The "Alburnus Maior" National Research Program

           Reports of archaeological objectives on the central area of the Carpeni Hill are rather numerous, namely the place called "Bisericuţă". Although this location has been mentioned several times in the literature, the data concerning the archaeological investigations that were carried out there are inaccurate; that is why we resorted to confronting this data with verbal information from those who know the area [1].
           The discovery of materials from a building equipped with hypocaust installation was briefly mentioned on several occasions. Most of the times, such historical attestations were related to the presence of Legio XIII Gemina at Alburnus Maior [2]. In an extremely brief note, without mentioning the exact spot where the excavation took place, V. Wollmann mentions the discovery of a brick bearing the stamp of the Gemina XIII legion: "In SII (17. 07. 1986), in a massive layer of stone debris and fragmentary tiles, at -0,40 m" [3]. V. Moga, at its turn, mentions that this discovery of tegulae in 1986 was made during an archaeological (?) research on the massif plateau - the point known as Bisericuţă - the authors of the intervention mention that they discovered debris belonging to ancient edifices (?), from which they recovered a fragment of a brick with the Legio XIII Gemina stamp, which seems to indicate a small fortification, erected for the surveillance of the area" [4]. However, the only Roman fortification in the Roşia Montană area was discovered at Abrud ("Cetăţeaua" point) [5].
           In October 2001, was opened the control section S22 (14 x 2 m), oriented East-North-East West-South-West (Fig. 1). The results of excatation in this section were modest. At the western part of the section we intersected one of sections excavated thirty years ago, and at the other extremity was identified an almost circular pit, filled with large quarry stone, among which we found the upper part of a votive or honorific monument.
           In order to investigate the entire area assigned to us on its entire length, we traced a new section continuing S22 to the West: S23 (14.50 x 1.50 m). This yielded the expected results, namely the identification of three walls: Z1, Z2 (perpendicular on Z1) and Z3 (parallel to Z1). Under these circumstances, we expanded the surface of excavation as it follows: a surface to the North, S24, of 4 x 9 m, keeping an 1 m earth witness to S23; this was eliminated later on, after drawing the profile, the extension of S23 towards the South, between meters 0-7 of the initial section (without keeping an earth witness), as well as the expansion towards the East and South-South-East, by the opening S26 (10 x 1.50 m) and S25 (10 x 2 m).
           A large building, belonging to the Roman era, was identified (Fig. 2). The course of the seven walls that we investigated reveals the buildings orientation on the East-North-East West-South-West direction; also was identified the perimetric wall on the northern side, Z6. This is doubled, towards North, by a linear stone substruction - Z7, not faceted, approx. 0.75 m wide. Z7 is parallel to Z6, 1 m away from it, and was also identified at the northern limit of S26 and in the western limit of S22. In fact, Z7 proved to belong to the third stone phase, see below.
           The first archaeological level was identified, sporadically, in the eastern part of the S23 section, as a brown-gray level, approx. 0.05-0.06 m thick (Fig. 3). In this level, was discovered Roman vessel fragments, a stamped clay lamp, manufactured on the potters wheel and a bronze item fragment.
           The first stone phase (Fig. 4) was discovered in the S24 surface and in the S25 section; it consists of two walls belonging to a rectangular building. Thus, Z3 was uncovered over a length of 0.70 m (in S24 alone), representing part of the building western side, 0.60 m thick; the wall was built using the opus incertum technique, and it was placed straight on the native rock. Z1 represents the northern wall of the building, and it was uncovered both in S24 and in S25 over a length of 8 m. It was built using the same technique as Z3, and it is 0.75 m thick. The treading level of this building was not identified, because the building was demolished and these two walls were reused during the second stone phase. We did not discover any archaeological materials that might allow us to date this phase.
           The second stone phase partially consists of three rooms with hypocaust heating system: room A, room B, room D and a room with no heating, room C (Fig. 5).
           Room A (delimited on three sides by Z1, Z2 and Z8) is rectangular, 5 m long on the East-West axis, while only a length of 4.5 m was uncovered on the North-South axis. The Z1 was reused from the first stone phase. The walls are approx. 0.80 m wide (only two rows were preserved out of Z8 wall). The room was built with a hypocaust installation, made upon a brick concrete substruction (cocciopesto grossolano [6]). 7 rows of piers were discovered, with a maximum number of 6 piers per row. The bricks at their base are square and most of them are 5 cm thick and have 26 cm side (pedales ?). Z1, which is the northwestern limit of the room A, was plastered on the inside up to 0.20 m high, as in the case of Z2, using the opus signinum technique.
           Room B (delimited by Z2 and its extension towards the North - Z2`; Z3, Z5 and Z6) has the dimensions of 4.35 x 4 m.
           In this phase, Z2 represents Z2's extension towards North. We can notice that this portion of the wall was later subsequent doubling to Z2 course. Z2 northern limit is marked by an interruption 30 m away from Z6, intended for the c1 drain to function in that space. Z3 extends 1 m towards West, over the native rock, thus widening room B, whose western limit is now Z5. This one, like Z6, with which it closes in a right angle in the northwestern corner of room B was found in a good state of conservation, with an elevation of six rows of stones bound with mortar.
           In room B, on Z3 surface, up to a height of 0.20 m, was identified white lime plaster, very similar to the substruction of the room with hypocaust installation. The hypocaust alignment of piers was placed on this structure and a channel for water drainage was also built, its covering bricks being set when the pavement was cast.
           The drain c1 was discovered 1.85 m away from the southwestern corner of room 1 and continues up to northeastern corner of the room, where a space was left between walls Z2 and Z6 to allow water to be evacuated room C. The channel had a depth of 0.30 m from the hypocaust substruction level and a width of 0.20-0.25 m; it was made out of stone bound with mortar. Eastwards of the channel, we discovered three alignments of 4 bricks each, with the dimensions of 0.28 x 0.22 m, placed differently form the other pier in the room. On the eastern side of this building the trace of a passageway was preserved, which allowed the hot air to flow into room A. This passageway was built out of two rows of three bricks each, with the dimensions of 0.28 x 0.22 m.
           Only the northwestern corner of room D was uncovered in S25. It is limited by Z1 at the North and by Z8 at the East. This room was also heated; only two rows of two hypocaust pier each were discovered.
           Room C was delimited by Z1, Z2 and Z6. The eastern side of this room could not be found because of a modern-time pit. This room was not heated and its treading level was not discovered either. The known dimensions are 2.5 x 3.7 m.
           West of room B we discovered the traces of a paving made of bricks and stone slabs, preserved in situ, which seem to come from the edifice's entrance.
           The hypocaust installation was implemented in three rooms, as was mentioned.
           In room A the pier substruction was made of an approximately 0.15 m thick layer of cocciopesto grossolano, consisting of a row of river stones placed on their edge, as we could see in a sounding made at the base of Z4, in S25. Therefore, the concrete was cast over a layer of river stone, placed aslant. The stone was placed on a layer of well-settled soil. This type of concrete has a very high resistance and that is why it was especially used for hypocaust installations.
           On room A surface, the brick which forms the base of the pier is square, with a 0.26 m side. The pier was placed in parallel rows; some of them were slightly displaced due to the installation's collapse. The bricks which formed the actual pier were also square, with a 0.19 m side.
           In room B, the pier substruction is similar to the one in room A, but when the hypocaust installation was built a water draining was also made, whose lid of rectangular bricks of 0.28 x 0.22 m was set into this mortar substruction. The hypocaust pier in this room has a broken brick of the above-mentioned type at the base, so that they have a square shape. As in the case of the pier in room A, the bricks which formed the pier body were square, of the bessales type (with a 0.19 m side). A pile contained seven bricks, including the base of the pier; this resulted from the identification of a complete pier collapsed at the northern limit of room B, next to Z6. Its height corresponds to the treading level identified, in situ, west of Z5.
           The c1 duct was used for water evacuation. We were able to establish the water flowing direction, namely from south towards the north, but we cannot specify where it collects the residual water from or where it is disposed of.
           The third stone phase. In the S23 section, on the S24 surface and in the S25 section, we discovered a massive 0.85-0.90 m thick stone wall (Z4), which intersects the demolition level of the second stone phase and, at the same time, is built directly on the hypocaust substruction of the second stone phase (Fig. 6). The wall upper part was discovered 0,25 m below the present treading level. The wall is made of limestone, bound with mortar; among some rows we discovered opus signinum fragments resulting from the debris of the second phase. The building treading level in the third phase was identified in S25, South of Z4 and in relation to it (Fig. 7, western profile of S25-S26), at 0.45 m below the present treading level; it consists of a (maximum) 0.15 m thick flooring, mostly made of mortar, with small brick pigments. North of this wall and relatively parallel to it in the S24 surface, in the S26 and S22 sections, we discovered the substruction of wall with a maximum width of 1.2 m, placed directly on the ancient humus. This substruction only consisted of limestone slabs that were not bound with mortar. The upper part of this substruction was discovered 0.10 m below the current treading level and it goes deep about 0.80 m in S26 (Fig. 7, S25-S26, West profile).
           The building that functioned during the 3^rd phase was partially superposed on the building from phase 2; it had almost the same orientation as the latter, but it extended towards the southern area of the plateau - an area that we have not investigated exhaustively during this campaign.