The "Alburnus Maior" National Research Program

           The Archaeological Research

           One of the most important gold exploitation sites in the Roman province of Dacia was Alburnus Maior (Roşia Montană, Alba county). Over the last two centuries a rich archaeological and epigraphic material was discovered in the area [1] (Fig. 1). One of the most well known sites in the region, particularly from an epigraphic point of view, is the Hăbad plateau.
           Lying northwest of the Brădoaia massif, the plateau is situated at 878 m altitude, having the shape of a spur and surrounded by forest on three sides.
           The first archaeological researches were started off as a rescue-mission and took place during 1983-1984. On this occasion, 27 intact or fragmentary altars, archaeological materials (fragments of pottery, two Roman lamps), parts of the Roman road connecting Ampelum and Alburnus Maior, as well as a Roman gallery (Figs. 2-3) were discovered [2].
           In year 2000 new researches were carried out on Hăbad plateau, leading to the discovery of river-stone walls (partially dislocated), ceramic material and parts of the above-mentioned Roman road [3].
           In year 2001 the plateau was archaeologically investigated for the third time by two teams from the Institute of Archaeology and History of Arts in Cluj-Napoca [4] and the Museum of Dacian and Roman Civilisation in Deva [5] (Figs. 4, 10). The archaeological research of the area was assigned to the team from Cluj and the process was rendered more difficult by the bulldozer excavations carried out before 1983 and after 1984. Also, the western side of the excavated area had been narrowed down further to the widening of the access way built for the geological research works.
           In order to establish the overall stratigraphy, 4 sections (S1-S4) had been initially drawn on the eastern side of the plateau, next to the skirts of the forest. The full excavation of the plateau was achieved using 4 x 4 m cassettes, with 0.2 m earth witnesses (C1-C32). Exceptions to this working procedure were cassettes C14 and C27-C30, that have different sizes because of they neighbour on the northern and western sides of the plateau. A total area of about 800 sq.m. was investigated (Fig. 4-5).
           On the following lines will be presented the four sections (S1-S4).
           S1/2001 - 13 x 2 m, North-North-East-South-South-West oriented, stratigraphy: 0-0.30 m - black soft vegetal layer with sporadic ceramic material; 0.30-0.50 m - yellow, solid clay layer, with scarce traces of pigment; 0.50-0.70 m - sterile, light yellow layer, representing the geological sub-layer of the promontory. Between meters 1-4 of the trench was identified at -0.10 m a level of debris laying on the very yellow clayish layer. After removal of the debris the foundation of a Roman wall (Z6) was revealed at -0.4 m at the southern corner, on a 1.8 m span, 0.80 m in thickness and North-North-East-South-South-West oriented (Figs. 6 and 10/6). Foundations were built out of river stone and bound in yellow clay.
           S2/2001 - 10 x 2 m, was opened as an extension of S1, has the same stratigraphy. One 1 m wide earth witness was spared between the two Sections. About 4 m away, on the very grounding of the trench, extending beneath the western wall, we identified a slab which seems to bear certain marks of craft, while 9-10 m away another ruin was found laying 0.2 m deep (Fig. 5).
           S3/2001 - 10 x 2 m, continuing the previous two sections described above, having the same stratigraphy. At the 1^st-3^rd mark meters the ruin found in S2 continues a possible course of an unfilled wall, presumably 3 m long and 0.4 m deep.
           S4/2001 - 15 x 2 m, North-West-South-East oriented, located on the brink of the promontory (according to the layout of the site) and relatively out of center in comparison with the previous sections. The following stratigraphy was observed: 0.30-0.40 thick yellow clay layer, with pigment, scarce archaeological material; 0.40-0.50 m sterile light yellow soil with the same structure as the previous sections. 14 m away, approx. 1 m from the brink of the terrace, the vegetal layer revealed the track of an unfilled wall (Z5), 0.4 m thick, probably part of a building fallen on the northern slope of the terrace promontory. Only one row of the foundation was preserved. In order to reveal it, a small cassette (2 x 3 m) was opened on the north-eastern side of S4 (Figs. 5 and 7).
           The first row of cassettes, four in number (C1-C4), was oriented parallel to the western side of S1/2001 section, up to the brink of the promontory. The stratigraphy is similar to section S1, except for the sterile layer that is slightly different in depth. There is sporadic archaeological material, mostly ceramic, as well as some small iron items, poorly preserved. The ruin reveals itself in all the cassettes right beneath the vegetal layer, just like the first two meters in S1.
           After the ruin was removed, were uncovered the foundations of a Roman wall in C2 (Z3) all over the extent, 0.8 m thick and 0.3 m of preserved height (Figs. 5, 6 and 10/6).
           The second row of six cassettes (C5-C10), similar to the previous one, was oriented parallel to the western sides of S1 and S3, with a 0.20 m wide earth witness. In the West profile of C6, at 0.30 m deep, the pedestal of an altar was discovered. This seems to be in situ. The ruin extending along the 9^th-10^th meters of S2, i.e. the 1^st-3^rd meters of S3, was identified in C8-C9, right under the vegetal layer. The stratigraphy is the same, as well as the archaeological inventory. In C7, more precisely in its south-western corner, two Roman inscriptions were discovered, at 0.6 m deep (Figs. 23/2 and 24/7). In C10 cassette, after removal of the ruin, a new unfilled wall becomes visible from beneath. It is East-West oriented, being 0.70 m thick. Cassettes were conventionally numbered, from C11 to C32. The stratigraphy discovered in the cassettes matches the one identified in C1-C10, i.e. at 0-0.30 m deep - black soft vegetal layer with sporadic archaeological material; 0.30-0.50 m - yellow, solid clayish layer, poor archaeological material; 0.50-0.70 m - sterile, light yellow layer, representing the geological sub-layer of the promontory.
           The main characteristic that becomes apparent over the entire searched area is the ruin, more or less compact, which does not allow us to see very clearly the duct of walls. In principle, the compact ruin becomes evident at the periphery of the plateau that has been assigned to us. The clear course of a wall (Z3) partially exposed, which can be distinguished in C2 cassette, also extending in C5 cassette with a North-East-South-West orientation (Figs. 6 and 10/6), it continues in C30 cassette which is part of the sector assigned to the Museum of Dacian and Roman Civilisation in Deva. Considering this aspect, the colleagues from Deva agreed that we should at some point investigate the respective area. We can notice on the respective area the course of a wall, almost entirely revealed, conventionally labeled Z6, which can also be distinguished in section S1 (Fig. 6). This wall has the same aspect as the other walls described above, 0.80 m thick; it also ends "abruptly", just like the Z3 wall. The above-mentioned walls represent an apsidal building, of c. 9 x 5.5 m (Figs. 5, 6 and 10/1-6). Considering that the sidewalls end "suddenly", it is very likely that some rods, not revealed yet, support the front side of the building. Approximate sizes of the walls are 0.80-0.90 m. Approx. 0.50 m away from the revealed wall (Z7), which draws an apse, in the same cassette C30, there are some quarry rocks extending about 0.5 in width without the possibility of distinguishing any wall course.
           Another wall having an average width of 0.40 m, East-West oriented , conventionally labeled Z4, appears in cassette C22 (Fig. 5). The joints of this wall could not be identified mainly because of the aspect of such structures, i.e. walls which are hardly preserved at all and their foundations generally appear at the same level as the ruin. Moreover, is seems that the area has been "disturbed" in the '80, during certain works performed on site. Anyway, in the bordering cassettes there is a lot of the stone that was initially part of those walls.
           However, we can clearly see walls on the northern area of the plateau, i.e. cassettes C10 and C15 - C19. These walls (Z1 and Z2) can be distinguished over c. 12-15 m in length, extending almost in parallel, at an average distance of about 2.10-2.80 m (Figs. 5, 7 and 11-12). It is unusual that the wall Z1 does not form a straight angle in cassette C15, but it extends in a rounded angle to cassette C10 (where it can no longer be distinguished because of the relatively compact debris). The landslides that have occurred in the area account for the trajectory taken by Z1. The building discovered in Valea Nanului [6] makes a similar case. Such walls could be traced almost up to the brink of the plateau without distinguishing their northern extremities. West of these walls, in cassette C13, we can see the trace of a sustaining rod, under the shape of a blaze, about 0.10-0.15 m in diameter, surrounded by a small ring of quarry rock. Yet, another rod having a different aspect appears in section S4, going down 0.70-0.80 m deep. In the latter case there is no blaze print and no ring of stone, but only its lodging hole in the ground, leading us to believe that the two rods are not contemporary.
           Another possible wall appears in cassette C28, in the very vicinity of the area disturbed by the works carried out in the previous years. This is why we can only follow it on a 4 m span (Figs. 5, 7).
           In cassette C14 were discovered 18 stone objects (two intact votive inscriptions and three fragmented ones, while the rest are fragments from similar monuments, as well as 3 plinths (Figs. 7, 19/1-6, 20/1-4, 21/1-3, 22/1-3, 23/1, 3, 24/4-6). In the North-West corner of cassette C14, under the debris which extends over the whole western half of the cassette, was identified a print mark under the appearance of firing pigments; the print was labeled X; it is more compact in the northern side and has a relatively oval shape (diameter 2.20 x 1 m), but its southern limit is not very clear and its short western side extends behind the cassette profile, which makes it impossible for us to follow it (Fig. 5). This print is 0.10-0.15 m thick, and its profile is slightly alveolar.
           The stratigraphy (under the current conditions of the terrain configuration, after numerous disturbances) in the North-South (Fig. 8) and East-West direction only indicates one level of inhabitancy.
           Archaeological research has revealed the existence of several buildings on this plateau.
           The first one, of an apsidal shape (Figs. 6 and 10), is present in the Roman province of Dacia, as well as in the Empire [6a]. In Dacia, we have similar buildings at Ulpia Traiana Sarmizegetusa, Porolissum and Germisara [7]. They were considered temples [8] or, more recently, even the location of collegia [9]. Both such functions could be possible in the case of the building on the Hăbad hill, if we take into account the inscriptions discovered on site.
           The second building is enclosed by three walls (Fig. 7). We believe it is a classic temple. The edifice is rectangular, with the long side North-South oriented. The perimetral wall probably enclosed a yard. The epigraphic altars (found in situ) inside the temple yard are oriented to the west, according to the canons (Vitruvius IV, 5, 2). Similar plans for temples can be found at Roşia Montană [10], as well as in other centers of the Dacia province [11].
           As for the other three hypothetical buildings (of which only one wall of each was preserved at the North and West extremities of the plateau) we cannot draw any conclusion yet. It is not impossible that such buildings would belong to consular beneficiaries attested in the Hăbad inscriptions. Most certainly, a statio led by beneficiaries consularis [12] used to be on this site.
           On the same plateau, the archaeologists from Deva discovered other three buildings [13].
           With respect to the construction techniques of such buildings, we can say that they had quarry-stone foundations on which probably wooden structured would be erected. The roof was made of wood, as well. The small fragments of burnt wood and the iron nails which were found on this site, as well as the absence of any mortar traces and shingles from the entire investigated area plead for this theory.
           Such edifices cannot be dated exactly, yet the thin layer of material culture indicates a short-term inhabitance. Most of the epigraphic discoveries seem to date to the 2^nd century (see below), therefore the above-mentioned buildings could have the same chronology.
           The movable inventory consists of Roman pottery, objects made of iron and stone.
           The pottery is of the provincial Roman type, almost entirely decomposed because of humidity and acidity of the soil. From the analysis of the ceramic material we identified three types of vases: vertical rim bowls (Figs. 14/1-2), turibulum (Fig. 14/3), plates with an almost horizontal rim and interior grooved decoration (Figs. 14/5-8), bowls with almost horizontal rim or slightly slanting to the inside, oblique wall and flat bottom (Figs. 15/1-8), pots (Figs. 16/1-5 and 17/1-6). Such fragments of vases were found over a large area, they actually occupy the entire investigated zone. Chronologically speaking, we can only determine a longer time-span, between the 2^nd-3^rd centuries A.D. The iron materials consist of nails of different sizes (Figs. 18/1-8, 10-17) and a hook (Fig. 18/9).
           Stone objects are more numerous, yet most of them are fragmentary because of their partial re-use as well as disturbances of the area. Among such items we could find: four upper corners of an altar crown ornamented with small corner ornaments and very stylized vegetal motifs (Figs. 19/1-4), which seem to have been parts of the same monument; a fragment of volute from an altar-gable (Fig. 19/5); eight items from different monuments with traces of mouldings (Figs. 19/6, 20, 21); three plinths (Fig. 22). All discovered objects were made of tuff gritstone of Orlea or Carpeni.