| Entry: | Today we had a lot of action in the trench, as five students from Southampton University are joining us for three days.
Three of them were working with DLG, moving a lot of soil in Building 107 (16988), uncovering features and finding loads of artefacts. We are not taking back 16988 all the way to the western section, but will leave a 70cm step to increase safety. After 16988 is finished, which will be tomorrow as the excavators are so fast, we can clean the room, 3D it, and start taking out wall phases from the top.
KTX had jack working with him in Building 106 today. Jack scraped the face of southern wall F.3351 to see the brick laying, and then started removing 1.5 layers from the top to expose the top of the second lowest layer. He started in the middle where things are clear; the wall has a concave base, it get thinner and also more confusing towards east and west – we will see.
KTX worked on the central part of wall F.5058, and by the end of the day we had a clear picture how this wall was built. He exposed parts of three brick layers, and he is leaving the easternmost slice (thickness ca. 3cm) of wall F.5058 against F.2424 to have a section of F.5058. It seems like F.5058 was built from north and south simultaneously, and then joined in the centre. One indication could be that the base of F.5058 is also concave, but that could also be the result of unevenly prepared construction site. Another is that the western face of the wall makes a slight bend here. Also, the southern part seems to have built with a different style, as we actually found one brick in oblique angle to the others, with the narrower side facing the room – never seen in the rest of the wall. Also, in the centre, we seem to have bricks of one layer not nearly set next to each other as usually, but overlapping, as if they had to make them fit. KTX also exposed one layer in plan view of the connection of F.5058 with buttress F.3301. In the front part of F.3301, he found an unusual brick – it was reddish, rather small in plan but quite thick. One smaller brick was put about 4cm into the wall, and then runs into F.3301. When we sectioned the upper part of the connection a few days ago, we could see that the mortar layers of F.5058 continued into F.3301, where they slope down following the base of the buttress. In the connection point, the mortar and brick lines were slightly offset, though. They stopped, then there was a 0.5cm gap, and then the continued ca. 1cm further down. The brick layers of F.3301 have a curved shape when seen in section. If they were manufactures outside the wall, and then applied to the wall, they must have been wet and kind of pushed into their curved shape. This curving was even more violent in F.3302 – one still wonders if it had not been easier to prepare a level surface for the buttress before construction, rather than having to compress bricks into weird forms. Maybe it was not acceptable for them to cut the tops of old decayed walls when they wanted to build on top.
EUR finished cleaning the faces of F.3303, F.3353, F.3366 and F.3368 in southeastern Building 5. Normally, none them that has signs of plaster, and I began to think they never were plastered. The very corner of F.3368 and F.3366, though, EUR found a patch of very thick plaster attached to F.3366. Also very interesting is the base of buttress F.3363. It seems to be separated from F.3368 by an L-shaped fill layer, meaning the wall makes a step down here, which is filled, into which the buttress as built. F.3363, however, seems directly abut by the lower buttress F.3366, and connected with it by a tiny rest of plaster in the corner.
CMF worked in the southwest corner, and mainly removed fill that was interpreted as parts of a wall beforehand, but turned out not to be. What might be the southern continuation of wall F.3346, or another wall (see discussion below) is quite clearly visible now. What might be southern F.3352 over F.3346, or some wall in line with F.3352, kind of wants to be visible, you can feel it, but it is not quite there yet. The uncovering process is difficult as the fill surrounding it consists mainly of refused building materials, which we do not just hack out as we expect possible cut and destroyed walls here.
We have been discussing phasing in B.105 for several days with several people. All people involved in the discussion agree that walls F.2424, F.5051, F.3352, F.3341, F.3364 together with buttresses F.5062, F.5061, F.5063 and F.3363 form one stratigraphical complex (A), as they are on one level and together form one regular structure. Walls F.3346, 3311/3310, F.3303 and F.3368 together with buttresses F.3365, F.3353 and F.3366 formed another complex (B). There are currently three main ideas about how A and B relate chronologically: Hypothesis 1. A is younger than B; B represents the old building whose outlines were copied in the younger building A, which was, however, larger and had an entrance in the south. Main arguments for this hypothesis, without having taken any of the features out to actually check their connection, is that A is higher than B, and that the features of B are more destroyed than quite well preserved A. Hypothesis 2: B is younger than A and was built to hold up unstable complex B. Hypothesis 3: A is the upper storey to B. To my eyes, this seems improbable as the bases of A are so irregular.
Without removing any walls, there are still several points that can help us solving stratigraphy. There is no western buttress in complex B; buttress F.5062 seems to either sit on a long wall F.3346 with an irregular base, going deeper in the east (in the room) than in the west (on top of the wall), or F.3346 and a wall with same orientation south of the buttress were built abutting F.5062. Buttress F.3363 seems to be sitting on fill on wall F.3368, but seems to be abut by F.3366 – unless the latter was cut very neatly and F.3363 built perfectly into the cut.
GWN and JHB worked further on wall plaster phasing in Space 341 and Space 340. Also, Emily scraped back the dried out front of wall F.3333 so we could see its internal structure. Even after spraying it in the end of the day, though, no layering was so obvious. I hope it will become more apparent to us once we stared at it for longer, and sprayed it several times. However, F.3333 was always a bit different, its plaster was only preserved in the lower part.
The phasing recognition in Sp.341 and Sp.340 takes up a long time. The summary of information gathered so far: Space 340: The lower parts is plastered by a layer (17299) which also connects to the floor 16977 and to the plaster of F.5052. This plaster layer was covered by material which also had a lot of whitish marl in it and was rather homogenous (17293), so JHB interprets is as another layer of plastering. I am not so sure about this, but agree it could represent intentional coating. As we discovered these layers when scraping down the wall, we cannot see any more how this layering connects to the upper part, so people have to believe our memory. The plaster layer (17252) in the upper part of the wall stopped ca.10-15cm above the floor, where the ambiguous plaster-fill 17293 was found under it. This would mean that the plaster we saw first, in the upper parts of the walls, is either not preserved at the bottom, or was applied at a time when the foot of the wall was covered by fill, and this would make the newly discovered plaster layer 17299, which connects to the floor, older.
This situation seems to be the same with wall F.2413/5055 in Space 340, we therefore have been using the same numbers for this wall as well (17252 upper plaster, 17299 lower plaster, 17293 fill-plaster at the bottom).
In Space 341, GWN has discovered a similar situation, with more details. A thick plaster layer 16897 (F.2413/5055)/ 17297 (F.5056) covered walls F.2413/5055 and F.5056, but not their bases. After he scraped off this layer, more layers of plaster became visible, so far summarised as 17298. 16897 and 17297 no not go down to the bases of the walls; interestingly, they stop exactly on top of the ash layer that spread in the entire room. The fill under this layer (17290) was mixed (ash, lumps, burnt clay) in its upper part, and very homogenous in the lower part – similar to 17293 on Space 340. 17298 runs behind 17290 and is abut by the plaster layer, which sticks to it, but can carefully be scraped off.
This was partially found out yesterday, and today GWN scraped the different layers off the western half of F.5056, leaving the eastern half intact as a reference point. By the end of the day, the wall was exposed in three stages: the westernmost part was scraped back to the mud brick, the central part was scraped fresh so all layers present were visible, but nothing was removed, and the easternmost part was left untouched and remains a bit grey and dusty. We still were not sure about how far down plaster layer 17298 goes. After scraping and scraping around details, with still some questions unsolved, we decided to change strategy.
Towards the end of the day, we decided that not more information could be gained out of scratching back the walls from their faces; we have the general picture, I guess, but still there are some details missing, sometimes one plaster layer has a gap that could be meaningful, sometimes we see something that might be even another layer. To speed up, we decided to section the walls in strategic places. GWN is going to take out the corner of F.2413/5055 and F.5056; JHB is going to remove the southern half of F.2413/5055 in Space 340 including half of its connection point with buttress F.5053/5054. We will thus not only see the walls and all their plastering and coating layers in section, but also tackle points important for the reconstruction of the construction techniques of the building (walls and buttresses binding or not?). One indication are the mortar layers in the upper parts of the walls; in upper F.2413/5055 and F.5053/5054, we see up to three layers of mortar made from dark grey crumbly material, one of which seems to run all through F.2413/5055, on the same level through buttress F.5053/5054, F.5057, and only stop when coming into F.3333, because this wall is not preserved high enough. It does not seem to be present on F.2428. This one mortar layer could be the indication of a so for undiscovered construction phase, during which the walls mentioned above were toped up with so far unused brick and mortar.
Before the section cutting, they cleaned the building and took another 3D photo-session. Everybody is looking forward to some picking tomorrow, after days of scraping.
In the sections, we will miss some layers that we already removed entirely (the thick upper plaster layers in Spaces 340 – 341 = 17297, 16897, 17252); if we had cut the section without previous wall scraping, we would have the entire section present, Because we tried to expose the phasing coming from the front, we now miss them – this experience encourages me to chose sectioning over flat recognition in contexts where we already assume it might be difficult, or in lack of time. |
