The Suba Water System as a Clay-Production Plant
Lecture by Shimon Gibson
On 30th Oct 2008
University of North Carolina at Charlotte
Ancient underground water systems have been the focus of numerous archaeological investigations at tells in Eretz Israel and some of them have been excavated, notably at Hazor, Megiddo, Jerusalem, Gezer, Gibeon, and at Beer Sheba. The general assumption made by scholars is that quite a few of these systems, if not all of them, are of Iron Age date. However, any certainty regarding the chronology of these systems has diminished in the last decade as a result of a great deal of new research and excavations. Many water systems show signs of having been reused, enlarged and adapted during the progress of time, and some of them were already fairly well-developed underground complexes when they came to be used in the Iron Age II. The complex water system at Jerusalem (the Gichon Spring), for example, has been the focus of new investigations, with an enormous amount of new information shed on its early development in the Middle Bronze Age II. Today researchers believe that the origins of many of the underground water systems in the tells of Eretz Israel are to be sought in the Middle and Late Bronze Ages, with a few founded perhaps even earlier in the Early Bronze Age. Indeed, it would appear that the inhabitants of Iron Age towns were adept in using and extending pre-existing water systems, or imitating those seen elsewhere.
A unique underground water system was uncovered at Suba, west of Jerusalem, and is the subject of the present lecture. It differs from the other underground water systems known from Eretz Israel, which I have just mentioned, owing to the fact that it is located in the heart of the countryside and not within the confines of an ancient town. Unlike those water systems in towns which were in use during a number of periods, the Suba system was only in use in the Iron Age, except for one part of the system – the reservoir – which was reused for other purposes in later periods, namely during the Second Temple and Byzantine/Early Islamic periods.
Where is the site?
The site is located on the edge of a valley which is used by Kibbutz Tzova for cultivation purposes, and is situated 9.5 km to the west of the Old City of Jerusalem, and 3.5 km to the west/north-west of En Kerem. The water system is positioned at the bottom of a steep hill (Har Tzova), next to a narrow valley known in Arabic as Wadi esh-Shemmarin and in Hebrew as Nahal Tzova. The hill on the opposite side of the valley is Har Eitan.
The only part of the water system that was visible before the excavations was the reservoir, which was full of soil and rubble, almost to its ceiling. The rest of the water system was completely buried and covered by agricultural terracing. The reservoir was investigated during an archaeological survey in the winter of 1999 by Shimon Gibson and Shimon Dar on behalf of Bar Ilan University. Excavations were subsequently made at the cave and its surroundings over five years, continuously from 2000-2003, and then during two separate seasons in 2005 and 2006. This work was conducted by Shimon Gibson and James Tabor, with the assistance of Egon Lass and Rafi Lewis. The work was sponsored by the University of North Carolina at Charlotte, and funded by the Whitehead Foundation, The Foundation for Biblical Archaeology, and Kibbutz Tzova. My thanks go to Reuven Kalifon, Yaaqov Ha-Tsubai and to Yael Kerem for their help during the work. Conservation of the reservoir was conducted at the site by the Israel Antiquities Authority in 2005.
The plastered reservoir was part of a large rock-cut water system which included the additional components: an upper spring-cave accessed via a hewn tunnel; a corridor which was open to the sky; a vertical shaft linking the upper area to the reservoir below; channels; and three external rectangular basins. In addition, eight standing stones (perhaps commemorative) were discovered leaning against one of the corridor walls.
Let’s have a look at the various parts of this water system in more detail.
The porch leading into the reservoir:
Access to the reservoir was via a porch (3.8 x 2 m) with a flight of seven steps. The porch had rock-hewn and partly built walls, with a roof made of large flat slabs laid one on top of the other (corbelling). The floor and walls of the porch were coated with a thick layer of greyish-white plaster.
A large entrance led from the porch into the reservoir, with a series of ten steps extending down to the floor of the chamber. The reservoir is elongated in shape (24.5 x 3.5 m, 4.5 m deep), in the form of a gallery, and was hewn out of rock following natural seams between thin geological layers of bedrock. In the wall immediately within the reservoir on the right was a small plastered alcove. In the ceiling was a shaft linking the reservoir with a corridor carved into the rock above it. The walls and floor of the chamber were coated with lime plaster, but not the ceiling. Trowel marks of the plasterers are visible on the surface of the plaster. The type of plaster has been identified by the Geologist Arieh Shimron as typical of the Iron Age. A radiocarbon dating of charcoal taken from within the plaster on the walls, made by Elisabetta Boaretto, established a date between 770-400 BCE. There are no signs that repairs were made to the plaster of the reservoir after it had been abandoned, even though the cave was reused for other non-water storage purposes in Second Temple and Byzantine/Early Islamic periods.
I estimate that the reservoir when full was able to contain some 300 cubic meters of water. At the back of the reservoir stalactites seen on the ceiling and travertine build-up on the walls (flowstone deposits), above the original plaster surface, are a sure indication that the main source of water derived from the slope of the hill above the reservoir. Thorium-Uranium testing of the first build-up of travertine on the walls established a date of 598 BCE.
On the floor of the cave were mud and silt deposits which indicate its final use as a reservoir. Large portions of collapsed wall plaster were found within these deposits and, lower down, closer to the floor, a few shattered storage jars in situ. They resemble types known from Gezer and Tel el-Ful dating from the second half of the second century BCE. A handle with a Yehud type stamp came from the uppermost part of these deposits.
Tunnel and corridor above the reservoir:
A rock-hewn corridor open to the sky was traced in the area above the reservoir. At its northern end was a tunnel hewn into the side of the mountain. The end of the tunnel is cross-like in plan and this was done presumably in order to enlarge and divert water from the aquifer into the tunnel. In antiquity a fair amount of water must have seeped into the tunnel based on the travertine deposits visible on the walls, but today only a small amount flows into the cave in the winter months. Midway along the corridor it widens and the top of a vertical shaft, protected by a surrounding wall built of one course of stones, is visible extending down to the plastered reservoir below. Judging by the appearance of the shaft and the adjacent corridor, some kind of lifting device must have been used in order to raise water from the reservoir through the shaft and into the corridor. Whether or not this was a kind of shaduf lifting device or a water-wheel is uncertain. Eight large stone slabs are arranged against the western wall of the southern part of the corridor. These served some function which is unclear to us. One possibility is that they were masseboth. Conceivably these stones were set up either by the owners or by the stone hewers involved in creating the water system. Perhaps –
to quote Graesser in his 1972 article – these stones served “…to commemorate an event, and more specifically, to call in mind the participants in all the honor and glory of that event.” A number of complete Iron Age II pottery vessels were found within the corridor area.
At the southern end of the corridor a narrow tunnel with a rock-cut channel in its floor extended a distance towards a small relay pool (1.80 x 2 m). Water collected within this pool could be diverted in three different directions: (1) into a large pool (5.30 x 2.70 m) to its east; (2) to a drainage channel covered with small slabs of stone which wrapped around the pool to its south and into the valley; (3) or back into the porch at the entrance to the underground reservoir. Another large pool (5.10 x 3.05 m) exists above the lower large pool, but how it received its water is unknown. The two large pools were rock-hewn and plastered. Judging by the height of the plaster on the walls of the pools they probably held no more than 6 cubic meters of water each. A hard deposit of silt-clay (15-20 cm thick) was observed stuck to the plastered interior walls of these pools, making it clear that not only was water contained in these pools but also quantities of soft clay. The lower pool was abandoned already in the Iron Age, judging by a living surface with pottery and a fireplace established inside the pool. A radiocarbon date for the charcoal from the fireplace provides a date between the 8th and 5th centuries BCE.
What about the general appearance of the valley beyond the water system? There can be no doubt that the place chosen for the water system was not accidental. It was established at the base of a steep hill and close to a point in the valley where it narrows. Aerial photographs suggest that a massive barrier wall once existed across the valley, extending to the point close to south-western external corner of the porch at the mouth of the water system. Hence, we suggest the barrier wall served as a dam and was used to contain water behind it and to the east. Geomorphological pits/sondages cut at a number of locations along the length of the valley and down to the mouth of the cave, have shown that in the Iron Age the level of the bed of the valley was much deeper than it is today by a couple of meters. It would appear that soil only began accumulating there in the Hellenistic period, perhaps as a result of deforestation activities on the slopes of the adjacent hills.
What was the function of the Suba water system? The location of the water system within the countryside and at some distance from the nearest Iron Age settlement, would immediately suggest that it had some rural function, perhaps for irrigation agriculture or for the watering of herd animals. However, the shape and appearance of the system goes against such an interpretation. Indeed, water was evidently collected intentionally within the reservoir so that at given times quantities of water might be raised with some water-lifting device to the higher level of the system. Subsequently, the water was allowed to descend level-by-level and into a system of pools. The remainder of the water flowed back into the reservoir, except for the dregs which were allowed to drain through a channel into the valley outside. Such a Sisyphean arrangement only makes sense if the water system was being used for some industrial function which would require a constant circulation of water. This industrial activity could not have produced polluting substances, otherwise the water would not have been allowed back into the reservoir as it evidently did.
A clue to the possible industrial use of the system is the thick and hardened layer of very fine marl clay silt which was seen adhering to the sides of the two large pools. This suggests the pools were used for clay-cleansing and preparation procedures, but on a very grand scale. Nearby there were probably pottery manufacturing buildings and firing kilns. The latter have not yet been found, but it is not surprising that pottery kilns from a later period (Byzantine) have been investigated not too far away, further up the Tzova Valley, close to Ein Suba.
The Iron Age II date for the water system is certain based on three types of evidence: (1) Thorium-Uranium dates obtained from the travertine covering the plastered walls of the reservoir; (2) radiocarbon determinations of charcoal from the plaster of the reservoir, and from a fireplace within one of the external pools; (3) pottery vessels of 8th to 6th century date from within the fills in various parts of the system.
What terminology should we use to describe the various parts of this water system? I would suggest that the term ashokh is appropriate for the entire water system. This word is not mentioned in the Old Testament, though it does appear in the Mesha Stele. I would use the term bor for the underground reservoir, with the term geb to describe the ponding of rainwater behind the barrier wall in the valley.
What parallels may we suggest for the Suba water system?
As far as I am aware, this is the first time that a water system devoted to the cleansing and preparing of clay for pottery manufacture, has been uncovered in an archaeological excavation in this country. Certainly ethnographic parallels may be adduced for the function of the pools at Suba as installations for clay preparation. A few such pools still exist at a number of pottery-production locations in Gaza and near Sebaste, but they are fastly disappearing.
In terms of the overall plan of the Suba system, I think we can favorably compare it to an Iron Age water system from the Amman Citadel. However, the closest match is the early Iron Age water system uncovered in recent excavations at Beth Shemesh. Its features are remarkably similar to those of the Suba water system, including a flight of steps leading down to the reservoir entrance, an alcove on the right of the cave as one enters the interior, the consistency and trowel-finish of the plaster-coated walls, and the appearance of its horizontal galleries. In the case of Suba, the underlying reservoir consisted of only one gallery, but the mode of hewing is remarkably similar.
To which settlement did the Suba water system belong to?
On the basis of archaeological surveys conducted in the vicinity of Tzova, we can eliminate Sataf and Har Herat which have only a few sherds or non at all. There is only one large site with substantial remains from the Iron Age II and that is site of the Arab village of Suba which is situated at a distance of about half-an-hour’s walk from the water system. Its exact identification and ancient name is uncertain. Iron Age pottery and artifacts were recovered there during excavations conducted by a British team working in the late 1980s on the medieval fortress. Iron Age II pottery was also found in excavated fills near Ein Suba. Lower down to the north, in the area known to kibbutz members as “bamat ha-sela”, close to 40 caves have been identified, with at least three undoubtedly of Iron Age date, with benches and carved headrests, and repositories. Finds from within these caves included pottery and jewelry, notably finger and ear-rings. This site is situated slightly less than one kilometer to the north of the underground water system that we have uncovered in Nahal Tzova.
• The Suba water system was hewn during the Iron Age II, probably in the 8th century BCE. The system (which I call an ashokh) comprised a number of components, notably a large plastered enclosed reservoir (bor) which received water through a tunnel from a mountain aquifer. Outside the cave and in the valley was a very large external pool open to the sky (geb).
• The water system was not used for agricultural irrigation or for the watering of animals, but had an industrial function. We believe it served for the purification and preparation of clay reserves necessary for large scale pottery-making procedures. The hewing of this system was probably an enterprise instigated by the state of Judah and it was probably controlled by them (see I Chronicles 4:23).
• The clay-purification plant was abandoned in the Late Iron Age, perhaps during the 6th century BCE, but the reservoir continued to be used for water storage until Hellenistic times (2nd century BCE).
• In the first century CE the abandoned system was reused for water-purification rites. It was also used as a Christian memorial cave in the Byzantine and Early Islamic periods but water no longer accumulated within the cave.