PIA GULDAGER BILDE

< Phase 1, Late Republican, (mid?) 1st century BC (Caesar?)

< Phase 2, Early Imperial, ca. 20-40 AD (Caligula?)

< Phase 3, Late Neronian-Early Flavian, 60-80 AD

< Phase 4, Hadrianic, 120s AD

Overall layout of the villa

The villa is as already mentioned situated by the southwestern shore of the lake. The backbone of the villa is an artificial, oblong terrace measuring ca. 260 x 60 m. The villa platform consists partly of levelled rock, and is partly built in Roman mortar. Towards the lake, the wall is more than 5 m high. It has not yet been excavated to its full depth. In its final phase the area covered by the villa measured ca. 100 x 450 m.

The main terrace of the villa is oriented roughly north-northeast to south-southwest with the main view overlooking the lake. The orientation of the villa towards east provided an intended coolness at summertime. The villa’s long side is parallel to the lake, and the changed orientation of the villa’s two wings implies that its overall layout follows the curvature of the lake, a feature probably established already with the first villa phase (Fig. 2).

Access to the villa was seemingly two-fold: via a paved road and by boat. The paved road of access entered the villa plateau from the north. This road intersected the road diverging from via Appia and leading to the Sanctuary of Diana, probably the via Virbia.^[ix]

The level of the lake was regulated with care with the emissary,^[x] the 1654 m long passage cut through the mountain. This overflow kept the level of the lake constant at ca. 322.25 m above sea level, almost 3 m higher than the present level. The lower floors of the villa were situated very close to this ancient level of the lake signifying that it must have been a main point to be able to moor at the villa’s quai. The villa was at least in its later phases accessed also by boat from the lake, probably for representational purposes. At least from the second phase, the villa plateau could be accessed via a tunnel cut through the terrace wall and leading visitors directly to the heart of the villa: the central building block. We know very little of this central block as it has been almost completely destroyed by recent agricultural activity.

The area north of the villa plateau consists of two parts: towards the north an area apparently without buildings. This open space is framed by a terrace supporting the access road towards west, and towards east close to the lake a long wall. The space between these two walls was probably a walled(?) garden (Fig. 2). Immediately south of the proposed garden is a building block with a structure in two stories being at least in the final villa phase 4 a bath (see the contribution by E.M. Viitanen).

On the villa plateau were the living quarters and the villa’s representational rooms. Furthest to the north was a long A-shaped portico with mortar floors behind which we find small cubicula with black and white mosaics and white-washed walls. Further south was in the phases 2-4 a closed garden peristyle probably substituting the atrium of the first phase. The peristyle had opus sectile floors and an impressive water channel in front of the columns. Opening into the peristyle were several rooms again with opus sectile floors, and at its northeastern corner, opening into the long facade portico, was a sizeable triclinium also this with an opus sectile floor. This building block characterised by its rich opus sectile floors and coloured stucco walls (Fig. 8) contrasts to the previously mentioned building block further north (Fig. 5).

South of the peristyle is another building block, which is situated in the axis of the villa. This block had either two stories or a change of level, as we can see from the presence of a staircase leading from the peristyle to the rooms in the central block. Towards west at the villa’s highest point and almost in its axis lays the villa’s cistern (Fig. 1). The layout of the villa plateau from the central zone to its end in south is unknown to us, as it has not been investigated. However, south of the villa plateau, another wing had been added at a later time, probably in phase 3. This wing contains a huge horseshoe-shaped exedra, and here we find the opening of the emissary (Fig. 2). The overall use of the various building blocks of the villa plateau may be suggested as the following: in the northern end of the villa is the servants’ quarters, whereas a representational part may be found in the central area, and probably a private zone of the owner with living quarters and recreative areas (the exedra) in the southern half of the villa (Fig. 2).

We have no proof of a pars rustica indicating production in the villa. However, the site seems to feature several gardens. One of these, the one north of the proposed servants’ quarters may have been a productive garden with fruit trees. Between the main villa terrace and the lake is an extended, flat terrain. With the exception of a small channel, we have not yet been able to verify any building remains on this vast expanse, and it is probable that this was a garden too. The gardens in the open space between the A-shaped portico and the edge of the terrace and in the closed peristyle verified by the American garden excavation team in 2001 were probably ornamental gardens rather than productive ones.

With above remarks of a more general character, we’ll turn to the structures of the villa more in detail presented by phase.

Phase 1

The villa plateau

The first villa was, as already mentioned constructed on a huge terrace. This terrace remained the main feature of the villa. It was constructed in opus caementicium with an upper facing of opus quasi reticulatum made of basalt cubilia.

The early villa probably had an extension more or less similar to that of the later phases. At least in the northern half of the villa, in the area of the later bath, several structures incorporated as supportive elements under later floors, point to this conclusion (Viitanen Fig. *). These structures have the same orientation northeast-southwest, differing from the orientation of the main terrace, and from later orientations in the bath area. They were probably substrutions extending the villa north of the main villa plateau. We have no means of determining the character or use of this part of the villa in its first phase.

The villla possibly featured a long facade portico finishing at least part of the villa terrace. This is suggested by a monumental Doric capital found in trench CD (Fig. 4). It is made of peperino, with a maximum width of 0.91 m and with a diameter of 0.715 m. It is of the type with a tall, plain neck in common use in Central Italian architecture of the second half of 2^nd century BC to ca. 75 BC, in both private and public buildings.^[xi] In the Sanctuary of Diana Nemorensis, it was intensively used for the large and small porticoes all dated to the last quarter of 2^nd century BC.^[xii]

On the villa plateau, the remains of the first villa are very badly preserved due to the fact that the walls of the first villa were eventually torn down to soccle height or to the very level of their foundation. However, it is certain that the long foundation made of long, squared peperino blocks had its origin in this phase. In trench CB and DD the front wall of the later portico was found to be constructed directly on top of a phase 1 wall. This original wall was furnished with a 2nd style wall painting very little of which is preserved. It shows a Doric half-column without base on a dark bluish grey background. The wall is investigated to a length of ca. 6 m. We ignore, whether the portico was A-shaped in the first phase.

Whereas the building block of the A-shaped portico was basically unaltered in the later periods, the early structures south of this block were in later phases completely overbuilt with structures of a new layout. The understanding of this part of the building remains, therefore, very fragmentary. Under the garden soil in the closed peristyle a sizeable rectangular structure measuring ca. 5 x 4 m and with ca. 0.9 m wide walls came to light (Fig. 3 and 8). The walls were razed to the foundation. In the middle of the structure’s eastern side was found a drain. This probably connected with the old drain preceding the vaulted passage in the terrace constructed in phase 2. It is an attractive idea that the rectangular structure indeed are the remains of the impluvium of the first villa.

North of the ?atrium and under the later walls and floors of the triclinium in trench CE were found the remains of an, again, obliterated but wide foundation wall oriented perpendicularly to the terrace wall. South of it were traces of two rooms connected with a door opening. In the rooms were evidence of two different types of opus sectile floor, one with a pattern showing a square in a square, the larger squares being separated with a narrow fillet. No traces of the actual stone slabs are preserved, only the deep imprints in the mortar floor. The pattern in the second room consisted of one or more rows of small squares separated by multiple fillets, and of larger rhombs surrounded again by thin multiple fillets and intersected by a square in a square. Some of the fillets made of black slate unfit for reuse are preserved in situ. The square in square pattern is as such very popular over a long time, however, the use of the fillets framing the squares is an early trait, and it is found almost identical in the only opus sectile floor of the villa dei Volusii from 60-50 BC.^[xiii] The pattern with large oblong rhombs bordered by multiple fillet and combined with squares was particularly in vogue in the mid-1st century BC, but only preserved in form of the more common mosaic floors.^[xiv] Also the technique of the flooring points to an early date. It is characteristic that the back side of the slabs is convex and very thick in the middle leaving a deep central impression. A number of slabs with these characteristics have been found in later fill, particularly in trench CI, mainly consisting of hexagons but also smaller rhombs both made of palombino were found (*Birte). None of them come from the two mentioned floors, as they vary in size and shape. However, they attest to the apparently widely spread use of marble floors even in the first phase.

Tearing down the first villa

There is, as described above, no doubt about the fact that the first villa was torn down. When exactly, this took place, and why, still remains an open questions. It may have happened as late as the early Augustan period. Building elements from the first phase were widely reused in the villa of the second phase, e.g. as caementa in the walls and for the brick columns.

Phase 2

The villa plateau

In the early Imperial period, the villa was rebuilt more or less from the ground, even though features such as the terrace wall itself remained standing. The facade portico was probably reconstructed reusing the old capitals.

As found in trench CM, at the latest by this period, a ca. 3 m wide road paved with trapezoid blocks of basalt in parts supported on an artificial terrace with a facing of opus reticulatum was constructed. The road continued in a straight line from trench CM to the villa plateau where it was also found in trench CI.

In this phase a vaulted passage was constructed by cutting through the main terrace wall, thus providing access from the lower (garden?) terrace to the central part of the villa. The passage measuring 3.5 m in height and 1.9-2 m in width was furnished with very crude reticulate walls of basalt and with an irregular segment vault. At its present state, it is completely filled with soil ca. 8 m from the present entrance. However, a hole cut in recent times by illicict diggers through the mortar foundation of the stair case found in trench DA, reveals that the passage stops 26.6 m inside the terrace turning sharply at an angle of 90⁰ towards south and hence the central building block. We do not know where the vaulted passage surfaced.

At the villa plateau, the general level was heightened ca. 30 cm. As the portico, which was probably at least from this period A-shaped, was reconstructed on top of the old walls, the level of the portico was heightened by one block, and a new mortar floor was laid in the corridor. At least by this period, behind the portico were built a series of rooms with a common back wall and (reused) peperino thresholds. The technique of the walls is a rather sloppy opus reticulatum with corner blochetti.

Small rooms, approximately 3 x 3 m, were added to the northern facade of the portico (Fig. 3). These rooms were decorated with geometric black-and-white style mosaics. In one room found in trench BA, the pattern consisted of hexagons outlined with black on a white ground, and one room found in trench CA decorated with groups of rectangles outlined with black clustered around a black square. Both types of floors are well known from late 1st century BC to at least 2nd century AD.^[xv]

The layout with a closed peristyle probably originates in this period, but the remaining stuctures were much destroyed when the now standing peristyle was built in phase 3. In the centre of the later peristyle, in phase 2 a small basin was constructed, 2 x 2 m with an internal size of ca. 1.35 x 1.35 m. Its depth is unknown as it is later filled with mortar as a foundation of a statue base. The basin was cut into the ground and furnished with a facing of reticulate at its inside.

The northern wing

In trench DF, which was opened in order to investigate the relationship between the villa and the lake in the hope of finding part of its quai constructions a huge filling layer was encountered. This fill layer containing spolia of the first villa indicate substantial construction work in connection with the quai or embankment in phase 2. However, the actual walls were not found. With much probability, they are situated either beneath the modern track or under the present terrace constructed of retiform cages filled with boulders (Fig. 1-2). Neither can be demolished in order to verify the ancient remains.

The dating of this phase has not yet been established with certainty. However, the general profile of the fine ware pottery, in particular of the Terra Sigillata pottery, suggests a date in the late Tiberian period or slightly later (*Birte).

Phase 3

As far as our evidence goes, the layout on the villa plateau was basically created by phase 2. Phase 3, however, saw major restructurations and in the wings north and south of the main terrace, the construction of new major building elements.

The terrace support

The tall terrace wall had by this period apparently become partly instable. A support consisting of 59 fornices was therefore added in front of the terrace in its entire length (Fig. 1). Very little of the fornices’ original facade is preserved. However, by the entrance to the vaulted passage we can see that at least the new entrance and the northern pillar of fornix 35 was furnished with a facing of opus mixtum. A few traces of red paint, moreover, show that the visible part of the fornices were painted red.

Apart from the stabilising effect, the fornix structure had further two functions: a practical and an aesthetical. On its top it carried a well constructed gutter, which can be followed in the entire length of the terrace from north of the vaulted passage in plot 166 to the end of the terrace in plot 180. The terrace wall proper is not visible plots 154-155 and 198 as it is almost completely covered with soil (Fig. 1). At an earlier stage of the excavation, it was thought to be an aquaduct; however, its position and the fact that it was open to the sky show that it was a gutter in which water from the roof of the columned facade fell and was led away.^[xvi] Further towards the lake, it moreover supported a balcony-like walkway of which very little is preserved.

The addition of the fornix structure is fairly well dated. It is with certainty added later than the vaulted passage was cut through the terrace wall in phase 2. And in trench AG, we had the opportunity to excavate the fill inside the channel which was filled with material of the early Flavian period providing us with a terminus ante quem for the construction of the terrace support.^[xvii]

The cistern

The villa’s completely preserved cistern, situated almost exactly in the villa’s central axis and at its highest point, was at the latest constructed in this period also (Fig. 1).^[xviii] Its lower parts were cut into the rock, and inside this large cut rectangle, the cistern was built in opus caementicium with two barrel-vaulted aisles. The exterior length is 37 m, whereas the internal length is 34 m and the internal width is almost 8 m and its height almost 6 m. Internally and externally the cistern was furnished with nine pilasters securing the stability of the walls. The inside of the cistern was lined with a thick, finely smoothed layer of opus signinum. There were access from the cistern’s flat roof via two pozzi in the western aisle. Due to the soil and vegetation covering the cistern’s roof, it has not been possibly to verify the position of the aquaduct bringing water to the cistern. However, in all likelyhood it entered the cistern in its northern corner. This part of the cistern cannot be studied from the inside, as soil continuously washing down through one of the pozzi positioned in that end of the cistern, impedes the studying of the roof and back wall of the cistern in its northwestern aisle.

The water was conducted from the cistern in lead pipes. They are no longer preserved, but in the wall of the facade of the cistern impressions of two pipes can be seen. These pipes were originally conducted in a vaulted mortar channel, situated between the fourth and fifth buttress of the facade. The channel was 1.2 m wide and 1.8 m high, and a rectangular shaft measuring 0.6x0.66 m at the facade of the cistern provided access to the vaulted channel. The vaulted channel led the water towards the central or southern half of the villa. In antiquity, an attempt was made to construct yet a conduct for an outlet between the sixth and seventh buttress of the facade. However, this was given up and the work was abbandoned.

The cistern was used for living and shelter during World War 2, in occasion of which the bottom was cleared of all ancient remains. We do, however, have some very precious evidence providing at least a terminus ante quem for the dating of the cistern. The rectangular shaft and the vaulted channel are not built in one with the cistern. A hole was cut into the cistern facade at the level of its floor from the outside. Through this hole a lead pipe was conducted, and the larger hole cut filled with mortar so that water would only exit via the pipe. In the channel, the pipe was positioned on a sloping bed of mortar, evidently added when the pipe was actually placed in the channel. This mortar bed was filled with large chunks of pottery, but also a few pieces of fine ware pottery providing a date in period 3 were found. The cistern is, accordingly, at the latest constructed in this period.

In both of the short sides of the villa at each end of the villa plateau, as mentioned above, the building was significantly altered in this phase. In north, a structure in opus mixtum was added as a support of the road terrace consisting of two parallel fornices in total 9 m wide, which may have carried a monumental entrance to the villa (Fig. 1). The upper part of the structure is unfortunately not preserved.

The exedra

In the southern end of the plateau is the villa’s most impressive standing element, a huge horseshoe-shaped structure built in opus mixtum (Fig. 1).^[xix] The preservation of the building is precarious.^[xx] During the campaign in 1999, the visible parts of this grand building were cleaned and measured manually and digitally with a total station.^[xxi]

The building is made by cutting the porous pozzolana rock and founding the structure directly against it. The cubilia used for the facing of the walls are of tufa of a very bad quality and accordingly poorly preserved. The exedra’s preserved visible height is 10.5 m, and the total width including the two wings, is 48 m. The wings are 13.5 m wide and the horseshoe-shaped room itself has a diameter of 21 m and a depth of 17.5 m. Towards the lake in both sides of these walls were two shallow niches. The horseshoe-shaped room and the two wings are constructed in two stories. The wings originally had 2 + 2 fornices placed on top of another, whereas the central room had two annular corridors also placed one on top of the other.

Cleaning revealed that the lower annular corridor had a well preserved brick profile including a dentil frieze situated immediately beneath the vault. It is preserved to a height of ca. 2.4 m which gives an indication of the original height of the wall. The original height of the entire structure was probably ca. 15 m, perhaps higher.

We have not yet been able to verify neither date nor function of the room. A possible parallel may be the so-called sepolcro di Agrippina, a theater-shaped nymphaeum belonging to one of the rich 1^st century AD seaside villas in Campanian Bacoli (ancient Bauli) and of almost the same size as the Nemi structure.^[xxii]

The villa plateau

It was not just in the outer ends of the villa, the building was significantly altered in phase 3. Also on the plateau itself several structures were either constructed or rebuilt. The rooms behind the portico as excavated in trench DD were altered with the construction of narrow mortar walls dividing the larger room(s) into smaller rooms. It is unknown what necessitated the alteration of the layout of the rooms. With the new walls followed also new wall- and floor decoration (Fig. 5).

The southernmost room was probably a service room with a pavement of opus spicatum. It was connected with a door to the larger room north of it. This room was equipped with a floor featuring a black and white mosaic with the same pattern of rectangles grouped around a square also employed floors of the preceeding phase. The narrow room further north had a white mosaic floor with a narrow black fillet following the walls. Finally, the last room in trench DD, had also a black and white mosaic with a pattern of double-T maeanders. This pattern is not very common, but it is found e.g. in the Villa of the Volusii at Lucus Feroniae (room 7),^[xxiii] and in Pompei in the triclinium of Casa di Lucretius Fronto.^[xxiv]

All of the walls were left undecorated (white), only with a stucco profile featuring a stylised leaf-pattern at the transition from wall to ceiling (*Birte). Also the ceilings were white.

Behind the rooms of the portico was in this phase constructed at least one, but probably two parallel but separated corridors. They were divided by a ca. 0.80 m wide wall, constructed ca. 1 m from the back wall of the portico. The 1 m wide corridor behind the portico was furnished with opus signinum on walls and floor, whereas the parallel corridor (or room?) further west was furnished more pretentiously with shallow niches, 1.65x0.6 m. We ignore the wall- and floor decoration of this period, as it was substituted in phase 4.

In the building block of the closed peristyle, new walls functioning as back- and front walls of the rooms opening into the peristyle were constructed. In trench DA, the remains of three rooms have been established (Fig. 3). The two northern rooms were connected with an internal door. From the middle room was also access to a room or corridor situated west of the mentioned rooms. It is highly probable, that it is during this phase, a staircase leading from the peristyle to a room at a higher level or an upper storey in the central, but unexcavated building block, was constructed.

The peristyle itself measure ca. 15 x 22 m (Fig. 3 and 8). In all probability, it featured 6 x 10 columns with an enlarged intercolumniation in the two axes. We ignore the architectural style of the columns. The columns were monoliths of peperino furnished with stucco cannelures, and in the following phase, a torus was added as a base to the columns indicating their style to be either Ionian or more likely Tuscan. It is at the latest in this phase a wide gutter channel was constructed immediately in front of the peristyle, 1.1 m wide and 75 cm deep. The gutter was open to the sky and it functioned as a decorative element like a basin, as it was probably continuously filled with water and thus creating a small artificial island in the centre of the peristyle connected with the corridor of the peristyle with one or more “bridges”.

In this or in the following phase the small basin in the centre of the peristyle was filled with mortar in order to reshape it into a statue base.

The entrance area where the access road entered the villa plateau was changed considerably. The old access road was interrupted, and a new suite of rooms opening towards the north were constructed against the back wall of the cubicula of the preceding period. The newly built rooms were furnished with black and white mosaic floors with geometric patterns, one with a pattern like the game Nine Men’s Morris in black on a white background (Fig. 6) and one with a pattern of white four-pointed stars. Whereas the last-mentioned pattern appearing in the mid-1st century AD is fairly common, particularly in the second half of the 1st century AD and early 2nd century AD,^[xxv] the first is only infrequently found. ^[xxvi]

The building of the rooms signifies an extension of the living quarters beyond the old villa plateau. The rooms probably opened towards yet a portico. Part of its foundation wall was located. However, time did not permit us to investigate this part of the villa further. It can, though, clearly be seen that this entrance area carries much resemblance to the villa of Lucus Feroniae, and the function of this part of the villa was probably also the same: to provide habitation for the servants.

The new suite of rooms added to the older block of the villa, was not the only addition in this area. In the lower part of plot 150, already from the first phase constituting part of the villa, the area of the later bath was completely rebuilt in opus mixtum and with a new orientation of the walls (see the contribution by E.M. Viitanen).

In order to connect the structures at the villa plateau with this lower-laying zone, a ramplike road supported by a terrace wall was constructed. The road probably entered the zone of the A-shaped portico in its eastern corner. However, in this zone the road is completely removed due to recent agricultural activities, and the modern surface level is lower than in antiquity, so this cannot be verified. The 3.1 m wide road made of trapezoid blocks of local basalt and with well preserved kerb stones at its western side is preserved to the length of 40.75 m (originally ca. 57 m), as it stops immediately to the north of the building which is evident from the large rectangular blocks placed across the road at its end in trench AB.

Phase 4

The villa plateau

At the villa plateau, rennovation and embellishment are the keywords for the activities taking place here during phase 4, rather than new construction. It is during this phase all visible marble floors and marble wall decoration were added. In the building block of the A-shaped portico, in trench DD the largest of the four rooms opening into the portico was redecorated with a figural mosaic added to the existing floor as central decoration (Fig. 7). The new mosaic consists of a square with an inscribed circle. Inside the circle are four exotic birds, and in the centre is yet another bird smelling a flower. In the four corners between the square and the circle are stylised vegetal scrolls, and above them inside the circle, four equally stylished palmettes. Facing the entrance is an inscription reading: M.PA(V)IMENTUM.FECIT.

The shape of the room, the addition and position of the bird medallion and the find inside the room of a miniature Corinthian capital in white marble (*Birte) lead us to suggest that the room was the household lararium.

The shallow niches of the western corridor were walled up til in order to create a uniform wall. The new wall was furnished with a thick layer of mortar, which provided the backing of soccle and wall marble veneer. Also the floor was in all probability furnished with marble plaques. However, in connection with the plundering of the villa, the marble was removed from wall and floor only leaving behind the impressions of the plaques in the backing mortar.

Further south, the peristyle was also furnished with marble decoration (Fig. 8). Between the columns were added slabs of white marble, and the rim of the gutter basin along with the channels sides and bottom were also clad with slabs of white marble. The rim of the channel was constructed as a low “box” of mortar revetted with white marble in order to create a low parapet separating the channel from the “island” inside the peristyle.

The floor of the peristyle corridor has been verified in several trenches (AR, AS, CD and DB). In trench AS and DB it is partly preserved, and in the remaining trenches only preserved as impressions in the mortar as the marble plaques have been robbed. The pattern consists of rectangles 59x29.5 cm surrounded by a 4.5 cm wide fillet. Alternating rows are displaced half a rectangle of a the type called isodomo listellato. The marbles employed are portasanta for the rectangles and pavonazzetto for the fillets. This type of flooring is well known in Italy, and the combination of portasanta and pavonazzetto is considered cannonical by Guidobaldi.^[xxvii] The type of floor is mainly found in the 1^st century AD and in the first half of the 2^nd century AD with prominent examples in the in the Domus Aurea on the Colle Oppio,^[xxviii] not to speak of the many examples found in the Villa Adriana by Tivoli, predominantly in the Biblioteca Latina and in the Canopus.^[xxix]

The same pattern but with a larger module (rectangles: 74x44.5 cm; fillets: 6.5-7.5 cm wide) was employed in the entrance to the facade portico and perhaps used in this portico also as found in trench AT, but as this floor is only preserved as impressions in the mortar underfloor, we have no means of deciding which types of marbles were employed.

It was in all likelyhood in this phase the sizeable room at the peristyle’s eastern corner was furnished with a build bench(?) indicating the room’s function at least by this period as a triclinium. It opened into the peristyle and also into the long facade portico and, thus, provided the room with a view over the lake. Also this room had an opus sectile floor with a pattern resembling the above mentioned, but with a displacement of the rectangles of only 1/3 in the second row.

At least two of the rooms in the peristyle’s southwestern corner (in trench DA) were also rebuilt. The passage between the middle room to the corridor or room west of it was blocked. The middle room and the connected room to the north were furnished with marble on walls and floor. The wall decoration has been completely plundered with the exception of marble fragments sitting by the floor showing that the soccle was made of cippollino marble and only the impressions of the marble veneer is left in the backing mortar. The floors, on the other hand, are completely preserved, whereas the marble threshold between the two rooms had been removed too. The floor of the middle room consisted of narrow, parallel bands separated by fillets of a bluish marble (Carrara?). Inside the bands were black slate triangles bordered by narrow yellow giallo antico triangles creating a pattern of series of black arrows on a yellow background.

In the room to the north, the floor consisting of rows of rhombs of porta santa marble placed at an angle with triangles of bluish marble by the edges was also left in place. The southen of the three rooms was differently equipped. The western wall still maintained the white wall stucco of the 2nd phase and the northern wall the red stucco added in the following phase when the wall was built. A mosaic floor made with tiny tessera had been removed apart from a few remains along the walls.

The room or corridor west of the suite of rooms was furnished with a thick layer of opus signinum painted red on the wall, and the floor was a plain mortar floor providing the impression that the room was a service corridor.

The bath

In the lower, eastern part of plot 150, the building constructed against the terrace wall SU 2 during phase 3 was completely rennovated. The new orientation established in the preceeding phase is retained, as is the main layout of the area. In phase 4, the building with certainty functioned as a small bath (see the contribution by E.M. Viitanen).

The villa is abandoned

Apparently, the villa was abandoned in the mid- or perhaps third quarter of the 2^nd century. The latest brickstamp found in the villa is of the year 134 AD, and ceramic finds that can be dated later than ca. 150 AD are neglegible.

When the villa was abandoned, it was apparently emptied of all valuables and it was partly ribbed of its marble ornaments. The reason for the villa’s abandonment is not yet completely understood. However, it may have fallen subject to a natural catastrophe that also necessitated repairs in the nearby Sanctuary of Diana in the same period.^[xxx] It may have been despoliated in order to provide material for the reconstruction of the sanctuary. The finds of well preserved ceiling stucco in the room with the bird medallion immediately above the floor and a wall fallen on top of a pile of marble chips in trench DB show beyond doubt that the villa hardly survived long from the period of abandonment to its final collapse. Also the sanctuary was abandoned in the second half of the 2nd century AD.^[xxxi]

Phase 5, late antiquity, probably 6th century AD

In late Antiquity, perhaps in the 6th century, the villa site was used for a widespread necropolis. At least 10 tombs were found: eight in trench CE, one in trench CM, and one in trench DD. All of the tombs are inhumation tombs. They in part reuse the ancient villa walls, in part are they build with recycled building materials from the villa.

In trench CE were excavated two adult’s graves, both of which probably pertained to males, and one child grave. One of the male graves was found completely intact including its cover tiles. The other male tomb had been disturbed, and the skeleton had been pushed aside. However, the last mentioned tomb contained a small vessel of a local polished ware (*Birte). Remains of one further deposition was found. It was almost completely destroyed, but the remains of the jaw and two bronze earrings in their original position at each side of the jaw suggest the remains of an interred woman. In the child grave, no finds were made with the exception of the more solid parts of the skeleton. Further four tombs were noted in the trench. However, time did not permit the opening and excavation of them.

In trench DD a single tomb of a newborn was found. It was of a very poor construction compared with the above mentioned tombs. A few minute snail shells were seemingly included as grave goods.

In trench CM remains of yet a tomb were found, unfortunately with three-quarters of it sticking in the baulk. Again, time did not permit the removal of the 2 m high soil baulk above it and therefore, it could only be partially excavated. No grave goods were visible, but two adults were interred in the tomb. This tomb shared the same characteristics as the ones in trench CE situated no less than 130 m away as the crow flies. It is very interesting that apparently contemporary tombs are spread over such a vast distance. The whereabouts of the contemporary habitations are completely unknown.

Phase 6, Middle Ages, Renaissance

In one of the rooms in the bath, extensive remains of calcified marble dust on walls and floor suggests the presence of a lime kiln.^[xxxii] This corresponds with finds in the bath trenches of many amorphous pieces of marble destined to be burnt to lime. In the same area, moreover, the excavation’s highest concentration of Medieval and Renaissance pottery was found providing us with a date of the lime-burning activities. This lime-burning is probably to be viewed in connection with the building activities in the local churches and monasteries.

Pia Guldager Bilde

University of Aarhus

Denmark

Bibliography

Andersen S.F. & Guldager Bilde P. 2000: “Nemi, loc. Santa Maria and the application of computer technologies in field excavation”, CSA Newsletter 13.1,

Ghini G. 2000: “Ricerche al santuario di Diana: risultati e progetti”, in: Nemi - Status Quo, 53-64.

Guidobaldi F. 1994: Mosaici antichi in Italia. Sectilia pavimenta di Villa Adriana, Roma.

Guldager Bilde P. 1999: “Ved Dianas hellige sø. Nordiske udgravninger i Nemi, loc. Santa Maria”, in P. Guldager Bilde, V. Nørskov & P. Pedersen (eds.): Hvad fandt vi? En gravedagbog fra Institut for Klassisk Arkæologi, Århus, 181-194.

Guldager Bilde P. & Poulsen B., “Caesar’s villa? Nordic excavations of a Roman villa by Lake Nemi, loc. S. Maria”, AnalRom in print

Moretti M. & Sgubini Moretti A.M. (eds.) 1977: La villa dei Volusii a Lucus Feroniae, [Roma].

Nemi, Status Quo, Occasional Papers of the Nordic Insitutes in Rome 1, Rome 2000.

Illustrations

Fig. 1

Overview over the site. The large numbers identify the plots. Harri Kiiskinen & Pia Guldager Bilde (April 2002).

Fig. 2

Overview over the site, interpretation. Harri Kiiskinen & Pia Guldager Bilde (April 2002).

Fig. 3

Overview over the northern half of the villa plateau. Harri Kiiskinen & Pia Guldager Bilde (April 2002).

Fig. 4

Doric capital, found in trench CD (2000).

Fig. 5

Overview over trench DD from east (2001).

Fig. 6

Mosaic floor (SU 222) in trench CI (2001).

Fig. 7

Medallion mosaic inserted into earlier mosaic floor (SU 177 and 178) in trench DD (2001).

Fig. 8

Overview over trench DB from * (2001).

Notes

^[i]The project is directed by a steering committee consisting of the directors of the four Nordic institutes in office: the Danish Institute: Jan Zahle (1998-2000), Gunver Skytte (2000- ); the Finnish Institute: Christer Brun (1998-2000), Christian Krötzl (2000- ); the Norwegian Institute: Rasmus Brandt (1998-2002); the Swedish Institute: Anne Marie Leander Touati (1998-2001), Barbro Frizell (2001- ) and the field director, Pia Guldager Bilde, and its logistic base is at the Danish Institute (Karen Ascani).

^[ii]We would like to express our heartfelt gratitude to the soprintendente, Anna Maria Reggiani and to our local collaborator, the director of the Museo delle Navi at Nemi, Giuseppina Ghini for their openness and confidence in and full support to the excavation project.

^[iii]It is a great pleasure to extend our sincere gratitude to the foundations for their generous economic support.

^[iv]A group of six American archaeologists directed by Irene Romano (University of Pennsylvania Museum of Archaeology and Anthropology, Philadelphia) and Kathryn Gleason (Cornell University, Ithaca, New York), assisted by James Shryver. They are financed by the University of Pennsylvania Museum of Archaeology and Anthropology, Philadelphia. Their participation will continue also in 2002.

^[v]The Norwegian group is directed by Birgitte Skar, and their work is financed by the Foundation for Cultural Heritage Research (NIKU). Provided it obtains the necessary economic funding, this project will continue in the coming years.

^[vi]Published articles on the Nordic excavations have mainly been in the Nordic languages and of a popularising character.For a complete list, see http://fc.hum.au.dk/~klapg/bibliography2.htm. Contributions in English or Italian:

^[vii]A. Arnoldus-Heyzenfeld, Rocca di Papa, and crew.

^[ix]G. Lenzi 2000: “Il territorio nemorense dalla preistoria al medioevo”, in Nemi, Status Quo, 171-172, no. 24.

^[x]G. Ucelli 1950: Le navi di Nemi, Roma, 45-54; V. Dragoni & W. Castellani; “Opere archaiche per il controllo del territorio: gli emissari sotterranei artificiali dei laghi albani”, in M. Bergamini (ed.), Gli Etruschi maestri di idraulica, Perugia 1991:43-60; K. Greve, Licht am Ende des Tunnels, Mainz 1998, 82-87; M.J.T. Lewis, Surveying Instruments of Greece and Rome, Cambridge 2001, 113, 197, 200, 202-203, 208-209, 215.

^[xi]R. Delbrück, Hellenistische Bauten in Latium II, Strassbourg 1912, 151-152.

^[xiii]Moretti & Sgubini Moretti 1977, pl. XLV.

^[xiv]See references in Guldager Bilde & Poulsen, in print.

^[xv]See references in Guldager Bilde & Poulsen, in print.

^[xvii]Guldager Bilde 1999, fig. 7 (support in relation to vaulted passage); figs. 9-11 (some of the finds made in the channel).

^[xviii]L. Devoti, Cisterne del periodo romano nel Tuscolano, Frascati 1978, 21, figs. 19-20; Guldager Bilde 1999, 190-191, figs. 12-14.

^[xix]Guldager Bilde 1999, figs. 15-16; Andersen & Guldager Bilde 2000, figs. 1-5.

^[xxii]A. Maiuri, “Il teatro-ninfeo detto “sepolcro di Agrippina” a Bacoli”, Antemon. Studi Carlo Anti, Firenze 1955, 263-271.

^[xxiii]Moretti & Sgubini Moretti 1977, pl. XXIX.

^[xxiv]Pompei pitture e mosaici IX, Roma 1999, 254, fig. 168, 256, fig. 170.

^[xxv]See references in Guldager Bilde & Poulsen, in print.

^[xxvi]See references in Guldager Bilde & Poulsen, in print.

^[xxviii]L. Morricone Matini, Mosaici antichi in Italia I, Roma 1967, 70 no. 66, fig. 26.

^[xxix]Guidobaldi 1994, 103-104, rooms 18‑20, 22‑25, 89, 97, 115, 118‑119, 121, 126‑127, 138.

^[xxxi]G. Ghini 1997, “The New Excavations in the Sanctuary of Diana”, in I Dianas hellige lund. Fund fra en helligdom i Nemi (In the Sacred Grove of Diana. Finds from a Sanctuary at Nemi), Copenhagen, 181.

^[xxxii]The identification was proposed in 2000 by a visiting colleague, the name of whom I, unfortunately, never got hold of.