Ayreh Shimron answers questions and moderates an archaeological discussion
Our article explaining how the two teams of tunnelers who dug the sinuous path of Hezekiah’s Tunnel from opposite ends managed to connect (Hershel Shanks, “Sound Proof: How Hezekiah’s Tunnelers Met,” BAR, September/October 2008) produced many interesting reader responses. While we were unable to print much of this material in the magazine, the responses to these readers’ letters are written by Aryeh Shimron, one of the scholars on whose scientific publications the BAR article is based. The discussion is well worth studying.
Among the questions raised by BAR readers: How did the tunnelers from the southern end gauge the correct elevation to meet the tunnelers digging down from the spring? Could hammerings from the surface really be heard in the tunnel? How was the tunnel vented during construction to prevent the tunnelers from suffocating? Why was Hezekiah’s Tunnel so difficult to dig when other water tunnels—at Hazor, Megiddo, Gibeon and Gezer—were so relatively straightforward?
I was very interested in the latest article on Hezekiah’s tunnel (“Sound Proof: How Hezekiah’s Tunnelers Met,” BAR, September/October 2008). There are similar tunnels at Hazor and Megiddo. It would be interesting if someone could write an article comparing these water systems.
Tipton, United Kingdom
Thank you for your very interesting query. Let’s see if I can provide a satisfactory answer. There are a number of important differences between Hezekiah’s tunnel and those at Hazor and Megiddo. Besides their ages, the Hazor tunnel was dug toward the water aquifer and is 25 meters long; the Megiddo tunnel is 80 meters long and was excavated towards the spring. Both tunnels were carved out of soft chalk with a singular purpose: to link these cities with the water source that lay outside the city walls.
In contrast, the 540-meter-long tunnel of Hezekiah was constructed with two principal objectives in the minds of the engineers: (1) to divert the waters from the Gihon spring and bring them first into the core of the city, where they could be reached by digging a deep well, and (2) only then to continue the flow south toward what is now the Silwan [Siloam] pool at the city’s boundary. I can cite two sources of evidence for such an admittedly speculative scenario: First, the directly westward trend (into the heart of the city) of the northernmost leg of the tunnel (which also has two corrections made along its route), and second, the presence of a deep cistern very close to where the tunnel makes the 90-degree swing toward the southeast. I argue that this still-undated cistern may represent the start of what turned out to be an unsuccessful effort to reach the tunnel some 50 meters below. It is probable that at this point the decision to continue the tunnel in a southeasterly direction (that is, toward shallow rock overburden for purposes of acoustic communication) was made by the engineers while an identical change of direction was made by the southern team. Needless to say, the chance of meeting via a direct route, (that is, south for the northern team and north for the southern team) under some 50 meters of rock cover would have entailed a highly risky venture.
As to why the engineers started the tunnel at both ends, I think here is where we have to consider the possibility—perhaps even probability—that the Old Testament narrative regarding the threat that the Assyrian monarch Sennacherib made toward Jerusalem in the year 701 B.C.E. is historically correct. We have dated the tunnel by radiogenic techniques (Nature, Vol. 425, September 11, 2003) and shown that, in terms of its age, it can be linked with the Hezekiah-Sennacherib story in the Old Testament. If so, then Hezekiah’s tunnel is precisely that; there was little time to spare, and the job was a rather hurried effort that needed to be carried out by two teams, as the results so clearly indicate.
For more information about this topic, subscribe to our BAS Library. Read Hershel Shanks’s “Sound Proof” along with Dan Gill, “How They Met: Geology Solves Long-Standing Mystery of Hezekiah’s Tunnelers,” Dan Cole “How Water Tunnels Worked” and thousands of other articles on ancient Israel.
I enjoyed Hershel Shanks’s piece on Hezekiah’s Tunnel and would like to make several comments.
His main thesis that the tunnelers moved to avoid the higher overburden and used sound guidance from above was already suggested by N. Shaheen in 1979 (“The Sinuous Shape of Hezekiah’s Tunnel,” Palestine Exploration Quarterly 111, pp. 103-108) but the consensus has been that, unless the tunnelers had the most sophisticated modern instruments, they could not have benefited from the vague sounds that might have emanated from the surface.
Having walked the tunnel many times, I was never able to derive directions from various hammerings by colleagues from above, even where the overburden was shallow. And the inscription makes it clear that the tunnelers could only hear each other when they were about 3 cubits (less than 2 meters) apart.
From inside the tunnel it can be seen that the first 80-meter length of tunnel from the north, and the last 150-meter length of the tunnel to the southern pool, which both have uneven ceiling heights, are natural dissolution channels that were widened and deepened by the engineers. They were then left to cut the straight line between these two ends. They did this by means of a rather primitive form of survey, which entailed transferring the line up onto the level ground to the east, by the Israelite City wall, and then transferring it back underground. That was not easy and, thanks to frequent checks and corrections, it took them in and out on a line of 300 meters, while the direct route, if they had achieved it, would have been only about 250 meters.
But it did mean they were able to hone in on a common location where, according to the inscription, they finally used simple sound effects (“axe against axe”) to complete the successful meeting.
[For details see my abbreviated article in Tel Aviv 25, no. 1 (1998), pp. 116-130.]
Stephen G. Rosenberg,
W.F. Albright Institute of Archaeological Research
Ayreh Shimron responds:
Thank you for your comments.
I have acknowledged Shaheen’s (1979) contribution to our work in the first publication pertaining to Hezekiah’s Tunnel published in New Studies On Jerusalem (Proceedings of the Fourth Conference, December 10, 1998). I am not aware of any consensus that “unless the tunnelers had the most sophisticated modern instruments, they could not have benefited from the vague sounds that might have emanated from the surface,” and I would appreciate a reference to such a consensus. Indeed, if this consensus is of equivalent value as the statement in your publication (above) that “karstic activity must by definition be random,” then I cannot take your statement seriously. Karstic activity is anything but random, and karstic tunnels are without exception controlled by sedimentary layering (bedding contact) or a fracture or system of tectonic fractures. As I mentioned, already in our 1998 publication (above), we had tested the hypothesis of sound communication and found it to be a valid and powerful tool in the Hezekiah tunnel scenario; no other explanation justifies the swings in both the northern and southern tunnel segments towards the east and shallow topographic levels.
Of course, acoustic communication is not precise as the source of sound emanating from the surface and tunnel cannot be pinpointed, which at least partly explains the sinuosity and numerous “correction niches” along the route. I suggest you put a surface team with a good geological hammer at the columbarium above the tunnel and you stand in the tunnel beneath the columbarium (also with a good geological hammer) and I promise you that you will discover that sound between surface and tunnel carries well and very much beyond the 2 meters mentioned in the Siloam inscription. The latter incidentally refers to the voice of the tunnelers that was heard when they were 1.35 meters apart and not the sound of hammers—a significant difference.
You write that “It can be seen that the first 80 meters from the north, and the last 150 meters to the southern pool, which both have uneven ceiling heights, are natural dissolution channels that were widened and deepened by the engineers.” I don’t know by whom this fact can be seen, but this statement is nonsense. A brief glance at the figure of joints and fractures in the tunnel reveals that all of these trend at a high angle to the trend of the tunnel, thus the tunnel cannot be a natural dissolution feature as there is no fracture or fractures which parallel the tunnel. It is probable that the first 6 meters, that is, from the southern pool until the site of the inscription, is a dissolution channel and the proof can be seen just opposite the inscription site, where a conspicuous fracture veers off towards the north-northwest while the tunnel swings toward the north-northeast. This fracture, the Karstic Fracture in our 1998 publication (Figs. 1 and 2, referenced above), was probably the starting point for the excavation from the south. Along this still-active geological feature, water has been percolating downward, depositing flowstone along its walls since ancient times. Tunnel engineers, seeing that the fracture contained water, probably hoped that it would lead them directly to the Gihon spring in the north. Alas, the idea was probably dropped when the fracture veered toward the north-northwest, not in the direction of the spring and the northern excavating team. Regarding the rather primitive form of survey that you suggest, which entailed “transferring the line up onto the level ground to the east, by the Israelite City wall, and then transferring it back underground”, I am afraid that I just can’t see it.
The following three letters raise similar questions, and have therefore been addressed by Dr. Shimron in a single response at the end.
I found the article “Sound Proof: How Hezekiah’s Tunnelers Met” quite an interesting overview. It did raise one question in my mind, however: If the tunneling was begun from opposite sides of the hill, what would have been the method available at that time to gauge the elevation at which to start from the Siloam Pool side? I can understand that water running downhill would be a guide as to elevation for the tunnelers on the Gihon spring side. But if they had no means of establishing the elevation on the Siloam side, then there would be a huge potential for much wasted effort from that side, as both tunnels needed to align horizontally and slope down towards the pool. In any event, that they started from both sides, rather than taking the easier route of tunneling downhill from the Gihon side only, shows how desperate they were to finish this quickly in the light of Sennacherib’s threat as described in the Book of Isaiah.
Kevin N. Daniel
Nowhere in the article about the construction of Hezekiah’s tunnel do you explain how the engineers, working from both ends, managed to keep the tunnel floor sloping at the exact angle needed to enable water from the Gihon Spring to constantly flow gently downward for 1,700 feet to the Siloam Pool. In fact, the contours shown on the map seem to indicate that the spring is actually at a lower elevation than the pool, which would require the water to flow upward! What light can you shed on this?
Robert M. Delancy
As a longtime retired civil engineer, I found the article on Hezekiah’s Water Supply Tunnel by Hershel Shanks particularly interesting. I kept looking to learn of the respective elevations or differences of the intake source (Gihon Spring) and the outfall (Siloam’s Pool). As I vaguely recall from my hydraulics course and my water supply course, that’s one of several factors critical to the flow of any kind of liquid. A few others factors are coefficient of the roughness of the surface over which the liquid flows, the shape of the conduit and the distance of flow.
I looked carefully at the illustrations on pages 52 and 57 depicting the location of the spring, the pool and contour lines of unmarked elevations. Now, I recognize that these illustrations are less than accurate in order for one to fully appreciate the exact locations and respective elevations. But nevertheless, the contour lines shown in the illustrations indicate that the spring is considerably lower in elevation than the pool. Without a pump, water simply does not run uphill. And so I am asking: “What are the respective elevations?”
Silver Spring, Maryland
Although the tunnel engineers had many problems to deal with (a possible emergency situation was one of them), they also had a few things going for them. For example, simple oil levels, and a very efficient surveying instrument called a gromar that was used for surveying straight lines and simple angles, were already available at the time. The usage of the gromar is implied by the many 45 and 90 degree angles cut during execution of the tunnel.
The elevation for the start of Hezekiah’s tunnel in the north segment was the nearby Gihon spring. Using the combination of an oil level and occasionally allowing waters from the spring to seep through would have defined the elevation and grade for the northern tunnel segment. The elevation for the concluding southern meters of the tunnel and the catchment basin—that is, the Siloam Pool—was a little more difficult, but not impossible, to define.
The keys to understanding this segment are, in my opinion, the excessive height of the southernmost 50 meters the tunnel and the exit of Canaanite Channel II. The north-south trending Channel II, also referred to as the Siloam Tunnel, is a near-surface channel that passes into a tunnel, which was constructed by the Canaanites during the Middle Bronze period as part of their waterworks around the Gihon spring. Hezekiah’s tunnel runs nearby to the west and about 3 meters beneath this channel along most of its route, with the exception of the initial and final segments. The channel begins above the Gihon Spring at an elevation of approximately 681 meters. That is about 3 meters above the spring (blocking walls raised the water to a level so it could flow along this channel), which erupts at an elevation of about 678 meters almost directly beneath this channel’s entrance. The channel exits at an elevation of approximately 676 meters, about 253 meters due south and about 100 meters east of the present remains of the Siloam pool (the exit of Hezekiah’s tunnel), at an elevation of 673 meters. Knowing both the location of the exit of Channel II and the desired catchment location for the waters from Hezekiah’s tunnel, it was a simple matter to survey in the location and approximate elevation (which had to be at least 3 meters lower then the exit of Cannel II) of the future Siloam pool and starting point for the southern segment of Hezekiah’s tunnel.
The start of excavation was along the present ceiling and well-defined karstic fracture. The ceiling was then progressively lowered to reach the desired elevation, some 5 meters beneath the exit of Channel II. Unfortunately, the lovely fracture along which water seeped to the surface had to be abandoned since it veered to the north-northwest, which was not the direction desired by the excavators.
In summary, Canaanite Channel II was probably at least partly instrumental in pinpointing the elevation and location of the Siloam pool as well as the starting point of excavation for the southern segment of Hezekiah’s tunnel. Since the tunnel passes directly beneath this channel at two critical locations, the Round Tower columbarium and the southern Shaft to Surface (see our Fig. 2 in New Studies on Jerusalem, Proceedings of the Fourth Conference, 1998), it must have also played an important role during the excavation of some segments of the tunnel. Yet, all things considered, it is quite incredible how the two teams managed to meet almost head-on, at virtually identical elevations as evidenced in the very small difference in ceiling elevation at the meeting point.
I have read many articles over the years relating to Hezekiah’s Tunnel. Your article “Sound Proof” in the September/October 2008 issue renewed my interest.
The one thing I have never seen addressed in any of the articles I have read is the need to vent these tunnels. This is especially true if the openings are to be kept secret. I remember reading about the construction of one of the tunnels under the East River in New York City. One of the main problems engineers faced was venting. They solved that problem by not going in a direct line but by forming it in the shape of an S.
The people living in Hezekiah’s time were quite familiar with caves and the like. Air rushing in and out as pressures change can sound like a vacuum cleaner. Also, bats flying in and out give the openings away, so keeping these openings secret would be a major problem.
Proper venting may have been one of the reasons for the shape of the tunnel, as in the case of the East River tunnel in New York. Also, the shaft would help in the venting.
Perhaps the builders did not even consider the venting and just followed the most convenient way to go. But then we must consider the one who created the most convenient way to go.
Richard A. Hollmann
Ronkonkoma, New York
The bedrock around Jerusalem is limestone, a lot of which is karstic, which means that it contains many openings caused by the dissolution of the limestone. These are manifested as small and large caves, occasionally connected by narrow channels. In addition to such dissolution features, the mountains in the area are cut by myriad fractures, joints and faults, most of them linked to the major faulting along the proximal Jordan-Dead Sea rift valley. Hezekiah’s tunnel is cut by many such fractures in addition to the presence of karstic voids. Undoubtedly these, and others such as the Shaft to Surface and Warren’s Shaft, acted as passage-ways for the access of air during tunnel excavation. Nonetheless, it probably was not easy going. I doubt, however, that the “S”-shape of the tunnel had anything to do with air access or control in this case, although the idea is very interesting.
Your article titled “Sound Proof: How Hezekiah’s Tunnelers Met” was of much interest. You write very well!
However, I disagree with the theory presented. Your article attempts to prove that the cutters of Hezekiah’s Tunnel “were guided by communications from the surface, that is, by hammering on the bedrock above.” This is questionable. Imagine a tunneler deep in bedrock: He certainly can hear knockings, but can he detect direction? Unlikely. I understand that the originators of the concept came to their conclusion following a test on the completed tunnel. This is no guarantee that a tunneler faced with bedrock can detect direction to cut by knockings on the surface.
Also, how does the surface worker know exactly where the cutter is positioned in order to know where to knock? If knocking from the underground to the outside is used today to locate persons trapped in mines or earthquake collapses, it undoubtedly requires special instruments more sensitive than the human ear, which certainly were not available in Hezekiah’s day.
Also, I would suggest that it is likely that the tufa on the floor of a natural crack was simply cut away as the tunnel floor was created and therefore was not found in the drill cores.
One cannot ignore the essential need for oxygen for the workers and their torches while cutting deep inside the rock. Vincent tells how the workers could clear debris from Channel 2 for only a limited period due to lack of air. As is known, there is only one shaft (not manmade) open to the surface along the entire length of the tunnel. This alone should be enough proof that some kind of continuous crack was vital. This can also explain the strange shape wherein the originators “… find the change in direction taken by the two teams as somewhat puzzling.” Enough said.
As a matter of fact yes, he can detect direction, perhaps not the precise point, but certainly the direction of sound. And I don’t have to imagine; I worked underground in a number of mines many years ago. Regarding the “natural crack,” I have tried to show you that natural cracks (that is, geological fractures and joints) are very visible features. I have mapped them quite carefully—all of them in fact—along the full length of Hezekiah’s tunnel, and they are shown as those tiny little lines on the map included in Hershel Shanks’s article. Even a cursory glance of the sketch will reveal that most of these little lines trend roughly north-south, many west-northwest to east-southeast, and some even east-west, collectively almost all at a high angle to the trend of the tunnel. Although there are, indeed, a few fractures which parallel the tunnel walls (and the excavators tried to make use of such fractures), they were only local features during the execution of the tunnel. In addition, there is most certainly no continuous (or discontinuous) crack of any sort, natural or unnatural, along Hezekiah’s tunnel. This is unequivocal. I have tried to show and explain to you these very simple, and oh-so-plainly visible, geological features years ago, Mr. Abells. It’s quite clear that I failed totally in my efforts. Enough said indeed.
Sign up to receive our email newsletter and never miss an update.