II. Lower terrace – a. Architectural remains and structures
On the lower terrace of the Wadi al-Farasa East two trenches (trenches 2 and 3 on fig. 1) were opened in order to verify the hypothesis of a peristyle courtyard between the two main monuments of the terrace as described by the Deutsch-Türkische Denkmalschutzkommando in 1916 Bachmann – Watzinger – Wiegand 1921: 75ff. , or in more general terms to locate the original floor level.
Fig. 1: General plan of Wadi al-Farasa East with lower and upper terrace, after Bachmann et al. 1921, digitalised and modified by S. Fachard, Lausanne
In both soundings the ancient floor level was reached at about 930.60 meters above sea level and the hypothesis of a courtyard surrounded on three sides by stoai was confirmed, at least for the northern side of the area (figs. 2–4).
The original floor was paved with huge rectangular stone slabs put on a foundation of smaller and irregular slabs founded on a very hard layer of clay-containing earth The so called samaga that was used in the village of Wadi Mousa until recent times, see Kolb 2000: 258 n. 467; Zayadine 1974: 135. .
Fig. 2: Trench 2 from NW showing floor slabs and column drums (photo: S. G. Schmid)Fig. 3: Trench 2 from SW showing floor slabs and column drums (photo: S. G. Schmid)
In sounding 3 in the east corner of the courtyard, the floor slabs were put directly on bedrock which had been cut away in the entire area in in preparation for the installations of the lower terrace (Fig. 5).
Fig. 4: Trench 3 from NW showing bedrock, foundations and column drums (photo: S. G. Schmid)Fig. 5: Trench 3 from SE showing bedrock, foundations and column drums (photo: S. G. Schmid)
Sounding 2, however, revealed that at this spot the original bedrock was not cut away by the builders; here, the floor slabs were embedded on a layer of huge and irregular stones that were carefully filled in with the above described foundation matrix of smaller slabs and clay-containing earth (Fig. 6). Apparently at the position of sounding 2 the bedrock is on a lower level than at sounding 3. This is because at sounding 2, the valley lies at its lowest where the water had eroded the original wadi bed long before building activities began in this area. In order to create a even surface for the installations of the lower terrace, the original wadi bed had to be back-filled with huge blocks of stone at the lowest point while in the eastern part the bedrock had to be cut away to match the walking level. In trench 3 the level of the bedrock is at 930.58–66 m above sea level, while in trench 2 the level of the back-fill of the wadi bed is at 930.31–45 m above sea level. As can be seen on fig. 7, the floor subsided in some parts of sounding 2 because with the collapse of the Nabataean hydraulic systems (see below) the water gained back its old way and washed out the artificial fill and the foundation of the floor level.
Fig. 6: Trench 2 from NW showing small sounding beneath the level of the floor slabs (photo: S. G. Schmid)Fig. 7: Floor slabs in trench 2 (photo: S. G. Schmid)
As can be seen on fig. 7, the floor subsided in some parts of sounding 2 because with the collapse of the Nabataean hydraulic systems (see below) the water gained back its old way and washed out the artificial fill and the foundation of the floor level.
In both soundings it was found that special care had been applied to the area of the columns that formed once the north stoa of the lower terrace. In trench 3 the rock was cut some seven centimetres deeper and was then carefully refilled with the above described foundation. In trench 2, the foundation in the area of the columns is also very carefully executed and slightly higher than in the surrounding areas. The columns were not standing on bases but on a kind of plinths of which two were found in situ in trench 2 and a third in trench 3. The level of the plinths is at 930.73 (trench 3), 930.72 and 930.71 (trench 2) m above sea level, respectively. The three plinths are about 20 cm high and were found either with a column drum still in situ or still showed the raised circular surface on which the columns once stood (figs. 2–6).
The first column of the north stoa is a half column abutting the rock cut surface which had already been posited by Bachmann et al. 1921: 75ff. Drums and the capital of the half column were found in trench 3 which have provided the missing elements to explain the carved out rock features (Fig. 8). In addition, their place is confirmed by the corresponding measurements of the carved out space on the one hand and the dimensions of the half column and the half capital on the other. In two cases, the rock carvings show a lateral enlargement.
Fig. 8: Rock carvings for half column in trench 3 (photo: S. G. Schmid)
Into these enlargements perfectly fit analogous enlargements of two half columns drums of which one was actually found (Fig. 9). According to the dimensions of the enlargement this must be the higher of these half drums because the carving of the lower one is too narrow.
The architectural remains found in 2000 and the measurements of the carved rock for placing the first half column, make it now possible to reconstruct the basic layout of the north stoa on the lower terrace of the Wadi al-Farasa East. The diameter of the columns is 60 cm, the distance between the columns (intercolumnium) is about 187 cm with the bay measuring 247 cm (The better preserved column drums showed careful carving that points to the application of stucco. If we consider a layer of approx. 1 cm of stucco around the columns this would give an original diameter of 62 cm and therefore an intercolumnium of 185 cm; however the measurements of the bay would still be 247 cm.). This means a slight modification of the reconstruction proposed by Bachmann/Watzinger/Wiegand who proposed smaller columns and intercolumnia of about 160 cm. As referred to above, the columns have no bases but are directly put on plinths. The height of the columns up to the capitals can be calculated by comparing the plinths (930.71–73 m above sea level) with the rock carvings for the half capital (lower level: 934.74–79 m, top level: 935.04 m above sea level): this comes to a height of ca 4 m without capitals and with capitals of about 4.27–4.30 m. The half capital from trench 3 measures 27 cm in height, the complete capital from trench 2 is 29 cm (The difference of 2 cm within the two capitals would perfectly fit the difference of 2 cm within the plinths between the two trenches [cf. above]).
Fig. 9: Half column drum from trench 3 (photo: S. G. Schmid)
All column drums and the two capitals were badly damaged by the wind when still standing and further worn due to the seasonal alluvia when the stoa had collapsed and they were laying on the ground. Their bad state of conservation makes it difficult to say much about their form and type. However, they appear to be plain Doric capitals and the stoa must have looked similar to the porticoes of tomb no. 4 (the so called Khan) at the Bab es-Siq Brünnow and Domaszewski 1904: 195f. no. 4. showing Doric capitals and no bases, although there the columns are slightly bigger in diameter. This is an other modification of the reconstruction as proposed by the German scholars who thought, according to the size of the rock cuttings they correctly identified with the capital, of as Corinthian capitals Bachmann – Watzinger – Wiegand 1921. . However, as became clear from the half capital found in 2000, the hewn slightly larger space was to make the back part of the capital fit the carved out socket. Although these chances seem very small at first seight, they lead to an overall modified plan of the entire installation (Fig. 1).
The larger Doric columns instead of the supposed Corinthian and their wider intercolumnia result in the removal of three free-standing columns on the plan, that is there are now 12 Doric instead of 15 Corinthian columns Compare the traditional plan of the area: Schmid 2000: 336 fig. 1. .
As mentioned earlier, the level of the original paved floor was 930.61–66 m above sea level. The floor level of the so called „Soldier Tomb“ is on 932.27 m above sea level and therefore about 1.65 m higher than the courtyard. The interior of the opposite triclinium is 931.58 m above sea level, that is about 0.95 m higher than the courtyard. This means the courtyard was accessible from the two rock-cut monuments by steps no longer visible, but in all likelihood still in place.
In trench 2 some pavers of the original floor level were found in situ, but exclusively in the stoa (figs. 2. 3). These slabs were about 40 cm wide with different lengths sometimes exceeding 80 cm and were roughly 10 cm thick. In the area of the courtyard no flag stones were found in situ but some fragments from the destruction debris suggest that they were the same size if not bigger. The fact that the original floor slabs were completely missing in trench 3 and to a big extend also in trench 2 suggests that part of the lower terrace was robbed out, but only after maintenance had already ceased for some time; this can be gleaned from the stratified sections in the two trenches.
