Late last summer, I received a message from the guards at Hierakonpolis that a hole had opened up near the Old Kingdom-Second Intermediate period tombs (located behind the Fort), and that I should be prepared. I was certainly intrigued, but without more information, it was a bit hard to know how or what to prepare. Nevertheless, I included an examination of the tomb area in the permission request to the Egyptian authorities, and then filed it away on my "to do" list for the coming season.

Returning to the site in late November 2005, I was vaguely aware that I should investigate the matter, but with so much to do--unpacking, setting up the dig house, organizing the bricks for the Fort, etc.--it slipped my mind until I took my annual tour of inspection around the site to assess its condition. Trudging up the hill to the tombs, there was no further need to ask about it, because covered with sticks and brushwood, in the center of the terrace extending out before the tombs, was an unmistakably large hole. Now, I was not only intrigued, but also very concerned about safety, as we had crossed over this very spot hundreds of time during the course of conserving the decorated tombs in this hill--the tombs of Ny-ankh-Pepy and Horemkhawef--unaware of what lurked below.

To be honest, I had long been skeptical of reports of donkeys or horses stumbling into unknown holes and the amazing discoveries that followed. I had always considered them to be euphemisms, or a less embarrassing way of saying: "You know that pit filled with garbage that we had been ignoring for years...well, golly, it turned out to be something!" However, just the year before, while returning from a brief tour of the tombs, one of these totally unexpected holes appeared when I--not a donkey!--fell into it! My legs dangling in space, I was saved from being swallowed up only by quickly jutting out my elbows. Luckily, my colleagues were able to pull me out, but had I been alone it would have been a much more serious matter. I was so spooked by the event, I had no interest in exploring that hole, and instead sent the workmen to collapse it with long poles, resulting in a frighteningly large cavity more than two meters deep surrounded by loose windblown sand.

Giving the new hole by the tomb a wide berth, I angled my tape measure down through the cover to assess its depth. It was three meters down to bottom; definitely a fall that would hurt. Clearly, I couldn't simply ignore this hole. We'd have to see what it was all about.

Early the next morning, we pulled back the cover and I had a chance to take a good look inside. On three sides were more or less well-cut rock walls of a square underground chamber about 3 by 3 meters in dimensions. The fourth (east) side, however, was made up of perilous piles of tumbled rock receding into the dark distance. A corridor filled with debris, leading down to another chamber at a lower level, could be seen on one side by an unstable pillar of stone, behind which there were evidently more chambers. Testing with a probe indicated that they extended for at least another three meters, but as they were completely choked with rubble from the roof fall, there was no way this labyrinth could be entered safely.

The east wall of the subterranean chamber, left, had been reduced to piles of unstable rubble. A corridor in the darkness, right, led down to a chamber at a lower level and more chambers to the east, but it was simply too dangerous to enter. (Rick Coleman)

In the main chamber, blocks from the collapsed ceiling lay all around. Dust and debris sloped up against the walls to within 20 centimeters of the ceiling, while a wide depression filled in the center. I was confused. Was this chamber an undecorated tomb chapel with a shaft in the center leading to the burial chamber below, or was this a burial chamber for one of the mud-brick chapels along the south side of the terrace above? There was only one way to tell for sure--we'd have to excavate.

And no sooner had I said it, than a tripod, pulley, ropes and buckets miraculously appeared. Clearly, the guards had been planning this all summer and I wasn't the only one who was curious.

Left to right: Setting up the tripod. Down the hole, there was only room for two workmen and a metal bucket. The debris was hauled out of the tomb one bucket at a time by Sayyid Bustawi. (X. Droux)

We began by hugging the west wall, where the ceiling at least looked more stable, staying well away from the east side where the debris seemed to be the only thing holding it up. Measuring out a strip one meter wide along the three meter expanse of the west well, excavations proceeded with Salah Mohamed el Amir, Hamdy Mahallal, and one metal bucket. Once filled, a tap on the side signaled Sayyid Bustawi to haul the bucket up on the pulley, decant the debris into a basket, and send the bucket down again. There wasn't much room down there, so I watched from above, which, frankly, was just fine with me--the view was much better and spending the day with a meter of crumbling sandstone over my head was never my idea of fun.

The upper levels of the debris were remarkably clean, giving us hope that no one had been there before us. About half a meter down we started to find the sort of thing one expects in tombs--bones. But we hadn't expected so many! Disarticulated bones were everywhere, covering every inch of the test strip, mixed with just enough pottery and painted plaster to keep it mildly interesting and provide a provision date of early New Kingdom (ca. 1550 B.C.). With our questions still not answered, more would have to be excavated, but how? It all looked so dangerous. I needed some advice.

Luckily, I was able to convince expedition veteran Joe Majer to come out and lend a hand. Joe has long been involved with construction and I knew I could rely on him to come up with a safe and practical solution. After a thorough investigation, he gave his verdict: it was definitely not going to be easy. We could shore up the roof, but first we would need to create secure footings for any scaffolds and that meant more excavation. While full of potential, this was not the most glamorous task on the site, but Joe was willing to undertake it, with the help of Rick Coleman, a new team member from Australia.

First, they had to supervise the removal of all of those bones. Damp and caked with dirt, each one had to be brushed and then boxed by Rick to await analysis by our physical anthropologist Bernie Dickman, who determined that more than 25 individuals, ranging from older adult to infants, were represented in the scatter. Could they all date from about the same time? Or did we have evidence of a Late Period (Dynasty 26, ca. 650 B.C. and later) reuse of the chamber? This was a question crucial for understanding what we were finding.

Aside from more potsherds, there was nothing of much interest with the bones, and particularly noticeable was the lack of faience beads or amulets that might help with dating. Yet below all the bones, resting on the stratum of shale that was the floor, was a scattering of painted plaster. We could see the blue and red pigments and the impression of the textile on the back, but the pieces were so small and fragile, lifting them individually was impossible. Instead, I employed a trick I learned from our conservator, Richard Jaeschke. Placing a piece of fiberglass paper over the plaster, I then applied several coatings of a 10% solution of a consolidant called Paraloid B72, which is the active ingredient in glues like UHU and Duco Cement, to the paper. Over the course of the next few days we applied more coating, which soaked through the paper, and then we left it to dry. We hoped that the plaster would adhere to the paper just enough for us to lift it up without bringing most of the floor with it. Remarkably, it worked! Gingerly prying the paper and the plaster up, once flipped over and cleaned by Richard, the banded collar of a mummy mask could be seen, far exceeding our original expectations. But could we learn more?

Clockwise: The bones now removed, Joe makes plans in the cluster of painted plaster fragments. Detail of the painted plaster. In an attempt to lift the plaster we covered it with fiberglass paper, which we then coated with consolidant and left to dry. (R. Jaeschke) Success! The plaster adhered to the paper and once cleaned, it revealed a painted beaded collar from a mummy mask.

Moving along the south side of the chamber, but still keeping close to the wall for safety, we found that the deposits were deeper. Beneath the loose upper debris, we first encountered the articulated legs and feet of another body, but just below them, miraculously, we found three complete bowls stacked together. Two were of a type called "flower pots," but really of unknown function, which were quickly made on a wheel. The third vessel was of better quality with a lustrous red coating and a white band at the rim. All three were characteristic of the Late Second Intermediate period and early New Kingdom. Expectations of better preserved deposits below were running high, as just below the bowls were three almost intact burials of small children.

Surprise was tinged with disappointment, when only more disturbed burials, preserved as two pairs of partly articulated legs, were found below. Just as all hope was being abandoned, things began to change markedly. Continuing downward through the debris, first, we uncovered a red object, which in the gloom of the tomb was hard to make out. After directing a sun beam down into the hole with the help of an aluminium foil-on-cardboard mirror, its identity became clear--a statue! It was caked with damp soil, so we immediately turned it over to Richard for conservation, who worked his magic to reveal the upper torso of a limestone statue of a man, his red paint still well preserved. Although headless and legless, it could be dated to the Second Intermediate Period based on style: the strong rendering of the pectorals, the medial line of the stomach muscles and the drawn-in waist are all typical of statues of that date.

The next pass with the trowel revealed something even more exciting--the glint of gold. At the time we weren't sure exactly what it was, but it was definitely part of the head, eyebrow and upper eye of a plaster face covered in gold foil. Later, we discovered this was part of a mummy mask of a type common in the Second Intermediate period and into the early New Kingdom, known as micro-face, so called because the face is so small, ranging is size from 12 to only 4.5 centimeters (4.75 to 1.75 inches). Cast in a separate mold, the plaster face was then incorporated into a helmet-like covering made of thinner plaster or cartonnage (plaster on linen) that was placed over the rest of the head, shoulders and upper chest. Needless to say, this type of covering is incredibly fragile, and often it is only the faces that survive, usually in fragmentary condition, both because of the fragile nature of the material but also because it was customary to cover them in gold leaf, making them attractive to thieves.

Many excavators have commented on how difficult it is to excavate this type of mummy mask, and how easy it is to overlook their tiny traces. To this we can certainly attest. With all the excitement over the gilded fragment, we didn't pay much attention to a rounded lump of white, chalky plaster also found at the same level. Covered in mud, and only six centimeters in diameter, I took it for a bottle stopper until Richard pointed out the face on the other side! Only then did we realized we had two micro-faces, both probably originally covered in gold foil. We could then begin to reconstruct the original appearance of the masks based on what we had recovered and comparison with better preserved examples found elsewhere.

Another plaster face from a mummy mask (left), originally mistaken for a bottle stopper! A remarkably well-preserved micro-face mummy mask (right) found in Nubia (Courtesy Säve-Söderbergh and Troy, 1991. New Kingdom Pharaonic Sites, The Scandinavian Joint Expedition to Sudanese Nubia, vol. 5:3, plate II)

In general, these masks often included a broad collar around the neck and over the shoulder, just like the plaster we had recovered earlier. Over and around the head it was customary to paint a feather, or rishi, pattern, thus explaining the loose fragments with this pattern we had also found scattered in the debris.

The little micro-faces are an amazing piece of artistry, but one must admit that the overall effect when in place on the mummy is a bit strange. Some have suggested that micro-faces were made to better facilitate their transport, especially to Nubia, where a large number of these masks have been found, but this seems unlikely as a general explanation. It has also been proposed that the face was supposed to depict not the person within, but his soul or spirit, what the ancient Egyptians called the ba. The ba, often depicted as a human-headed bird, could leave the tomb and travel, visiting the world of the living and partaking in offerings, but it had to return each night to reunite with the corporeal body. Without this periodic contact the deceased would perish. Perhaps the face of the ba on the mummy mask helped the ba return home or magically insured that the two would always be reunited.

A few more passes with the trowel along the south wall revealed another deposit of pots, and then we reached the floor. It was now easy to see the original entrance to the chamber, cut into the south wall and presumably leading up to one of the mud-brick chapels on the terrace above. The entrance had originally been blocked with a brick wall, but this had obviously fallen down over many occasions, allowing both for the interment of over 32 people in total (16 adults, 11 juveniles, and 5 infants) and also, at least on some occasions, the rifling of earlier burials. Painstaking mending and careful analysis of the pottery, which included at least three ceramic canopic jars painted with a scallop pattern in red to imitate alabaster, revealed no intrusive material from later periods. This was clearly a family tomb that was opened and used often into the early New Kingdom.

Left to right: Master pot mender Xavier Droux on the job. Pottery from the tomb. Canopic jar fragments painted to imitate alabaster.

But if this subterranean room was the burial chamber for a chapel above, why was there a depression in the center? Extending the excavated area to the north we quickly figured that out when we discovered the ragged edge of a large circular hole that had been created by the collapse of the floor into an earlier chamber built below it. It would seem that when the ceiling of the upper chamber collapsed, it hit the floor with such force that it broke through the floor into the lower chamber, done easily enough as only about 20 centimeters of rock separated them. Examination with probes suggests that the lower chamber is about the same size as the upper one, but to clear the debris and investigate more would have made the structure even more unstable and dangerous than it already was. We had always assumed that the hill was honey-combed with tombs and now we knew for sure!

Since the floor of the one chamber had now become the roof of another, without a strong and stable floor on which to place the supporting scaffolds, work in the burial chamber had to stop while we figured out a new safety plan. We are still working on developing it. In the meantime, we covered the hole with cement blocks to prevent anyone from accidentally falling in.

Joe said it wouldn't be easy and it certainly isn't, but for our hard work we have learned many new things. This chance discovery has answered long-standing questions about the exact date and function of the mud-brick chapels, which were originally investigated by Ambrose Lansing in 1934 but not well published. It also provides the only relatively undisturbed collection of pottery from Hierakonpolis with which we can investigate regional differences during the turbulent time of the Second Intermediate period. The continued wealth and importance of the site during this time is also evident. An unscheduled and unexpected facet of our work, it has allowed us to begin to understand the underground history of this remarkable site.

Preliminary plan and cross section of the subterranean tomb chambers. The layout of the tombs on the terrace above is shown in gray.