season to investigate the full extent of the undercutting that occurred during these early historical excavations.

The re-deposition of the fills and levelling layers from the 1910 excavation is evident across the whole AA. In the south-western extent, to the south-west of the impluvium and southern drain, a fill that sloped to the west from just below the southern drain’s surface was characterised as a medium brownish grey sandy silt with notably few ancient finds and inclusions. Although the precise date is not determinable, a fill was then deposited over this sloping layer, which raised and levelled the deposits to match the southern drain’s top surface elevation. This second fill was characterised by a medium yellowish-greyish brown deposit with inclusions of plaster, some pottery, brick and tile, and some bone. In the easternmost extent of the trench, a layer of rubble fill with an associated overlying medium greyish brown silty sand fill to the south was identified and also attributed to this time on account of lapilli found within its matrix. Finally, a layer of a slightly darker greyish brown deposit accumulated above these fill layers, which can be clearly identified in the eastern extent of the AA, to the north-east and east of the impluvium, and in the south-western corner of the AA.

Modern Phase 2 – Maiuri's 1930's Excavations and Subsequent Beautification of the viridarium
The history of modern interventions into the Roman viridarium continued with the excavations and beautification efforts conducted by Amadeo Maiuri in the 1930s. This event is documented in a photo found in the Visioni Italiche: Raccolta diretta da Cesare Rossi e Marco Varoli: Pompei di Amadeo Maiuri. Excavations in the northern part of the viridarium conducted by the Project in 2009 (AA004) revealed how extensive these efforts had been. AA004 uncovered the consolidated structure of at least three inhumation burials that had been uncovered or rediscovered by Maiuri’s explorations that were subsequently consolidated in modern cement, seemingly for the purpose of public display. It is therefore not surprising that Maiuri’s investigations also extended into the area of this year’s AA014.

The extent of Maiuri’s excavations and beautification efforts is clearly identified in various locations across the AA. In the south and south-western area of AA014, sub-linear and sub-circular cuts were made through the post-1910 accumulation deposit, exposing the north-west to south-east drain in the southern extent of the trench, the top surface of the impluvium and the associated inlet drain on its eastern edge, and the ancient column base, which is situated directly over the southern drain (Fig. 20). Located in the northern extent of the AA, a sub-circular cut was also made through the hard earthen deposit to expose the north-eastern ancient column base. This cut was also likely used to expose to depth the area adjacent to the northern face of the impluvium, a result of Maiuri’s further for inhumation burials, which were known to have existed in the immediate vicinity. The bottom of this cut identified in this northern area of the trench was not recovered, even at a depth of about 1.8 m below the modern surface.

With the re-excavation of the known ancient features complete, attention was then turned to the consolidation and beautification of the viridarium. Both the north-east and south-east ancient column bases were capped and consolidated in an opus incertum-style masonry with sarno stones and limestone, mortared with a characteristically modern light grey mortar filled with lapilli. To install water pipes for the intended fountain feature located within the impluvium, a shallow east to west cut was made along its base, through its associated drain inlet on the eastern edge of the impluvium, and cut through the post-1910 fill and accumulation layer to, and under, the eastern nymphaeum dividing wall, which was constructed only after these cuts had been filled. The distinctive lapilli mortar was also used to repair some slight damage to the eastern and southern surface extents of the base of the impluvium, possilby itself the result of the removal of the ancient opus sectile facing, known to have existed on its ancient-phase surface in 1910. It is also clear that the lapilli mortar used to consolidate and level the base of the impluvium was formed around the base of the fountain piping, holding it in place while in use. Interestingly, a deposit of light grey ashy silt and lapilli was identified at the base of the cut, overlying the exposed surface of the southern drain. It is likely that this location acted as the mixing area for the lapilli mortar production, with the residue of this process still extant and visible in section in areas across the southern AA extent.

Following the consolidation of the exposed ancient features, these areas were then filled with a medium greyish-brown sandy silt. It was upon this fill layer that the present-day concrete columns were reconstructed, although at a distance much closer to the corners of the impluvium than the ancient originals. A low wall and metal fence was also constructed over this fill layer, separating the nymphaeum feature from the rest of the space in an attempt to protect its highly decorated and fragile mosaic surface from the intended visitors to the viridarium.

Modern Phase 3 – Post 1950s Removal of Beautification Aspects
It is not known how long this viridarium remained in its modern reconstructed condition, but it is clear that it did not continue in such a decorated state to the present day. The photo identified in Visioni Italiche: Raccolta diretta da Cesare Rossi e Marco Varoli: Pompei di Amadeo Maiuri shows the finished surfaces of the impluvium and reconstructed columns, as well as planted trees on the outer edges of the impluvium space, none of which are present today (though many new trees in the area might be their descendants). Although no traces were identified this season, a cut must have been made through the topmost deposit from the centre of the dividing nymphaeum wall to the centre of the impluvium, to remove the piping used to pump water to this modern impluvium fountain. Following the removal of this piping, the eastern impluvium edge that was functioned as the drainage channel was filled with rubble and the entire area covered again by the modern topsoil and overburden.

Archaeological Area AA015
This trench was positioned in the fauces of the Villa delle Colonne a mosaico that lead into the viridarium space. The trench ran from the threshold stone at the entrance of the villa to approximately halfway up the fauces and was restricted to the southern half of the hallway. The location of this excavation was chosen with multiple aims in mind. The first objective was to investigate whether a pavement existed and to document any remains that were uncovered. The second objective was to identify areas for targeted excavation in order to clarify the chronology of the construction of the viridarium space and its alignment with the phasing of the Villa as established and discussed in the VCP 2015 report.

Phase 1 – Earliest Traces
Currently, the earliest evidence of activity in this area is the construction of the southern fauces wall (and presumably the northern wall), its foundations, and an early wall/threshold. The foundations are built exclusively in black lava in opus incertum and use a strong light grey mortar to bond these stones. No foundation trench was uncovered in the course of this year’s excavations. The foundations are wider than the wall above, which are also constructed in black lava opus incertum. A north-south running wall that is keyed into the foundations was also constructed at this time. It is built in a black lava opus incertum and uses the same strong light grey mortar as the foundations. This wall may have served as an early threshold. The northern and southern fauces walls use a light-yellow mortar with prominent white lime inclusions that weaker in quality than the mortar of the foundations (Fig. 21). Each fauces wall is set upon a large black lava plinth at the Villa entrance. This period corresponds to Phase 6 of the Villa phasing discussed in the VCP 2015 report. A hard-packed grey-brown surface that slopes downward to the north-east is present in the centre of AA015 and may represent an early work surface or street. At the time that excavations ceased in 2017, however, the precise stratigraphic relationship between the surface and the foundations of the southern fauces wall could not be clarified.

Phase 2 – Levelling Layers in the Villa
Following the construction of the foundations and southern fauces walls, the hallway must have been raised to the level of the final floor. No traces of a pavement have survived, possibly due to the large-scale removal of soils that appear to have characterised the excavations of the viridarium (and seemingly also the fauces) in 1910. The layers recovered shared several characteristics that include a light brown to bright yellow soil of very firm compaction and small amounts of pottery and plaster. A separate small deposit of plaster in the centre of AA015 also formed a part of these levelling layers. This produced several large fragments of polychrome plaster, primarily Third Style. It is as yet unclear whether these are ancient or modern levelling layers.

Phase 4 – Modern Alterations and Interventions
Modern excavation of the fauces and viridarium space cleared the eruptive material and reached depths below the AD 79 layer. A drain that spans the length of the fauces was cut into the centre of the hallway and constructed out of large modern pan tiles. This appears to have been done in order to act as a conduit for the modern impluvium and fountain that was built just to the east in the centre of the viridarium. The placement of two large white marble blocks as the threshold also seems to be a modern action. The eastern face of the southernmost threshold stone was extensively reworked and chipped in order to make it fit into the space. A deposit of these stone chips was removed with modern material (Fig. 23). The addition of an elegant white marble threshold follows the pattern of modern reconstruction in the Villa in order to increase its aesthetic appeal to visitors, and may have been necessary after the widespread removal of soils in the area in the excavations published in 1910.

The eastern extent of the trench also preserves evidence of these modern interventions. Deep cuts in an ancient hard-packed surface were made during the course of modern excavation and damaged the foundation of the southern wall by removing the stones that projected out from the wall. These cuts preserve where excavation of the viridarium began and extend to a significant depth, approximately 60 cm. It is believed that the excavations published in 1910 removed an enormous volume of soil within the viridarium in order to reach the depths of earlier Samnite burials. The exposed areas were then refilled with soil that is characteristically devoid of finds (but includes lapilli) in order to level the area for modern reconstruction (see AA014 Final Report above for more discussion of these activities). The evidence of modern intervention in the eastern extent of AA015 appears to mark the starting point of these operations. Final modern interventions are the consolidation of the Third Style wall painting along its base on the southern fauces wall.

Finds Processing, Ecofactual Recovery
As always, excavation in 2017 was accompanied by contemporaneous processing, analysis, and recording of all artefacts recovered from this and previous seasons. This produced a database of recovered materials that will serve to help coordinate further study by specialists

Charcoal, Bone, Shell, and Micro-Faunal Remains
Soils from all seasons have now been floated using a bucket-flotation method. As usual, light fractions have been reserved for study by environmental specialists and some heavy fractions were sorted throughout the field season, recovering material smaller than the 0.5 cm mesh employed in dry sieving. Finds included terrestrial and marine bones ranging from mouse to fish, shells, sea urchin and charcoal, including several seeds.

Pottery Analysis
The Via Consolare Project aims to create a complete typological overview of the pottery dataset recovered from the Villa delle Colonne a mosaico and Insula VII, 6. With this study it is possible to begin to characterise the consumption patterns within the excavated archaeological areas, while also contributing chronological information to the stratigraphy of the areas, and the overall phasing of the Villa delle Colonne a mosaico and Insula VII, 6. In addition to providing chronological analyses of the stratigraphic excavations, the study of pottery is used to gain a better understanding of the city in general. In particular this project focuses on the pottery consumption patterns within the city to gain new insights into the ancient Pompeian world and its use of ceramics.

This year we continued our extensive and detailed study of pottery recovered in previous excavation years. The 2017 field season witnessed the complete pottery analysis of AA001, AA007, AA009, AA010 and AA011. Simultaneous to the excavations, a preliminary study of certain interesting finds of AA012 and AA013 was undertaken to provide some useful dates for sequencing. As a result of the work this year, we the analysis of the pottery from previous excavation years was completed and can now fully concentrate on the new material of the 2017 field season. The 2017 field season recovered approximately 256 kg of pottery, roughly 380 litres in volume. The majority of the pottery, originating from the trenches AA012, AA013 and AA015, have been washed, processed, and accessioned to be fully analysed in the following field season of 2018. This excavation season witnessed the recovery of some interesting finds such as an almost intact urceus, a lamp decorated with Jupiter and an eagle, a partially complete Greek lamp, in addition to the normal classes and types expected from other Pompeian datasets, such as the Granai del Foro.

Pottery Analysis of AA001,AA007, AA009, AA010 and AA011
The analysis of the pottery comprised sorting the entire dataset into ceramic classes as a primary division methodology. Each class was further divided based on vessel shape and type. A detailed analysis was performed for each area through the study of diagnostic sherds. In addition to the creation of a typology based on diagnostic elements, a quantification of the entire assemblage was performed based on EVE’s, weight, and volume. The study of the material, included the secondary documentation of the sherds through drawings and photographs.

AA001: Overview
The pottery assemblage of AA001 (124 kg of pottery) was fully analysed and studied during this year’s field season. The general pattern noticed in AA001 produced a similar view of last year’s analysis of the pottery from both AA005 and AA006 (cf. the report of 2016). The general trend in AA001 demonstrated a high presence of locally produced fabrics, with inclusion of black sand, belonging to the Vesuvian area. The locally produced wares were found in association with some imports such as an African amphora of the type Maña C1 (cf. Bonghi Jovino 1984, 283). However the imported products are far fewer than the local productions. One of the focal points of this projects is to gain a greater understanding into the importance of local productions and imitations within the city limit of Pompeii. The assemblage of AA001 was dated to the first century AD based on the presence of the Maña C1 and coarse ware vessels used in the first century AD.

AA007: Overview
AA007, containing 50 kg of pottery, also demonstrated a high percentage of local production types such as a locally produced Dressel 2-4. AA007 became a very interesting dataset due to the high presence of Campanian Orange Ware in comparison to its equivalent Italian Terra Sigillata. Further research will provide more information into this phenomenon and the importance of local imitations. Due to the presence of a Black Gloss bowl type Morel 1412 and some Italian Terra Sigillata identified respectively as Conspectus 20 and Conspectus 36.1, the earliest phases of this area were dated to the late Republic and early Augustan period. The overall pattern in AA007 demonstrated a similar story to the other areas within our research domain. The excavation of the area of AA007 witnessed some interesting decorative lamp fragments depicting a charioteer, a boxer and a lion (Figure 25).

AA009: Analysis (15 kg of pottery)
The archaeological area of AA009 did not provide much information due to the lack of diagnostic elements. One of the only diagnostic elements present in this context is in the form of an olla of the type Di Giovanni 2311 c. As a result we can conclude that this area of excavation can roughly be dated to the first century AD. The most interesting find of the AA009 dataset is an amphora sherd, with an eastern-Mediterranean fabric, carrying a red titulus pictus containing the letters O.I. (Fig. 26).

AA0010: Analysis (7.5 kg of pottery)
The analysis of AA0010 did not provide sufficient diagnostic sherds to be able to obtain sufficient information to gain spot dates. As such AA0010 provides only a rough chronology based on the few rims of coarse ware dating to the first century AD.

AA0011: Analysis (1.2 kg of pottery)
The area of AA0011 had similar problems in comparison to AA009 and AA0010. The lack of huge amounts of pottery and diagnostic elements, caused an impediment to obtain a detailed overview of AA0011. One of the only diagnostic sherds providing any usable data belongs to a bowl type Conspectus 27, in use during Tiberian- Flavian times (15-70 AD).

Conclusions and Preliminary studies of the pottery from AA0012-AA0013: 2017 season
The study of this year further demonstrated the importance of local wares within the city of Pompeii in comparison to the import wares. Furthermore the local imitations of terra sigillata, known as Campanian Orange Ware, were frequently present in this year’s analysis of certain areas (cf.AA007). All research areas can be dated to the late Republican period, but more particular to the Augustan period. Our pottery analysis has amply demonstrated the development of the city during the Augustan times. This research summer has aided in our understanding of the research areas in question and provided more information about particular development phases. Contemporaneous to the excavation and processing of finds, a preliminary study of SU 38 in AA0012 was performed and certain interesting finds of AA0013 were studied, analysed and documented.

AA12: Voluminous pottery deposit
The AA0012 trench, on top of the shops adjacent to the Villa, uncovered a voluminous deposit consisting mainly of pottery. The deposit not only contained amphorae, Terra Sigillata, lamps and coarse ware but more utilitarian ware such as a fragment of a ceramic tripod and a hypocaust pillar. The preliminary study of this deposit focused mainly on SU12.038 and SU12.041. However based on this preliminary analysis we were able to obtain a chronology and date the deposit. The association of following diagnostic vessels date this deposit to the first century AD i.e. Dressel 20/8, Dressel 2-4, a caccabus type Di Giovanni 2211b and an olla type Di Giovanni 2311a, a Jupiter-Eagle lamp type Deneauve VB and a Terra Sigillata plate, orginating in Pisa, with a stamp of Zolonius (5BC-AD50). (Fig.27) However due to the presence of a particular variation of a Dressel 20 and the stamp of Zolonius, this deposit must have been created in post-Neronian times (cf. Bustamante et al 2011, 522). As a result we concluded that this deposit can be dated to the Neronian times or post-Neronian.

AA0013: A preliminary study
The AA0013 trench, situated in the triangular room inside of the Villa behind the garland tomb (Tomb 6), recovered an almost intact urceus in SU13.012. The vessel has been identified as a a two handled urceus of the type Schoene I (Fig. 28). The vessel is 32.5 cm in height and has a diameter of 13 cm in total. This type of table ware is a common table ware vessel in the first century AD. Approximately 16 vessels of this type were discovered in the house of Julius Polybius in Pompeii. The vessel was discovered in situ in a horizontal position. The placement of the vessel and its origin lead to the conclusion that the vessel was possibly placed here after being used a libation vessel. In 2016 a cremation urn was discovered in the vicinity of this urceus (cf. report of 2016). After a comparison between the levels of interment of both vessels, it can be concluded that both levels of placement are quite similar. For this reason it seems very plausible that this vessel could have been used in a ritual environment in connection to funerary rituals. However the urceus is missing both handles and has suffered a partial rim fracture (Fig. 29) A close analysis of the fractures have demonstrated that these fractures are not recent. As such it might be suggested that the handles were intentionally broken off, to enhance the placement of this vessel, since the vessel was placed on its handle side. We plan to create a 3D scan of this vessel during next year’s research campaign. Secondary to the almost intact urceus, AA0013 recovered a partially fragmented Greek lamp. This lamp was used as a mean to date the stratigraphic context and provide more important knowledge concerning a chronology for the part of the villa and its development over time. The lamp in question is regarded as a Greek lamp, datable to the fifth-fourth century BC. (Deneauve 1969, 48).

3D Topographic Survey of the Area of the Villa delle Colonne a mosaico
Having mostly completed the Total Station survey of the Villa delle Colonne a mosaico in 2015, the focus of survey this season was the features and deposits of the trenches. As in previous years, all 3D survey was accomplished with the use of a Leica TCR805power Total Station, in combination with a Leica GMP111-0 Mini Prism. While in many cases, traditional planning techniques were employed in order to record features in precise detail, Total Station survey supplements these plans and provides the overall framework into which they may be fit. In addition, particularly for simple planning of deposits, Total Station survey can provide an expedient alternative that can speed up the process of excavation. This was the most frequent use of survey this season. Finally, Total Station survey is needed in order to provide necessary scale and position for Structure from Motion point clouds, which form an important aspect of the Project's recording process.

3D Data Collection
Excavation in AA012, AA013, AA014, and AA015 was accompanied by complete recording in 3D using Structure from Motion technology. Each stratigraphic unit (SU), feature, and surface was recorded in this manner extracting millions of 3D points and colour information from an unordered series of photographs. Additional photos were taken after each SU/US was photographed, so that every stratigraphic unit was recorded. After processing, these point clouds will be meshed with Meshlab - software designed specifically for cultural heritage projects by the University of Pisa, and developed with the support of 3D-CoForm Project. Following this stage, each mesh will be coordinated into a 3D model of each excavated area that will permit the re-examination of the excavation in the future, in almost as much detail as the original one. This permits a complete reconstruction of the entire excavation and even the virtual ‘re-excavation’ of deposits recorded in 3D.

Conclusions and Current Interpretations
Archaeological investigations in the Villa delle Colonne a mosaico and the completed excavation of AA012, AA013, AA014, AA015 have helped to resolve a number of questions about the Villa and its development, particularly during two major phases of expansion. The first involved the creation of the long row of shops that face onto the Via dei Sepolcri and served to support the upward expansion of the villa and its addition of second and finally third stories. The second added lateral expansion to this upward growth, by expanding to the south, behind at least one pre-existing tomb and over the top of a long standing cemetery from the Samnite period. The exact nature and timing of this change, which must have been quite a dramatic transformation, gives the whole area outside of the Porta Ercolano a new, and perhaps darker complexion. In addition, the work of this summer also revealed further detail about the final phases in the western shops and their relationship to the villa itself. Plausibly caused by the earthquake(s), of 62/3, these changes, which involved the transfer of rooms that had previously belonged to the villa use of the shops represents a dramatic shift of control, and possibly a significant transformation in status. Finally, excavations in the viridarium and fauces both revealed much about the nature of two phases of previous investigations of the in these areas, which were clearly much more invasive than had previously been imagined. Despite this, it was nevertheless possible to piece together evidence that survived these remarkably aggressive excavations in order to shed light on the ancient phases that formed the motivation for these trenches.

An Expanding Villa - Upper stories and even more extensive vaults
Excavations in AA012 this year revealed more information about the western room in this area, demonstrating that there was also a lower vaulted room, in this area, similar to that under the eastern corridor discovered in 2016. It is therefore clear that the process of creating the vaults between the Villa and the shops involved lower stories not only under the eastern corridor, but also at least one small room to the west. This serves to resolve completely questions about the early elevations in this area, which can now be seen to be not so very different from the level of the lowest phases of the Villa core and the current elevation of the via dei Sepolcri. In fact, it is clear that the expansion of the Villa in Phase 5, which saw the creation of extensive vaulting in the western shops intended to support upper stories of considerable height, was part of a more extensive system of vaulted spaces that ran around the Villa on its western and northern sides. It seems likely that the lower storey of this involved a corridor on the east and rooms on the west, just as was the case with the upper storey, meaning that a series of cryptoporticus-like areas flanked the Villa, serving to connect it physically to the shops and producing a complex that served as a base for the upper stories of the Villa itself. While some of these observations were already made last year, the continued excavation of AA012 completed this picture, revealing that the whole zone between the Villa and the shops was filled with vaults such that the Villa sat atop a man-made hill. The scale of this undertaking was enormous, testifying to the tremendous wealth of the Villa owner.

An Expanding Villa - Building over Samnite graves
A subsequent phase of expansion in the Villa involved an equally dramatic transformation of the landscape. The Samnite tombs recovered from the area of the Villa testify that much of the space it occupied had originally been of funerary use. The final phases of the city witnessed the expansion of the Villa to include these areas, in form of the viridarium, Sacellum, and the space behind tombs 6, 7, 8 and 9. Clearly a change of this type could easily be seen as an act of cultural imperialism and oppression depending on how active or ongoing the relationship was between those buried in these tombs and their descendants. The timing of this change is therefore also vital to understanding the exact meaning of such a transformation. Excavations in AA014, despite considerable modern interventions, were able to identify aspects of these changes that reveal part of the story of how they took place. At the level of the Samnite tombs, which appear to have been cut into the natural soils at a fairly deep level in which the Mercato eruption's characteristic yellow soil was evident, were sealed by a thick layer of re-deposited natural soils that served to seal the tombs and create a new working surface. Before this layer was deposited, a seemingly widespread layer of charcoal and ashy soil was laid down, possibly he result of activities needed to decommission the cemetery and make its use for domestic purposes possible. Certainly, almost nothing is known of how such a transformation could have taken place. One must think only of difficulties that Cicero encountered regarding his house-turned shrine to imagine that some legal and religious action must have been necessary. However, given the expansion of the gardens of early imperial villas in Rome over the plebeian cemeteries of the Esquiline, it is also clear that such transformations were possible. In fact, the timing of this very similar change in Rome, which appears to precede the similar expansion of the Villa delle Colonne a mosaico by roughly a generation, may suggest the longevity of local Samnite traditions that prohibited such changes for some time. Certainly, no trace of activities or construction between the earlier funerary use of space and the later garden was recovered. At the same time, it is difficult to imagine that everyone in the cemetery had been forgotten only a century after the implantation of the Roman colony. Surely somebody must have been upset by these changes, and perhaps the layer of charcoal and ash documents efforts made to propitiate the dead. The thick layer of deposited natural served to separate the later phases from these earlier uses. At the same time they also served to raise the level of the soil in this area and to bring this part of the Villa more into line with the elevation of the core to the north. After the mosaic covered columns, drains, marble revetted impluvium, and nymphaeum had been constructed directly upon this layer, the transformation of the area was complete.

Similar changes were undertaken in the triangular space to the south of the viridarium, but here, the transformation of function was less pronounced. Indeed, though this area also witnessed large levelling layers, and may even have seen the same redeposited naturals over charcoal and ash, the activities that followed remained in the funerary sphere. Indeed, situated behind one pre-existing tomb, of the late 1st c. BC and over the Samnite tomb recovered by Maiuri, levels of soils were raised beyond the elevation of the viridarium, possibly making use of soils deriving from the excavation of the foundations for new tombs and enclosures that by the end of the city, had filled in the whole western side of the space. The widespread presence of Republican and even much earlier wares suggests an early, if perhaps redeposited origin for these layers. At greater depth, one particular fill of note was a midden of sea shells that must result instead from a particular activity, possibly related to the funerary activities of the area as a whole. Even within the upper strata behind the tombs there is clear evidence of the continuation of ritual, such as the deposition of a nearly complete wine urceus, laid in its side and buried within the deposits. The excavations in 2016 also produced a cremation urn from these upper fills. Since this area was reachable only from within the Villa, it seems likely that these funerary activities pertained particularly to that family.

Post Earthquake Property Changes in the shops
Continuing evidence for post-earthquake alterations in the Villa and shops was also found in AA012. The lower room of the vaulted spaces in this area was found to be filled with a thick layer of pottery, including complete, but broken, amphorae, which appear to have been stacked on their sides, perhaps in an effort to fill up the space more quickly. This decommissioning of the lower storey, which mirrors the use of the eastern first floor as the destination of a toilet or sluice as described in the report from 2016, would suggest not only considerable downgrading in the use of these spaces, but also a certain desperation of construction efforts. While there is no explicit indication of earthquake induced trauma, the timing of the changes, in the period of Nero or later, certainly makes this a possibly. Related to these changes was the addition of an unusual first storey cistern to this area, presumably intended to provide at least some of the liquid that was disposed of in the nearby sluice.

Extensive modern interventions
Though it was not the intention of this research to examine the nature of earlier interventions and sub-surface explorations of the Villa, this season produced considerable evidence for several widespread campaigns that have only been partially documented. Explorations that led to the publication of 1910, which were motivated almost entirely by the desire to recover Samnite graves, appear to have removed massive amounts of soil from the viridarium and fauces, often to the level of more than a metre. Features such as the drains and impluvium mentioned in this report must have been pedestalled or supported with chocks. Certainly considerable damage was sustained to the drains at this time, which appear to have been undercut, perhaps in an effort to recover underlying tombs. The extensive nature of these excavations is not mentioned at all in the report in the NdS. Following the backfilling of this area, further explorations were undertaken by Maiuri in the 1930-50s. These appear to have been intended to prepare the Villa for display. To the north, Samnite graves were rebuilt in concrete, as found in AA004 in 2009, while the impluvium and nymphaeum was given piping for a fountain, and a new drain was cut through the fauces. Evidence of this drain, including a new modern limestone threshold was found in AA015 this year. Concrete replicas of the mosaic columns were placed, albeit in different locations from the originals. One the whole, the explorations in the Villa have been extensive, producing a difficult and complicated archaeological record of surviving ancient elements.

The results of the 2017 field season have served to augment considerably the understanding of Villa and its dependencies. The expansion of the Villa fro its central core is now well understood, and the history off the area behind the tombs has also become somewhat clearer. Focus must now return to the Villa core, which represents not only the last area of the complex that has not received extensive modern explorations, but also holds the answers for questions of the earliest chronology of the Villa and its original construction. The Project will seek permission to excavate in these areas in 2018, as well as continuing the investigations and documentation of the urban development and chronology of Insula VII 6, and the Via Consolare as a whole. As always, we remain deeply indebted to the Soprintendenza Archeologica di Pompei, the Ministero per i Beni e le Attività Culturali, Soprintendente Prof. Osanna, Direttore dott.ssa Stefani, Arch. dott. Galeandro and Assistente Sabini and extend our warmest thanks for their kind and continued support and encouragement in our research activities. Our work could not have been done without their aid. Finally, we wish to thank our great friends at Bar Sgambati and Camping Zeus for their ongoing generosity and unending friendship toward the Via Consolare Project and its members since its inception.

1. Kockel, V. and B. F. Weber. 1983. ‘Die Villa delle Colonne a mosaico in Pompeji.’ Mitteilungen des Deutschen archaeologischen Instituts. Römische Abteilung: 60-61.
2. We are grateful for this interpretation by Caroline Cheung.
3. Kockel, V. and B. F. Weber. 1983.
4. VisualSFM by dott. ChangChang Wu and PMVS2 by dott. Yasutaka Furukawa and dott. Jean Ponce are normally employed by this Project. This year the commercial alternative PhotoScan was tested as an alternative. The results were found to be impressive but largely comparable.
5. Kockel, V. and B. F. Weber. 1983.