How They Met: Geology Solves Long-Standing Mystery of Hezekiah’s Tunnelers

020

Waterworks underlying the City of David, the oldest inhabited part of Jerusalem, have two aspects that have long puzzled Biblical scholars and archaeologists. The first is whether David’s general, Joab, captured Jerusalem from the Jebusites by entering the city through an underground passage, or perhaps a water system, that led from a spring outside the city wall to a point inside.

The second concerns the famous tunnel dug by King Hezekiah in the late eighth century B.C.E. to bring water into the city in preparation for an imminent siege by the Assyrian monarch Sennacherib, as recorded in the Bible. How did the two teams of tunnelers digging from opposite ends manage to meet, as the famous Siloam Inscription says they did? What accounts for the serpentine path of the 1,748-foot tunnel? And how did the tunnelers get enough air to breathe, especially since they had to use oil-burning lamps to see what they were doing?

In 1978, Israeli archaeologist Yigal 022Shiloh renewed excavations in the City of David and thoroughly reexamined the complicated waterworks beneath the city. I was privileged to serve as geological consultant to the excavation, for which I will always be grateful. Tragically, Shiloh died from stomach cancer in 1987 at the age of 50, without completing his final report. His 1981 article in BAR remains one of the best and most vivid overall descriptions of ancient Jerusalem’s underground waterworks.^a

What follows is a more detailed geological explanation of how these waterworks were formed. This is the necessary context for exploring the two puzzles mentioned at the beginning of this article, as well as several others discussed below.

The City of David, situated on a spur extending south of the Temple Mount, is bordered by deep valleys on all sides except the north. On the east is the Kidron Valley; on the west is the Tyropoeon, or Central, Valley, which separates the City of David from Mt. Zion farther west; the third, the Hinnom Valley, curves around Mt. Zion to join the Kidron Valley on the south side of the City of David.

There are actually three water systems beneath the City of David. All begin at the Gihon Spring, low on the eastern slope, near the floor of the Kidron Valley. Indeed, the city was originally founded here because of the Gihon Spring, Jerusalem’s only freshwater source.

The first of these underground water systems, the Warren’s Shaft system, includes a short, irregular tunnel that carries the water from the Gihon Spring, which gushes intermittently within a cave, to the bottom of a vertical shaft—Warren’s Shaft^b—that goes straight up for 37 feet. Someone standing on a platform at the top of the shaft could drop a bucket on a rope and draw up the water. At the top of the shaft is a gently sloping, horizontal, semicircular tunnel approximately 90 feet long. Toward the end of the semicircle is a 9-foot scarp; above it, a steep, stepped tunnel ascends toward ground level, ending in a vaulted chamber that is the upper entry to the system. This entryway is within the city, so that the water is accessible even though the source, the Gihon Spring, lies outside the city wall.

Ever since the modern discovery of this water system, scholars have wondered whether David’s general, Joab, used this system to enter Jerusalem and surprise the Jebusites. Those who argue for this scenario cite an intriguing Biblical account of David’s conquest of the Jebusite city that we will consider later.

The second water system, the Siloam channel, also starts at the Gihon Spring and runs south along the lower eastern slope of the City of David. The Siloam channel was used to irrigate the valleys east and south of the city (where the Kidron meets the Hinnom curving around from the west). In some places, the Siloam channel is an open, aboveground trough; elsewhere, it flows underground.

The third water system, Hezekiah’s tunnel, also begins at the Gihon Spring. Hezekiah’s tunnel utilizes the short, irregular tunnel from the spring to the base of Warren’s Shaft (thereby proving that this tunnel is later than the Warren’s Shaft system) and continues underground until it debouches 1,748 feet later into the Pool of Siloam (not to be confused with the Siloam channel).

According to the Biblical accounts (2 Kings 20:20; 2 Chronicles 32:2–4, 30), Hezekiah built this tunnel to provide a secure water supply during the 701 B.C.E. siege by the Assyrian forces of Sennacherib. Both the Bible and a cuneiform account by Sennacherib agree that the siege failed; Sennacherib was unable to conquer the city. No doubt the water that flowed through Hezekiah’s tunnel was a critical factor in the city’s ability to withstand the siege.

A famous inscription carved on the wall of the tunnel,^c near its southern end, states that it was dug by two teams working from opposite ends. (See “Siloam Inscription Memorializes Engineering Achievement” for the text of the Siloam Inscription.) Ever since American orientalist Edward Robinson rediscovered the tunnel in the mid-19th century, scholars have puzzled over how the two teams digging from opposite ends of the city managed to meet in the middle, especially after the winding route they took. And 023why did they follow such a circuitous route when it would have been easier—and shorter—to tunnel in a straight line? From the Gihon Spring to the Pool of Siloam in a straight line is only 1,050 feet; the actual length of the tunnel, however, is 1,748 feet, more than 66 percent longer than necessary.

Ever since these subterranean waterworks were discovered, the majority of scholars have regarded them as man-made and have tried to explain the puzzles about them in terms of human intent or error. Thus, for example, to explain why Hezekiah’s tunnel follows such a circuitous route, some scholars have argued that the southern bend in the tunnel was necessary to avoid the desecration of passing underneath what may be royal tombs above.¹ The northern bend of the tunnel was supposedly planned to meet a well within the city²—but no well was encountered anywhere along the tunnel. Another scholar has suggested that the frequent meandering near the meeting point was due to “false echoes” from the axes, which misled the workmen, and to their “nervous haste” to achieve the imminent meeting.³

Or take the semicircular course of the horizontal and stepped tunnels at the top of Warren’s Shaft: The straight-line distance between the semicircle’s ends is only 82 feet; the length of the tunnels, however, is 136 feet, an unnecessary 66 percent of additional tunneling. This curved route has been explained on the basis of the need to moderate the slope.⁴ If this was so, how can we explain the fact that the slope of the stepped tunnel is so steep, averaging about 33 degrees? There is a scarp nearly 9 feet high at the bottom of the stepped tunnel, which obviously would require a ladder to traverse; even then, the climb would be very inconvenient, especially when carrying buckets of water. This scarp between the stepped tunnel and the horizontal tunnel has been explained as a defense measure. An exit from the horizontal tunnel emerges on the eastern slope at a point outside the city wall. This tunnel was supposedly made for the disposal of excavation waste.⁵ However, since it required substantial extra labor and, in addition, compromised the security of the city, this explanation is rather weak.

Or look at the dead-end shaft on the plan. This 025shaft simply stops after a short distance and serves no useful purpose: It has been explained as an earlier attempt to do what Warren’s Shaft system eventually achieved, but the earlier diggers supposedly hit such hard rock that they could not penetrate it; they therefore tried elsewhere with more success.⁶

I believe there is a simple, consistent and unified solution to these and other anomalies. Instead of the frequently forced, unconvincing and speculative answers, tailored separately to contend with each anomaly, I suggest a single explanation that will account for almost all of them. Moreover, this explanation is based on solid physical evidence that can be verified in the field.

Previous explanations of these anomalies have been based on the assumption that these water systems were all man-made. However, as geology shows, the channels and shafts were already there for man to enlarge and utilize.

Let me explain: Underlying the City of David is a well-developed natural karst system. Karst is a geological term that describes an irregular region of sinks, caverns and channels caused by groundwater as it seeps and flows through underground rock formations. The ancient City of David was built on a limestone and dolomite hill. Knowledge of the geology underlying this hill will help us understand most of the apparent anomalies of the underground water systems.

Beneath the City of David lie two strata or layers of rock that dip to the southeast—Meleke limestone and, beneath that, Mizzi Ahmar dolomite,^d deposited about 90 million years ago in the Upper Cretaceous era.

The Meleke limestone consists of about 30 percent fossil fragments of mollusk shells and some coral. Originally, most of the matrix of the rock was fine-grained lime mud that has now recrystallized into rocky material. Because of the dissolution and removal of the fossil fragments, the Meleke limestone is very porous and permeable. By contrast to this porous limestone, the underlying layer of Mizzi Ahmar dolomite is very 027dense and hard-practically Impervious.

The limestone and the dolomite were each affected by the formation of the karst, but in different ways, depending on their lithologic properties and stratigraphic relationship.

In porous, permeable limestone, water can move relatively freely in all directions-not so in the practically impervious dolomite beneath. Thus when water reaches the lowest level of limestone and then encounters the hard dolomite, it is forced to continue its flow along the boundary between the dolomite and the less resistant limestone, resulting in the formation of caves and channels along the contact line. See, for example, the tunnel at the top of Warren’s Shaft as it continues to an exit on the eastern slope, previously thought to be an exit for removal of tunneling debris. This tunnel lies along the contact surface between the limestone and the dolomite.

Dissolution also occurs in the dolomite, but here it typically starts in faults, joints, cracks and fissures. Once an initial dissolution is formed, it draws more water and continuously enlarges, forming vertical shafts connecting narrow, subhorizontal flow nets and conduits. Karst features in the Mizzi Ahmar dolomite around Jerusalem include irregular dissolution cavities, small caves and short shafts filled with dark brown loam, clay and mud.

Our geological examination of the subterranean waterworks beneath the City of David indicates that they were fashioned essentially by skillful human enlargement of natural (karstic) dissolution channels and shafts that were integrated into functional water supply systems.

The Gihon Spring itself issues in a karstic cave. Gihon means gusher or gushing, suggesting something that bursts forth suddenly, intermittently and perhaps mysteriously. That is what the spring did, until recently.^e Its pulsating flow is caused by the periodic charge and discharge of a natural subterranean siphon, typical of karst terrains.

Warren’s Shaft, with its irregular shape and the calcareous crusts that coat its walls, is a natural sinkhole. The shaft developed along a joint that can be traced for its entire length. Its bottom narrows into a funnellike shape, with rough walls, entirely similar to the natural termination of a karstic sinkhole. In fact, the sinkhole called Warren’s Shaft extends nearly 9 feet below the level of the short tunnel that connects it to the Gihon Spring, an extension that is wasteful and unnecessary from a functional viewpoint.

To test whether Warren’s Shaft was in fact a natural sinkhole, we analyzed a fragment of calcareous crust from its irregular walls for carbon-14. It contained none, indicating that the crust is more than 40,000 years old: This provides unequivocal evidence that the shaft could not have been dug by man.

I mentioned earlier the dead-end shaft adjacent to the Warren’s Shaft system, which has been explained as an earlier attempt to create a shaft—like Warren’s—that had to be abandoned when the ancient diggers hit hard rock. This shaft is still filled with debris, and we were not able to reexamine it.⁷ But from 029Louis-Hugues Vincent’s meticulous description of it in 1911 (he noted its irregular shape, its rough walls, natural encrustation and gradual downward narrowing) it is clear that it too is a natural karstic sinkhole.⁸ In short, neither Warren’s Shaft nor this dead-end shaft are manmade.

The semicircular tunnel that leads from the top of Warren’s Shaft was originally a dissolution channel with its entrance upslope on the hill and its exit downslope, connecting several caves along the way.

By exploring the recesses of the cave in which the Gihon Spring issues, the ancient inhabitants of Jerusalem found an irregular fissure that led from the spring to the sinkhole we call Warren’s Shaft. By carefully deepening and widening this fissure, they caused the water from the spring to flow to the bottom of Warren’s Shaft. They probably sealed the bottom of the shaft to prevent leaking and partially filled the sinkhole that extended almost 9 feet below the level of the tunnel connecting it to the spring. In this way they made the sinkhole into a well shaft: By lowering a bucket on a rope, they could bring up water from the base of the shaft.

The semicircular karstic tunnel at the top of Warren’s Shaft provided an entry from the surface to the top of the “well.” In the steeper part at the beginning of this semicircular tunnel, the ancient inhabitants of the city created steps. Where the slope was more gradual, they simply enlarged the opening as necessary. They encountered the dead-end shaft on the way, but covered (or filled) it to avoid the pitfall. Leading out of this semicircular tunnel, another tunnel exits on the eastern slope outside the ancient city wall. Once explained as having been dug to allow the ancient tunnelers to remove excavation debris, this exit tunnel was in fact a preexisting natural tunnel that they blocked to make the system safe from invaders.

In short, the Warren’s Shaft system consisted almost 030entirely of natural preexistent karstic features. These natural elements were not significantly altered, which explains why this installation contains so many apparent “mistakes” and is so inconvenient to use. This also accounts for the circuitous shape of the tunnel leading from the top of Warren’s Shaft up to ground level.

David’s capture of Jerusalem is described in two Biblical passages—2 Samuel 5:6–9 and 1 Chronicles 11:4–7. The texts are difficult and in many places corrupt,⁹ but if we put them together in their most likely meaning, this appears to be the story: The Jebusites taunted David that the lame and the blind could defend the city; perhaps in scorn they even placed the lame and the blind on the city walls as make-believe defenders. David reacted by offering to appoint as commander of his army the man who would first smite the Jebusites by—doing something. Exactly what was done is one of the oldest scholarly cruxes; the Hebrew is yigga b-tsinnor, which means to touch or get to the tsinnor. The critical word is tsinnor. Whatever was required to be done, Joab succeeded in doing it. As a result, David captured the city and made Joab commander of his army. Jerusalem thenceforth became the City of David.

Aquila, in his second-century C.E. Greek translation of the Bible, translates tsinnor as “watercourse,” and the King James Version translates it as “gutter.” Yet these translators knew nothing of the existence of Warren’s Shaft. Highly regarded modern translations like the Revised Standard Version and the New Jewish Publication Society translation render tsinnor as “watershaft” and “water channel” respectively. Other eminent scholars disagree, claiming that tsinnor refers to some kind of grappling hook to climb the walls.¹⁰ The most recent and thorough study of the word tsinnor—for the first time using, for comparative purposes, Canaanite Ugaritic—concludes that “Ugaritic philology supports the best-attested sense of tsinnor in Hebrew, namely, as a kind of water-shaft or part of a water system” (see “Up the Waterspout: How David’s General Joab Got Inside Jerusalem”).

Scholars who translate tsinnor as some kind of water channel suggest that Joab daringly entered the Gihon Spring, found Warren’s Shaft (the tsinnor), clambered up and surprised the Jebusite defenders.¹¹ Charles Warren, who rediscovered the shaft, recounts how, with the use of a few boards, he and his able assistant Sergeant Birtles were able to ascend Warren’s Shaft.¹²

Yigal Shiloh argued that hewn, man-made waterworks are unknown in the pre-Israelite cities of the country, and that therefore Warren’s Shaft system could not have existed when David captured the city.¹³ Linguistic issues aside, our geological investigation establishes that ancient (Jebusite) Jerusalem could have been entered from at least two points outside the walls: one from the Gihon Spring and up Warren’s Shaft and another from the tunnel that exits on the eastern slope. Both of these routes led to the main tunnel above Warren’s Shaft.

There can be no doubt that this karstic system existed at the time of David’s capture of Jerusalem in about 1000 B.C.E. Even if the Warren’s Shaft system, including the exit tunnel to the eastern slope, was not yet used as a water supply system, it was probably wide enough for a person to pass through.

An understanding of the karstic system also explains most of the anomalies in Hezekiah’s tunnel, especially its serpentine route. It also solves another technical problem that is rarely mentioned in the literature: the problem of ventilation. How otherwise could miners working at the blind end of a narrow passage, using oil-burning lamps, get enough air to survive?

Finally, the presence of karst conduits also explains the puzzle of how two teams digging from opposite directions, as described in the Siloam Inscription, maintained accurate underground orientation, progressed toward a planned meeting point, and eventually joined up with one another. Although we know almost nothing about surveying methods and mining techniques of the time, both ventilation and orientation would appear to be insurmountable obstacles to digging a new tunnel.

The answer is that a natural dissolution channel ran a winding course under the hill, connecting an opening on the southwestern slope of the city with the 031Gihon Spring. Perhaps this opening was large enough for a man to crawl through from end to end. Or perhaps the channel was first enlarged to allow passage and ascertain that it connected with the Gihon Spring. Once the connection was established, massive two-way tunneling could be launched with assurance of success.

Other investigators have suggested—correctly, as now can be shown—that Hezekiah’s tunnel followed a natural fissure.¹⁴ We have reexamined the careful measurements of the ceiling height in the tunnel made by Vincent during the Parker mission (1909–1911).^f Where the channel’s original ceiling was too low, the tunnelers raised the ceiling, as is clear from their pick marks on the smoothly hewn rock plane. At other points, the original ceiling was more than high enough, and they left it irregular and undisturbed. Most of these high sections (often higher than 6 feet) consist of rough bedrock, relics of the original dissolution channel.

The drawing below shows the height of the ceiling over the course of Hezekiah’s tunnel. For most of its course, the tunnel has been hewn to a height of between 5.5 and 6 feet by raising the ceiling; along several stretches, the ceiling is too low to pass under without bending; at its lowest point, it drops to 4.75 feet. At irregular intervals, however, especially in the southern section, the tunnel sometimes rises to a height of 16.5 feet! Obviously this was not done by human tunnelers. Furthermore, this indicates that the 032opening of the original karstic channel at the southern end of Hezekiah’s tunnel was some 16 feet above the level of the Gihon Spring; hence, the original slope of the natural channel was toward the north.

The floor of Hezekiah’s tunnel now proceeds down a smooth but very slight slope from the spring to the Pool of Siloam, descending a mere 12.5 inches in the course of 1,748 feet. This precision suggests that anomalies like the varying height of the ceiling and the circuitous route of the tunnel were not the result of incompetence or carelessness. After the initial enlargement of the channel, the slope of the channel was reversed, probably by a single team of workers who progressed from the spring toward the south.

The Bible itself hints that it was necessary to reverse the inclination of the tunnel floor in order to direct the water flow southward. The tunnel is referred to twice in the Bible and once in the Apocrypha (or, in Roman Catholic terms, the Bible’s deuterocanonical section). It is first mentioned in 2 Kings, where we are referred to the Book of Chronicles of the Kings of Judah for more information about the “conduit”; clearly the tunnel was an outstanding achievement of Hezekiah’s 42-year reign (from 729 to 686 B.C.E.), especially because of its possible role in saving the city during Sennacherib’s siege in 701 B.C.E.:

“The rest of the deeds of Hezekiah, and all his might, and how he made the pool [of Siloam] and the conduit and brought water into the city, are they not written in the Book of Chronicles of the Kings of Judah?” (2 Kings 20:20).

The Book of Chronicles of the Kings of Judah has not survived. But, fortunately, we have two other references to this remarkable accomplishment. One is in the Book of Ecclesiasticus, a part of the Apocrypha for Jews and Protestants, but part of the Bible for 033Catholics. There we are told:

“Hezekiah fortified his city, bringing water within the walls. He drilled through the rock with tools of iron and made cisterns for the water” (Ecclesiasticus 48:17).

Another reference, in the Biblical Book of Chronicles, indicates that, in the course of tunneling, the slope of the tunnel floor had to be reversed so that water would flow from the Gihon Spring to the Pool of Siloam:

“When Hezekiah saw that Sennacherib intended to attack Jerusalem, he planned with his civil and military officers to stop up the water of the springs outside the city; and they helped him. They gathered together a large number of people and stopped up all the springs and the stream which flowed through the land. ‘Why should the kings of Assyria come here and find much water?’ they asked … Hezekiah closed the upper outlet of the waters of Gihon and directed them down to the west side of the City of David” (2 Chronicles 32:2–4, 30; italics added).

The Hebrew word here translated as “directed” is vayeyashrem, which would be better translated as “straightened” or “straightened down” (see the Jewish Publication Society translation). The text seems to mean that by careful downward cutting of the floor Hezekiah’s workers reversed its inclination. Originally, the Gihon waters drained into the Kidron Valley through a natural karstic conduit (see drawing above). When the Warren’s Shaft system was first discovered and adapted (either by the Canaanites/Jebusites or by the Israelites), the water still flowed out to the Kidron, and the tunnelers could work in dryness. When work was completed, they blocked the Gihon’s exit to the Kidron. This facilitated the flow of water to the bottom of Warren’s Shaft. The head of the Siloam channel (essentially an irrigation channel) is 7.9 feet above the level of the Gihon Spring. This opening is referred to in 2 Chronicles as “the upper outlet of the waters of the Gihon.” When the Siloam channel was constructed, it was necessary to build a small dam in order to raise the water to the appropriate height. This dam probably had a lower opening as well, at the level of the Gihon Spring. The two systems, Warren’s Shaft and the Siloam channel, could have functioned simultaneously by regulating the flow of water through this lower opening. When the Israelites came to dig Hezekiah’s tunnel, it would have been logical to proceed as follows:

• Reopen the original outlet of the Gihon into the Kidron Valley. This would drain the place dry immediately.

• Remove the Siloam channel dam, as a matter of convenience. It could have been left in place, but that would have made the evacuation of waste material more difficult.

• Rebuild the dam and restore the Siloam channel 064after the threat of Sennacherib’s siege passed.

The Israelites probably did rebuild the dam, because the Siloam channel served a very important function in irrigating the fields east of the city.

Because of the vaulted chamber (dating to the Roman period) at the top of the stepped tunnel, Shiloh speculated that the Warren’s Shaft system remained in use in the Roman period, simultaneously with the two other water systems. This is quite possible, although to use the Warren’s Shaft system simultaneously with Hezekiah’s tunnel, it would have been necessary to do something to raise the water level in the “water room” at the bottom of Warren’s Shaft. This could have been done easily by constructing a small dam somewhere along Hezekiah’s tunnel.

It does seem likely that the Warren’s Shaft system continued to be used even after Hezekiah’s tunnel was constructed. Those people living on the top of the hill and on the eastern side of the city, would find the Warren’s Shaft system more convenient. Others, living on the other side of the city, could use the water from the Pool of Siloam. At appropriate times, fields in the Kidron Valley east of the city and fields south of the city could be irrigated by means of the Siloam channel.

The design of the systems was determined, however, at nature’s direction. Man adapted these designs imaginatively and with great skill. Through an understanding of how these elements worked together, we can solve most of the puzzles that have heretofore stumped researchers.^g

Waterworks underlying the City of David, the oldest inhabited part of Jerusalem, have two aspects that have long puzzled Biblical scholars and archaeologists. The first is whether David’s general, Joab, captured Jerusalem from the Jebusites by entering the city through an underground passage, or perhaps a water system, that led from a spring outside the city wall to a point inside. The second concerns the famous tunnel dug by King Hezekiah in the late eighth century B.C.E. to bring water into the city in preparation for an imminent siege by the Assyrian monarch Sennacherib, as recorded in the Bible. […]