In the area of the NE-corner of the complex, corresponding to the last two 5m² squares within the N-portico, work started already in 2006. However, during the campaigns of 2006 and 2007 rather important constructions of the medieval period were excavated on that spot Schmid 2007B; Schmid in press. . Surprisingly, the medieval structures started to appear on a rather high level at first sight and therefore were connected to the rocky outcrop on the NE-corner of the complex. The NE-corner used to be the base for the second floor of the entrance building in the Nabataean period. During our 2009 season, the explanation for these observations was found, together with the NE-angle of the courtyard, an event we were looking for since our first season in 2000. Quite an important number of information regarding the construction and structure of the Nabataean complex was ever since expected from that area.
The first remains of collapsed architecture first appeared on a level considerably higher than on previous occasions (fig. 2). It soon became clear that it was partly due to the fact that the angle formed by the northern and eastern portico of the complex was built in shape of a massive heart shaped column (centre right on fig. 2). When the freestanding architecture of the complex collapsed during the earthquake of AD 363, the corner as well as the surrounding „normal“ columns fell into the space formed between the row of columns and the rock originally being the back wall of the portico and the support for the upper floor of the complex. On several occasions we could observe that the main direction into which columns and other architectural remains fell seems to be from East to West (cf. also below on the S-portico). Since the column drums, capitals etc. from the angle were literally „trapped“ in the corner of the complex, the collapsed remains are conserved at a considerably higher level than elsewhere what, in turn, also explains the higher level of the medieval constructions (cf. above).
Heart-shaped angular columns are quite well known from the region, especially from Herodian architecture, but also from some Nabataean examples. Heart-shaped columns in angles of peristyles can already be found in the palace of Demetrias (Thessaly. Greece) in a building phase that is dated around 200 BC. Marzolff 1996: especially 158–160; for a concise treaty of heart-shaped columns see Coulton 1966: especially 137–141. . From then on this particular detail spreads out rather widely, as may be underlined by analogous structures from the „Palazzo delle colonne“ in Ptolemais in the Cyrenaica in modern Lybia Pesce 1950: pls. I. XI. . It is generally believed that this element was taken over from Asia Minor and was especially successful in Ptolemaic architecture Lauter 1986: 255. and it would be from there on that it found its way into Judaea, where it was especially prominent within buildings attributed to Herod the Great Lichtenberger 1999; Schmid 2002B. . It is, therefore, not surprising that it is attested on other monuments, surely not related to Herod, such as the so-called complex of Absalom and Josaphat, the monolith of Zacharias and the „pillar“ of Absalom, all funeral monuments at Jerusalem on these and similar monuments see Bonato 1999. . The nearest, although not necessarily the closest parallel to the heart-shaped column from the Soldier Tomb complex can be found in front of the so-called Turkmaniya-Tomb, where a similar courtyard can be reconstructed on this see Schmid 2007A. . However, the angular column of the Turkmaniya-Tomb is considerably smaller than the newly discovered one from the Soldier Tomb complex. The basis form of our heart-shaped column is rectangular, measuring 60cm x 60cm, with an additional half-column of 30cm radius added on two sides.
One of the main problems with the reconstruction of the Soldier Tomb complex in previous years used to be precisely this angle between the northern and eastern porticoes of the courtyard. This is due to the initial reconstruction of the complex as proposed by the Deutsch-Türkische Denkmalschutzkommando in 1921 Bachmann and Watzinger and Wiegand 1921. . There, the eastern portico does not form 90° angle with its lateral counterparts from the northern and southern sides, but follows an oblique track (gray on fig. 1). Since the discovery in 2004 that the porticoes initially were covered by arches Schmid 2005. , this reconstruction became problematic, since the pressure of the arches would have caused major static problems when hitting the angular column in any other than a right angle. Suspicions that the initial plan of the Denkmalschutzkommando probably needed verification and modification were confirmed, when it became clear that the heart-shaped column from the angle of the N- and E-porticos indicated a perfectly right angle, as does the stylobat upon which it is built (fig. 3). Since the Denkmalschutzkommando did not carry out excavations, the plan published in 1921 had to work with the appearing rock, which of course did not form many right angles. Therefore, the German scholars proposed an oblique E-portico (gray on fig. 1). However, the sounding in the very corner of the complex showed that the Nabataean builders had constructed a substantial wall in front of the rock in order to outbalance the irregularities of the rock (on top of fig. 2). This wall stretched also eastwards in front of the huge triclinium BD 235, at least until the main entrance, from where on the rock is very regular and the built wall probably was not needed anymore.
When measuring these new elements as well as those parts of the rock that must have belonged to the initial construction with a total station and putting them into a new plan, it became apparent that the Denkmalschutzkommando must have simplified some elements of the initial plan in order to obtain what they thought to be a coherent picture of the area. The corrected preliminary version of 2009 (in red on fig. 1) shows that, as a matter of fact, the East-portico is in an almost perfect right angle compared to the North- and South-porticoes. On the contrary, the huge triclinium BD 235 is not, as proposed by the Denkmalschutzkommando, cut out of the rock in a right angle to the courtyard, but considerably turned out of the main axis of the complex.
Within the substantial wall, a small water basin measuring about 1m² was incorporated (top centre on fig. 2). This water basin – or a similar structure – was already suspected since 2000 when some initial observations and thoughts about the water management system of the Soldier Tomb complex were realised Schmid 2001; on the water management of Wadi Farasa in a wider frame see Schmid 2008. . Since the rock carved remains of what must have been a clay pipe pasted with mortar onto the still visible rock carvings indicated the existence of a water installation on that spot. The discovery of the basin was thus not really a surprise. According to the finds of several ashlars with integrated space for a water pipe, the last meters of that water channel leading vertically into the basin were built upon the substantial wall in front of the rock mentioned above. The basin itself was plastered with the typical Nabataean hydraulic mortar using small pieces of pottery in order to improve its hydraulic qualities. On the bottom inside the basin as well as in front of it rather substantial amounts of pottery and glass were found. As on other occasions, for instance in loci related to the small rock-cut triclinium on the western rock-plateau of the complex on this see Schmid in press. , the total absence of fine ware pottery is striking. Cooking pots, amphorae and jugs of coarse ware pottery were by far the most prominent finds. An explanation of this phenomenon can probably be found in the chronology of these finds. As indicated by the shapes of the pottery and by the important number of so-called Petraean-Early-Byzantine lamps Grawehr 2006: 340–349. , these loci must belong to the latest period of use of the complex, shortly before the devastating earthquake of AD 363. This chronology and the rather important presence of glass fragments could indicate that during the late Roman period the most important category of fine ware, that is drinking cups, were replaced by glass cups and beakers on this see some reflections in Keller 2006: 127–130. 176–179. .
The water basin is so far the first clearly recognisable point of access to water within the entire complex of the Soldier Tomb and is immediately next to the rock-cut room, which was partially excavated in 2006. Since the only structural remains within the excavated part of that rock-cut room was a taboun, we proposed to interpret it as a kitchen Schmid 2007B: 143–144. . The pottery types found around the taboun correspond to the ones mentioned above and coming from the water basin. Therefore, the taboun also belongs to the late Roman period. Although there were no indications as for the use of the rock-cut room during Nabataean times, its position next to the water basin on the one hand and the neighbourhood of the huge triclinium could indicate a similar function in order to prepare food and drinks for the people feasting in the banqueting hall BD 235, an activity that surely would justify the presence of a comfortable water supply.
In the area of the N-corner, the floor slabs were posed on a small bed of samaga directly upon the carefully cut rock, as indicated by fine lines on the rock. The lines obviously correspond to the dimensions of the different rows of slabs (figs. 4. 5). As observed on several previous occasions, most of the slabs already were missing when the architecture of the complex collapsed in AD 363. During this year´s campaign some further elements came to light that can sharpen the picture of the last years of our complex. From the rock-cut room a layer of fat, gray ashes is spreading out into the N-portico, surely corresponding to the successive cleaning of the taboun’s ashes. Where the floor slabs still were in situ, the ashy layer is directly posited on the slabs. However, where the slabs already were missing, the ashes lay directly on the rock. The collapsed architectural members fell upon the ash as well as a next layer composed of sandy earth (fig. 6). In other words, despite partial plundering of the complex, it was still in use – at least partially –, since the taboun obviously was still functioning.