II. Lower terrace – b. Stratigraphy and chronology
The stratigraphy in soundings 2 and 3 allows us to draw a fairly accurate picture of the relative chronology of the installations on the lower terrace of the Wadi al-Farasa East. Together with some small finds these stratified contexts can provide information on the absolute chronology of the monuments.
As referred to above, most of the original floor slabs were no longer in situ when the north stoa collapsed. This means that the monument was still standing upright but apparently no longer cared for. The same can be deduced from the fact that some of the column drums were found directly on the foundations of the paved floor while others had collapsed on 20–30 cm layers of sand above the floor level. Therefore, some of the columns must still have been standing for some time after the complex had been abandoned.
Especially in trench 3 the destruction layer above the floor level, containing ashes and many stones including broken architectural blocks, was clearly visible. The pottery from that layer contained some forms of coarse ware that correspond very well to the pottery found in the destruction layers of the houses on az-Zantur which can be associated with the earthquakes of 363 and 419 AD On similar pottery see Fellmann Brogli 1996: 242ff. figs. 730–732. 736–741. 760–763. 772–777. 793–795; on the houses and their destruction see Kolb 2000. . It is plausible to consider one of these earthquakes as the cause of destruction of the courtyard and its stoai on the lower terrace.
Above the destruction debris is a sequence of layers showing clear evidence of alluvian soil. The texture of these layers is very compact and layers of earth with clay alternate with more sandy ones. These layers do not contain any stones. This could be an indication that most of the Nabataean hydraulic system in the Wadi al-Farasa East was not functioning anymore which caused the build-up of alluvia. The retaining walls would still have prevented bigger pieces like boulders to be carried in flash floods of heavy seasonal rainfalls. The archaeological contexts corresponding to these alluvia were heterogeneous and no precise date can be attributed to them, except that they occured after the collapse of the stoa in late antiquity. A second unit of sandy layers was much thicker in trench 3 than in trench 2 and again contained no datable material.
Although this late occupation does not belong to the main objective of our project, it provides interesting information on medieval Petra. It has been suggested that on top of Jabal al-Madhbah (Zibb ’Atuf) there was an other crusader fortress Vannini and Vanni Desideri 1995: 512. although no detailed investigations has so far been done. This suggestion had already been put foreward in early literature See for example Brünnow 1909: 250; Dalman 1912: 12–14. and if it is found to be true – as it seems to be – then the Wadi al-Farasa East would have been one of two important access routes to this fortress and received special attention by the crusaders. The lower terrace with its huge retaining wall towards the northern part and the narrow steps leading to the upper terrace would have formed an appropriate place to install a guard post in order to control the way up to the castle. It is probably in this light that we have to view the crusader occupation evidenced in trench 3. These findings also put in a wider context a stone slab, maybe a tomb stone, with a carved cross that was found in the „Garden Triclinium“ and most probably belongs to the same period Brünnow – Domaszewski 1904: 275 fig. 307; Dalman 1908: 196 fig. 117; Brünnow 1909: 249f.; Lindner 1997: 104 with n. 10. .
Fig. 10: Hand made pottery from medieval occupation in trench 3 (photo: S. G. Schmid)
In trench 3, the alluvial fill is topped by a level of occupation 932.70 m above sea level which can be clearly distinguished from the underlying structures and layers. This occupation includes a small kiln or oven built from small stones into the NW corner of the stoa. The pottery from this occupation corresponds very well to the selection of pottery from the crusader fortress of al-Wu’ayra on the outskirts of Petra dating to the 12th century AD Vannini and Vanni Desideri 1995. . In our occupation the characteristic plain hand-made pottery was found (Fig. 10) as well as some painted coarse ware pottery and a few sherds of a fine but hand made red painted ware (Fig. 11) which may have been imported (see comparable pottery in Vannini and Vanni Desideri 1995: 531ff. figs. 16–20. . Similar pottery is apparently common in the Carmel and in Ceasarea from later 13th century AD contexts Pringle 1984: 95f. especially fig. 3, 11; Pringle 1985: 175f. especially fig. 2, 2. .
On top of both trenches was an ashy layer representing the latest event that could be assigned a precise context within the history of the Wadi al-Farasa East: These layers together with some dumped stones are almost certainly from cleaning the two rock cut monuments, that is the triclinium and the „Soldier Tomb“ carried out by the Department of Antiquities of Transjordan in the 1930s Horsfield 1938: 40 with nn. 5. 7. .
Fig. 11: Red painted pottery from medieval occupation in trench 3 (photo: S. G. Schmid)Fig. 12: Nabataean plain and painted pottery from soundings beneath floor slabs in trenches 2 and 3 (photo: S. G. Schmid)
As there is no earlier or later material from this layer we may conclude that the north stoa was built in the second or third quarter of the 1st century AD. This is especially interesting because the construction of the huge cistern of the upper terrace was dated to the same period by painted pottery found in the mortar of the basin in front of it Schmid 2000: 343f. . It would appear that most of the monuments of the two terraces belong to the same building programme. As the stoai of the courtyard of the lower terrace cannot be earlier than the rock-cut installations but are most probably contemporary or – although less probable – slightly later, this means that the same date can also be applied to the triclinium and to the „Soldier Tomb“. This is also supported by the cuirass of the central statue of the tomb that can be dated stylistically to around the middle of the 1st century AD as had already be pointed out by M. Lyttelton Lyttelton 1974: 62. .
Therefore, the term „Roman“ very often applied to the „Soldier Tomb“ has to be abandoned. There is no reason to date this tomb as well as the related installations from the lower terrace later than the third quarter of the 1st century AD and therefore to the Nabataean period.
However, the most important result in terms of chronology is related to the construction of the north stoa. In both trenches opened in 2000 the very hard samaga layer that was used as part of the foundations for the floor slabs contained some Nabataean pottery providing a rather precise terminus for the construction of the monument (The very hard texture of this samaga layer sealed it against intrusions from above. This is also underlined by the homogeneous aspect of the pottery found in it.) Beside some small rim sherds of plain Nabataean fine ware there were also a few tiny fragments of painted ware (Fig. 12), all belonging to phase 3 and more precisely to phase 3a of Nabataean pottery, dating from AD 20 to AD 70/80 Schmid 2000A: 25. 38. .
