| Entry: | Building 105 starts to get very complicated. CMF and FKJ continued excavation in its northern part. CMF found a layer of fill in the NW corner that was very different from its surrounding (darker, looser, with very many finds) and that she determined to be a small pit (17274). FKJ in the meanwhile started a new layer of room fill (17272) from the NE corner where she found many large lumps of grey ashy brick, orange clay, and plaster in a dense cluster. It took a long time to determine that none of these belonged to a construction feature. The lack of any other kinds of finds or material between these large clay lumps made them seem like collapse, but most of the materials cannot be found in the walls surrounding B.106, therefore it cannot be in situ collapse – unless the non-preserved upper parts of the building contained such materials.
In the southern part, the walls challenge us. EUR, SO and I spent the day scraping around them to determine their outlines. A single brick was found in front of (north of) buttress F.3363 abutting it. A cluster of clay and plaster in the corner of F.3363 and F.3364 turned out not to belong to a construction. Instead, we found a little feature of the size of roughly one brick put into the corner of the two walls west of the buttress, which we have been mostly ignoring to far as they seem not to fit with the rest of the building. They are made of yellowish brown brick and mortar, with yellow plaster. At the moment we have only a corner of two walls where we would assume the SW corner of B.105 – but we would have expected this corner also to be made from the same dark grey brick as the rest of the building. This corner, including the newly discovered feature in the corner do not have feature numbers yet as they cannot yet be connected to F.3364 and F.3352. It is still possible that they connect to these walls and just, for some reason, this corner was built from different materials. It is also possible that F.3364 and F.3352 were cut for the yellow walls to be built, though. We will see. Not very helpful is the fact that, while the yellow walls stand proud and clear, the grey walls, which normally are well-behaved walls, are shy and hide their faces behind fill that stresses us out because it has so much of (eroded?) grey wall material in that it takes time to separate it from the wall.
KTX started removing the upper part of F.5058, including its corner with F.3351. This part of the walls of B.106 is the one that is preserved to the greatest height, but it is still disturbed at the top. We tried to separate the disturbed material from the wall material but when finding a tile in the sieve bag noticed that it did not really work out. In order not to spoil the mud brick unit of F.5058, which could contain finds crucial for the dating of the construction or the determination of the mud source, we decided to call the entire wall material removed to day “brick contaminated with late disturbances” (16873) and open a new number for the bricks removed from now on (17276) which is not contaminated.
GWN removed the lowermost part of buttress F.5053/5054 together with the fill around it, a mixed unit that was still quite large in terms of volume. By the end of the day, we had a plan view of the lower part of the buttress, which, just like the other three in B.98, is a an older version that is smaller than the part we saw so far, surrounded by plaster.
JHB started removing the small wall F.3335 and during this process made several interesting observations concerning the stratigraphy of the construction features. Firstly, he found a plaster layer on the northern face of F.3335, where it abuts F.3320. The fact that the plaster layer goes further down than the base of F.3320, which is actually preserved only a few centimetres high, indicates that the plaster was coating F.3335 before F.3320 was built abutting it. The north of F.3335 therefore seems to have been facing an inside space.
JHB also noticed that we had seen the outlines of F.3335 wrong before. We had seen that F.5052 forms a buttress with normal rectangular outlines abutting wall F.2428, and that F.3335 was built abutting its eastern face. Now, this is not true. F.3335 was built first, alongside the eastern part of F.2428. F.5052 was then built abutting both, and therefore has an uneven back part (in the north). That changes everything about the northeast corner of the building we have been thinking so far. We always assumed that F.3335, as well as F.3320, F.3321 and F.3334 were later additions to the building, postdating the long outer walls and the buttresses. One of the reasons for thinking this was that we discovered early that F.3335 is sitting on fill, while it took us a bit longer to find that out in the rest of the building. Another was that we kind of assumed the original building had the “typical” form with four corners and four buttresses, and that the small room Space 446 was added to it later. Now it seems that Sp.446 was part of the building from the very start, even before the buttresses were added to it. One still wonders why they did not just continue F.2428 to the doorway with Space 446, but instead built a small wall feature slightly offset; maybe we’ll find something enlightening us about this issue later.
In the SW corner of Building 107, DLG uncovered a layer 17278 that seemed to be a walking surface, partially prepared by putting limy and clayey material down, partially created by trampling. This layer was first discovered in the southwest corner of the Building while cleaned the construction features here. While following the layer into the room, it started to seem less surface-like; it could just be a distinct layer of room fill that separates well from the fill on top (16988).
In an afternoon discussion, PFB, PTW and I decided to from now on try and record the outlines of units in a GIS-compatible way, so we will take several points from the top with the total station and several from the bottom to try and catch the outlines. I am not so sure this will actually give a good 3D impression of our units, but it is easy to do and can be a good help during data processing especially for the artefact specialists. |
