Preliminary Report on the 1999 Swiss-Liechtenstein excavations at ez Zantur
by Bernhard Kolb and Daniel Keller
II. Ez Zantur IV: The Nabataean mansion
The excavation on EZ IV focused on the western, central and southern parts of the terrace. 15 grid squares (PQ) of 5 x 5 m were investigated (figs. 1–2).
Fig. 2: EZ IV. Schematic plan of the Nabataean mansion (drawing: B. Kolb)
Squares 89–92/AP established the association of the representational rooms 6, 7 and 17 with the central structures east of room 1. The schematic plan shows that the building was laid out along a main axis running NS, consisting of courtyards 28, 15, 19 and exedra 7 which connect to the perpendicularly set banqueting halls 6 and 17. Despite the insertion of two narrow rooms 20 and 21, we can assume that the striking mise en scène of the stretch between the supposed main entrance to the north and the reception/banqueting halls at the southern limit was designed to impress the visitor with a grand reception. The alignment of rooms 20, 21 and 26 concealed a step in the bedrock surface of ca. 1 m height along the walls I, AL, AK and AJ (room 21: ca. 924.07 m; courtyard 28: ca. 925.05 m). Clearly set off from the stately part of the mansion are the private and servants’ rooms to the east and west of the central courtyards. Focal point of the western private wing was the small courtyard 5. The strict layout of rooms 30, 27 and 22 to the north of courtyard 5 repeats the arrangement of rooms 6, 7, 17 on reduced scale and prove that – contrary to the Nabataean dwelling on EZ I Stucky 1996: 46. –, the same architectural conception was at work in the realization of private and representational rooms. Worth mentioning is the discovery of wells underneath rooms 22 and 27. They present vital evidence in reconstructing the water resources of the building.
Rooms 19, 15, and 21 (PQ 89/AO–AP; 90/AM–AN)
The clearance of squares 89/AO–AP exposed courtyard 19 almost completely (Figs. 1–3). The area measures 7.7 m (NS) by 7.2 m (EW). The northern and southern sides are furnished with distyloi in antis. The preserved bottom column drums (Diam. 75 cm) do not rest on ordinary bases but are fitted with „base collars“ of the Ionic-Attic type (Fig. 3). The two separately carved elements were mortared to the columns and an additional layer of high-quality plaster covered the joints. The same technique is well attested in the Temple of the Winged Lions Hammond 1996, 44f., Pl. 13, Figs. 2–3. . Courtyard 19 and exedra 7 were, as already mentioned elsewhere Kolb – Gorgerat – Grawehr 1999: 264. , connected by doors in the intercolumnar spaces. The rendering of the intercolumnar spaces on the north side of courtyard 19 is yet unclear. While the bases attest to openings of ca. 45 cm, identical to those on the opposite side in the south, there are no indications of doors. Moreover, the stylobate does not show the slightest trace of wear which would point to a secondary function as threshold. Considering the spacing of the bases on the EW axis, a curtain wall between the columns, which was dismantled later on, may have been a possible alternative to doors. Such a reconstruction, however, would imply a clear separation of courtyards 19 and 15. Clearance of PQ 89/AN is expected to resolve present uncertainties.
In the earthquake of 363, the collapsing architectural members damaged the sandstone flagging mainly in the central and north-western sections of the pavement. Between the door to corridor 3 and the eastern column of the north side, a repair of approximately 1.2 x 2.2 m is discernible. The inserted limestone flags are longer than the original ones and in a much better state than the original ones. The findings in courtyard 19 are typical for the rather simple stratigraphy on EZ IV: A considerable deposit of earthquake débris covers a ca. 20–30 cm thick layer consisting of fragmented stucco decoration from the walls and columns. The latter context simultaneously seals the stratum of the last phase of occupation.
Fig. 3: EZ IV. Courtyard 19 seen to the south (photo: D. Keller)
In the western part of courtyard 15 the mortar bedding of the robbed pavement was exposed. The badly cracked remains of the bedding were in a rather sad condition. Nonetheless, they are of some significance as it was on their level that a hoard of bullae, which originally sealed papyri of the second and third centuries was found in 1998 Kolb – Gorgerat – Grawehr 1999: 269f. . A small sounding made clear that the subfloor lies on a sterile fill of 20–30 cm thickness consisting of stone chips and sand. The fill levelled the uneven surface of the bedrock. In square 90/AM the northern wall AL of courtyard 15 and two fragmentary pilasters were exposed. The pilasters divided the southern face of wall AL into three segments and were aligned with the columns of courtyard 19/exedra 7, highlighting the strict overall conception of the building. At the eastern end of wall AL a well preserved fragment of the original painted decoration is still in situ. Fragments of a red dado and a tripartite border are discernible. The latter consists of a central black band and a linear pattern which frames a panel of uncertain colour.
Walls AP and AC facing courtyard 19 were decorated with different schemes. Closer inspection of the exposed fragments suggests that one frequent scheme was based on white panels with red drafts. A further well represented design included white hexagons with central flowers delineated by red bands.
Dozens of well preserved fragments of Corinthian capitals of heterodox or floral type with rich scrollwork on the calathos were found in the destruction layers of the courtyards 15 and 19. These capitals originally crowned the huge columns compare Kolb – Gorgerat – Grawehr 1999: 265, Figs. 5–6. . The main colour of the capitals was deep blue as traces on many fragments indicate. Rather surprising were the finds of capital fragments obviously belonging to the Nabataean blocked out type in the northern parts of courtyard 19 and in courtyard 15. Further fragments were exposed in the neighbouring rooms 20, 21, 28, 25, 16 and 2. A series of 8 plug-shaped objects made of sandstone is of special interest. All eight items have the same cross section which is reminiscent of the hull of a ship (Fig. 4).
Fig. 4: EZ IV. Schematic abacus flower of Nabataean blocked out capital (photo: O. Jäggi)
All these separately carved items of slightly varying heights (7–10 cm) and widths (8–10 cm) share the peculiar feature of a smoothed and painted surface on their front ends and on 7–10 cm of their sides which ends in curving scratched line. A painted linear black decoration and remains of gilding in the intermediate zones as well as traces of mortar at the backside are preserved on one of the plugs (Fig. 4). Most probably we are dealing with schematic abacus flowers typical for the Nabataean blocked out capitals McKenzie 1990: 190, Figs. h–j. . It is obvious that the plugs/abacus flowers had been fitted into the capitals up to the curved scratch line. Although of no static function, they measure up to 25 cm in length, i. e. they were sunk up to 15 cm into the capitals. The technique of inserting separately carved abacus flowers into the capital is not limited to the blocked out capitals. The pine cones found by the American team in the so-called Great Temple at Petra prove that the same technique was used for Corinthian capitals of the heterodox type Sharp Joukowsky 1998: 228f. and Fig. 5.49. . Considering the quantity of the schematic abacus flowers found in the courtyards 15 and 19, the question arises as to which supports the blocked out capitals belonged. We propose to reconstruct the Corinthian capitals on the columns (courtyard 19) and the blocked out capitals on the antae and on the pilasters on the northern wall AL of courtyard 15 respectively. Furthermore, the remains of paint and gilding on some of the best preserved schematic abacus flowers give an indication of the former appearance of the capitals. We can now confidently rule out the possibility of added moulded stucco decoration see Lyttleton-Blagg 1990: 95). .
Fig. 5: EZ IV. Rooms 21, 20, 26 (left) and rooms 23, 22, 28 from the east (photo: D. Keller)
The only partially exposed room 21 is situated at the northern limit of the southern and central rooms on identical floor levels (Fig. 5). Contrary to all rooms excavated so far, room 21 was furnished with a plaster floor which is merely preserved along wall AK. Locus 3231 included the whole exposed floor space, except the areas with the plaster still in situ. The levels of locus 3231 range from 924.05 m to 923.95 m, corresponding therefore approximately with the level of the plaster floor (924.07 m). The chronologically mixed finds from the mentioned locus represent the time span from the first to the fourth century – as was to be expected. The chronological information given by the material from locus 3251 beneath the aforementioned are of greater value: Glass, lamps and pottery date homogeneously to the first century AD. The data relevant to room 21 are closely comparable to the findings in rooms 20 and 26 (infra) confirming the preliminary dating of the initial construction phase to the first century AD. It is hoped that a sounding in room 21, planned for the forthcoming excavation season, will produce even more precisely datable evidence concerning the history of the edifice.
