.bk A12
.fl L813jl.j
.fd feature descriptions
.ed L813
.ei jl

.ri jl
.rd L813

f 375
H2 It is clear from the interlocking of the stones of this ledge with the stones of the rest of the wall of the circular area that it dates from the original construction of the underground structure, and is in fact a continuation of the wall.  Furthermore, the intersection of the northern stone wall of the square, k26, does not display such interlocking, and instead appears now to have been added on later.
H3 If our observations are correct, then f375 was the original western wall of the underground structure.  This would mean that the building was originally circular, and perhaps was not even a building at all, but a stone-lined well or similar construction.  Hence f375 and f352, parts of this construction.  At a certain time, possibly late phase 1 or early phase 2, as the pottery seems to suggest, this wall either fell or was knocked down.  Instead of rebuilding it in the same shape, the square area was attached to the standing walls as an entrance area.  f353, and possibly f354 and f355, are other remnants of this original construction, the highest surviving level of the wall at that location.  Then, the sherd packing f349 was laid down to raise the floor level above the tops of these stones.
H99 If this was originally a well, why was it ultimately filled in?
M1 It seems fairly clear from the construction techniques that this was the original western wall of the underground structure, i. e. that it was a large stone-lined shaft, in the manner of a well.

f 349
H2 The sherd material of this feature extends in between and underneath the stones of f376, the north wall of a32 in k26.  The feature itself is very thick, but the top is at approximately the same height as the tops of the stones in f353.
H3 The sherd packing must therefore be earlier than or contemporaneous with f376.  It seems likely that it was laid at the same time as a36 was constructed in order to raise the level of f348, the first floor, above the tops of the rocks f353.  See f354 and a32 for more notes on this theory.
M1 Originally labelled as a floor, this sherd packing was leveling material underneath the floor f348.

f 374
H99 This part of the mud ledge had been laid on top of the stones of f375, obscuring it from view.  Hence, we originally believed that the stones were a foundation of sorts for the mud ledge.  It was not until the winter rains washed away most of the ledge that we began to suspect that f374 and f375 were touching through coincidence, not on purpose.

f 373
H1 This earth was melted, mixed, and deposited by rains and flooding during the previous winter.
H2 We found a number of pieces of plastic within this accumulation which apparently were deposited over the winter.
H3 We believe that this material was unstratified, hence the assignment of a new feature number.  Even though it is equivalent in elevation to parts of f351, f351 was well-stratified.
=l f368

f 372
nr We excavated very little within this feature, once we realized that it was different from f369.  There was too little time left within the season.

f 371
na This feature sloped up against the north wall of a33.  If it is actually equal to f372, it is curious that it suddenly levels out, especially given the sharpness of its slope.
nr Some of the lowest levels of f369 were probably f371.  At the time, we did not understand that it was a series of thin layers, rather than one thick red layer.  So, we were searching for a clear red layer, and could never find it.  It is also possible that it did not slope quite as sharply as our excavations revealed it, if we were actually excavating material from it on the north side.  Still, it is a difference of only a few centimeters.

f 370
nl This feature almost certainly has an MZ12 feature number, but until that data is reorganized, it is difficult to know what that number is.  It is possibly f258, which covered the entire underground structure during MZ12 and was mixed red and gray, according to lr.

f 369
H2 This feature is below the lowest depth of excavation last season, even the f351 trench.  There was also no plastic found within it, unlike the higher layer, f378 (except for the plastic bag found within the fissure; however, it is almost certain that this was washed into the fissure from above, especially considering that that section of the fissure was without dirt fill at the end of last season).
H3 I believe that we have reached the beginning of stratified accumulation again.

f 368
=l f373
H1 This dirt was melted, mixed, and deposited by rains and flooding during the previous winter.
H2 We found a number of pieces of plastic within this accumulation which apparently were deposited over the winter.
H3 We believe that this material was unstratified, hence the assignment of a new feature number.  Even though it is equivalent in elevation to parts of f351, f351 was well-stratified.

f 367
nr Stef essentially scraped out the contents of the fissure from top to bottom.  The fill was very loose and crumbly, probably having been damaged somewhat by the winter rains.  It was very enlightening to see, after he had done that, that the fissure was itself only about 10cm wide.  The confused mass at the top of it (f303) is only partly the fissure; the rest of it is sheets of accumulation, f304 through f306, that have split off from the main body and collapsed on top of the fissure, like sheets of ice from a glacier.

f 366
H99 This is the lowest level of the mud ledge, and is at approximately the same level as the tops of the stones in f353.  If so, then this might be the first layer of the mud ledge, resting directly on top of the stones f352.

f 258
H99 This feature number referred to a series of alternating red and gray layers, which were too broken and confused to be separated clearly.  It is also the highest preserved layer in the k27 south baulk.  In MZ12, it covered the entire area of the underground structure.  For this reason, it seems likely that the feature which we have labelled f370 is in fact a part of f258 from MZ12.

a 32
H2 The clarification of the stone ledge, f375, shows that it is a continous part of the north wall of a32.  It is also the upper half of f352.  A fairly clear seam of mud-mortar separates the stones of the square, a36, from the stones of the circle, a33.  The first floor in a37, f348, is much higher than the floor in a33 (which we have not yet reached).  f353, the remains of the previous circular stone wall underneath the sherd packing, f349, seems to be in a fairly continous alignment with the stone ledge f375.  a33 is becoming very deep, over a meter lower than the floor in a36.
H3 Finally, the underground structure's architecture is beginning to make sense.  It appears that f375 and f352 are the remains of the original western wall of a33, which was once the original extent of the underground building, i.e. it was a stone-lined circular shaft into the earth.  At some point, this wall either fell or was knocked down, and a36 was added on in the western half as an entrance area to the building.  Another possibility, of course, is that there was an earlier western antechamber, but that it was smaller than a36, and that they were forced to enlarge the hole in the western wall when they built a36.
M1 Our current understanding of the evidence suggests that a32 was an ancient well of some sort.
M99 This building continues to surprise us.  The clarification of f375 shows that a33 was probably the original extent of the building, and has brought us back to our original idea about the structure's function: a well.  Urkesh was clearly a large city even in the middle of the Early Bronze Age, and such a city would have needed a lot of water, which either a large well or many small wells could provide.  But if it was a well, why let it fill in?  Perhaps the well went dry at a certain point, or perhaps the city was sacked and the well despoiled by invaders.  That might explain the thick ash layer f307, which we currently believe immediately postdated the loss of the roof of a32.  However, there is no evidence except possibly f307 for such a situation; further, the apparent progression of the pottery within the circle suggests that the accumulation within it occurred over an extended period of time.  Further, why have one central well, rather than many smaller ones?  gb has pointed out that one large well would probably deplete the water table in its area much more quickly than many smaller wells spread around.  Of course, perhaps this is just what happened.  Its later use is also unclear; the contents of the fill within the building, especially the circle, resemble the contents of a dump or a pit, but the fill itself seems to clean and orderly to have belonged to a garbage pit.  gb has again jokingly raised the idea that it might have been some sort of cage for wild animals, such as a lion, but this idea also has its problems.  One, the bone specialists ruled out the possibility of a lion having gnawed on the bones within the circle.  Two, the mud ledge f326 seems to be precious little protection against the claws of a lion or other wild beast.  Even though this theory is mostly tongue-in-cheek, it illustrates an important point: at some point, presuming that the structure started as a well, it changed its function at the same time that it began to fill in with accumulation.

a 33
H2 We have always presumed that a32 (and hence a33) predated the AP building, since the walls of the AP building were clearly built to accomodate its presence, i. e. they change course to skirt around it.  The top level of the eastern wall of a33, being at approximately the same elevation as the first floors in AP, also gives a fairly good indication of the level of the tell surface at that location at the time that AP was constructed, i. e. the ground was already higher there than around the AK building.  The level of the floors of the AK building, however, are much lower than the top of a33, suggesting that the tell surface was lower around the AK building, or that there was a fairly significant slope already in existence at that time.
H3 If the hypothesized slope did in fact exist, and if the a33 shaft was dug straight down into the hillside, it seems possible that part of the western wall of a33 would have been exposed to view.  The walls of a32 do not appear to have ever been freestanding, nor do they appear capable of ever having stood without earth behind them for support.  However, since the original western wall no longer exists, it strikes me as possible that the construction technique could have been different, a difference about which we cannot comment because it is unobservable now.  If it was exposed to the elements and to view, then this might explain why it no longer exists, that it was not built strongly enough for such wear.  However, this theory has problems, for example, height of a36.  This construction also does not seems to have ever been freestanding, and so must have been dug into something, the height of which would have been nearly equal with the top of a33.  Perhaps the retaining walls which terraced the tell underneath the AP building also brought the level of the tell surface up to that elevation further out west, necessitating or precipitating the construction of a36.  This theory still needs more development and consideration.

a
36	26	s	k26	The square part of a32, i. e. the western half.  It seems to have been added on later to the original structure, a33, but the reason for this is still unclear.

a 36
M1 This room was probably the entrance to the underground structure a32, a later addition after the collapse/removal of the western wall of a33.
M99 We do not understand at this point the reason for the construction of a36.  If a32 was in fact a regular well, why was the stairway built so narrow?  It is too narrow to have been effectively used to supply water to any significant number of people.  Also, what need could there have been of an antechamber to a well?  Even if the structure had a ritual purpose, the stairs still seem unnecessarily difficult.

a 31
M99 Whatever the function of a32 was, it seems impossible that it was used regularly, since the stairway is far too narrow to have admitted regular access.  Such a stairway suggests infrequent use.

f 326
H2 In order to have any stabiliy, the mud ledge must have accumulated against another surface, presumably the accumulation within a36.
H3 Presumably, the mud ledge increased apace with the ^a within a36.  However, there are two possible reconstructions of this event: first, k26 and k27 increased at roughly the same pace, so the ledge was never more than half a meter (or whatever appropriate size) high.  In this case, it served to keep to the accumulation within k27 out of k26.  Second, k27 continued in use as a well after a36 was constructed, so k26 increased in elevation for a long time as k27 remained lower.  In this case, the mud ledge probably served to keep the contents of k26 out of k27.