The ARCE Sphinx Project – A Preliminary Report

The area of the Sphinx is one of the few in the Giza necropolis where the stratigraphic history of the necropolis can still be observed. Moreover, it is a site in which several major periods of Egyptian History intersect in both architectural and depositional stratification.

From the Newsletter of the American Research Center in Egypt, No. 112, 1980, pp 3-33.
Reproduced with the permission of ARCE.

The Sphinx Project of the American Research Center in Egypt grew out of our perception of the stratigraphic opportunities that came to light in the course of the excavations conducted in the Sphinx sanctuary in 1978, under the direction of the chief inspector of the Giza Pyramids, Zahi Hawass. The area of the Sphinx is one of the few in the Giza necropolis where the stratigraphic history of the necropolis can still be observed. Moreover, it is a site in which several major periods of Egyptian History intersect in both architectural and depositional stratification.

Reports of previous excavations, including the 1925-36 clearance of the entire Sphinx complex, are vague as to the stratigraphy encountered as well as in the location of structures that were removed in the course of excavation. More surprisingly, no accurate large-scale plans of the monument and its immediate environs yet exist, despite over 150 years of modern excavation. [1].

With these deficits and opportunities in mind, we approached the Egyptian Antiquities Organization for concession rights to the Sphinx complex, and these were generously granted, in the name of the field director and principal investigator, in January 1979. In November 1979, the Archeological and Research Expeditions Committee of the American Research Center granted official ARCE sponsorship to the project.

In September 1979, the project received the fieldnotes, plans, and 200 photographs compiled by Pierre Lacau during the course of the 1925-36 excavations. These archives, currently under the care of the Centre Vladimir Golenischeff of the Ecole Pratique des Hautes Etudes, Paris, were made available to the project through the generosity of the director of the Centre Golenischeff, Prof. Jean Yoyotte. We have made extensive use of them in planning and interpreting the results of our work. The information they preserve will, we hope, make it possible for us eventually to produce, in effect, the archeological report of the 1925-36 clearance fifty years after the fact. Commensurate with our goals of excavation and recording, the project is working to provide an authoritative assessment of the need and best means for consolidation and conservation of the Sphinx itself. In this regard, we have been fortunate to secure the cooperation of Dr. K. Lal Gauri, director of the Stone Conservation Laboratory, and chairman of the Department of Geology, of the University of Louisville. Dr. Gauri has donated his time and expertise as project geologist since November 1979. We have also been fortunate to receive the collaboration of Dr. Jihan Rigai, of the American University in Cairo, who has donated her time and expertise as project chemist.

The work of the ARCE Sphinx Project would not be possible without the active collaboration and facilities made available to us through the generosity of the Egyptian Antiquities Organization. We wish to express our appreciation, in particular, to its President, H. E. Dr. Shehata Adam Mohammed; and to Dr. Abdel Qader Selim, Director General; Dr. Ibrahim el-Nawawy, former Director General of Pharaonic Antiquities; Dr. Mahmud Abdel Razik, Director of Excavations for Cairo, Giza, and Saqqara; and Mr. Nassef Hassan, Director of the Giza Inspectorate. Special mention should be made of Mr. Zahi Hawass, Chief Inspector of the Giza Pyramids, who has offered to coordinate his own research at the site with our project.


1. 1:100 map of the larger quarry forming the Sphinx “ampitheater”, from S of Khafre’s Valley Temple to the quarry ledge N of the Sphinx cavity.

2. Mapping the structural geology of the site; documenting geological stratification and faults.

3. Reconstructing, as far as is possible, area plans and approximate stratigraphic relations of the site’s various periods from published and archival sources, including the documentation in the Archives of Lacau.

4. A photogrammetric study of the Sphinx statue, principally for complete N, S, and frontal elevations showing attached stonework, natural faults, and contours.

5. Drawings of the Sphinx and its site in section, including complete N?S and longitudinal E?W master profiles.

6. Analysis of the stonework on the statue: tool marks, masonry patterns, composition of the stone and mortar.

7. 1:50 detailed plan of the immediate Sphinx sanctuary (rock cut cavity) showing features in the floor such as holes, cuttings, and faults. Larger scale plans of specific features will be done as necessary to record more important details.

8. Excavation of undisturbed deposits on the site neglected during previous excavation, to salvage remaining stratification. This includes some minor deposits in the Sphinx Temple and the systematic clearance of artificial holes and cuttings in the floor of the Sphinx sanctuary.

9. An assessment of the need for consolidation work of the Sphinx and the steps to be taken for its preservation. Progress The Sphinx Project has been in progress for one year. Personnel includes James Allen, Principal Investigator; Mark Lehner, Field Director; K. Lal Gauri (University of Louisville), geologist; Jihan Ragai (The American University in Cairo), chemist; Christiane Zivie (CNRS), Egyptologist; Ulrich Kapp (DAIK), photogrammeter; Peter Lacovara (University of Chicago) and Cynthia Schartzer (The American University in Cairo), archeologists.

Work began on site 17 June 1979 and continued through July. The project was resumed 12 September 1979 and carried on until 7 February 1980. Dr. Gauri came for the project’s preservation study in April 1980, and work has continued from 6 June 1980 until present.

The highest priority was given in the first season to completing the large-scale architectural recording of the statue itself, since concern about its possible deterioration prompted restoration efforts by the EAO even while our project was in progress – efforts that necessarily change the state of the statue, concealing ancient stonework and part of the parent-rock core. At the same time, additional detailed studies and small-scale excavations were undertaken to sample the range of information on the site and the kind of recording necessary to extract it.

Our project’s objectives fall naturally into five main areas of focus. In order of priority, these are: (1) the statue itself; (2) the assessment of the need and means for its preservation; (3) the sanctuary; (4) the treatment of the undisturbed ancient stratification still remaining (in collaboration with Mr. Zahi Hawass); and (5) the chronology of the larger site, based on our mapping and on the documentation that exists from previous excavations.

1. Study of the Sphinx statue

The statue known as the Great Sphinx is formed of a parent rock core fashioned out of the surrounding gebel, and at least two layers of rectangular stone blocks forming a “skin”: larger stones, varying in size and up to 1 m. in length next to the parent-rock core; and smaller, brick sized stones covering the larger blocks in places.

General elevations of the N, S, and front sides of the Sphinx can be done most expeditiously through photogrammetry. The photogrammetric survey and plotting for these elevations were generously furnished by the German Archaeological Institute in Cairo, whose photogrammeter, Mr. Ulrich Kapp, joined the project on-site through September and October.

The surveying and photography for the photogrammetry were keyed to the project grid, surveyed onto the floor of the Sphinx sanctuary and Sphinx Temple. When the photogrammetry was begun on September 12, the survey grid was rechecked for accuracy. Surveying data and photographs were completed through October. Mr. Kapp has now finished plotting the elevations, which show the contours of the exposed parent-rock of the statue and all stonework attached to the core. N-S cross sections have been produced at 5 m. intervals along the length of the statue. The plotting is done at scale 1:50; Fig. 1 shows, in reduction, the elevation of the front with a section down the center.

Portions of the statue that could not be recorded with photogrammetry were plotted by hand. Fig. 2 includes the top-plan of the forepaws with all observable stonework measured in. This will be locked into a master plan of the statue showing the detail of the head, outline of the back, the masonry boxes attached to the sides of the statue, a stone-by-stone plotting of the base outline, and the surface of the layered stonework that attaches at different levels around the parent-rock core (Fig.3).

Such records will be useful in assessing and documenting future efforts at restoration of the statue. In addition, they reveal much about the building history of the Sphinx. The coating of masonry attached to the core body in ancient times, and now denuded to less than half the height of the existing core on the N and to about three-fourths of the height of the core on the S, shows at least two main periods, indicated by overlapping layers. There is also patchwork inter-phasing, probably of different periods, in the outer layer. In addition to the master sections through the statue at 5 m. intervals, detailed sections at 1:20 are drawn and located on the master plan for studying this stonework (Figs. 3, 4).

A good part of the history of the Sphinx may be read from this added masonry, from the condition of the core under the earliest level, the stratification of the stonework, analysis of tool marks, and possibly through analysis of the mortar bonding the different phases. Small samples of this mortar are collected and keyed to the 1:20 sections (e.g. Figs. 3, 4; mortar samples mE1, mE2, etc.) to preserve their precise provenance. The mortar analysis has been carried out by Dr. Jihan Ragai at the American University in Cairo’s Chemistry Unit, using X-ray diffraction, infrared and emission spectroscopy.

2. Preservation Study

In August 1979, a series of seven samples of loose flakes and efflorescenses from the stone of the parent-rock core of the Sphinx, the Sphinx sanctuary, and the Sphinx Temple were sent to Dr. K. Lal Gauri for analysis in his Stone Conservation Laboratory at the University of Louisville, along with a set of color slides documenting their exact source. Dr. Gauri’s preliminary analysis showed a high accumulation of water-soluble salts at the surface of the stone; through cyclic crystallization, these may be partly responsible for the stone’s deterioration.

In April 1980, Dr. Gauri made an on-site assessment of the Sphinx and carried out systematic sampling. Fig. 5 illustrates a sampling profile at the front of the Sphinx. The geological unit in which the core body has been cut contains several strata. Samples of flaking stone were taken for successive strata to determine the chemical situation at the surface of each, and to test for the hypothesis that rising and evaporating water is bringing soluble salts to the surface. Samples E23, 24, and 25 were taken from the sand bedding, the bottom, and the erosion line of the granite Thutmose IV stela to determine if a similar process is acting on its gradual flaking and deterioration (5 lines of text have been lost since Lepsius’ publication).

Dr. Gauri is presently studying these samples. Stone loss on the Sphinx over the last 50 years can be approximately determined by comparing profiles of the statue as shown in the Arch. Lacau photos with its present state. Dr. Gauri will send his results to several other specialists working in the field of stone consolidation to solicit observations and recommendations. A full report will then be submitted to the Antiquities Organization.

Dr. Gauri has forwarded some preliminary conclusions based on his study thus far:

The field investigation clearly reveals the intimate association of efflorescing salts and decaying stone. The stone is being reduced by continual peeling and every peel of stone has a deposit of whitish salts. This is true of the limestone of the core and veneering blocks as well as of the granite (Thutmose IV) stela and at the Khafre Valley Temple. Furthermore, one can clearly see patches of salt infestation spreading from the sites of bad mortar.

These salts, originally lodged in mortar, in the blocks of the various generations of restoration, and in the parent-rock from which the core of the Sphinx has been carved, have been accumulated by the under-water at and beneath the surface of the stone. Their repeated solution during condensation of moisture at night and crystallization due to evaporation of water during the day, mechanically disintegrate the stone. Also, these salts weaken the stone whereby it becomes more vulnerable to wind erosion. It follows from this that infested mortars should be removed, restoration blocks replaced or cleaned, and the parent-rock flushed of these salts. It also follows that the water table should be mapped and efforts undertaken to lower the water table in the immediate region of the Sphinx. Further, near replacement blocks as well as new mortar must be carefully studied and cleaned of salts if they occur in these materials. Precautions must be taken that new restoration materials are not in contact with infested stone and mortar.

My visual observations of the Sphinx, and its state as seen in the [Arch. Lacau] photographs of 1925-26 suggest that during this period a massive deterioration has not occurred which would warrant immediate emergency treatments. This is not to minimize the obvious seriousness of the problem relating to the occurrence of the salts which, like cancerous growth, are spreading into the neighboring relatively healthy stone.

3. The Study of the Sphinx Sanctuary

One of the immediate needs for recording the Sphinx is an accurate large-scale plan of the sanctuary (the floor area of the cavity in which it was formed). Much information still remains to be extracted through careful study of the many features cut into the parent-rock floor. These take the form of small holes, 15 20 cm. in depth and diameter (designated hl, etc.); linear or rectangular cuttings (c1, etc.); and irregular depressions (d1, etc.). Many contain undisturbed ancient fill that has been left intact by earlier excavations. By carefully mapping each feature and analyzing its fill, we hope to be able to appreciate larger patterns of distribution and to understand the purpose of the features.

For this study, the immediate floor area of the Sphinx was divided into seven sectors: NE, N, NW, SW, S, SE, and E. During June and July 1979, a series of these features in the NE corner of the sanctuary was excavated as a sample (Fig. 7). These occurred under the area labeled R4 in Fig. 6 (which includes the plan of excavation of a slope containing undisturbed deposit done under Zahi Hawass in 1978; the relation of these features with the overlying ancient deposits in this corner will be fully discussed in a separate report in collaboration with Mr. Hawass). Specific groups of features, such as those in Fig. 7, were planned at larger scales of 1:20 and 1:10. The small holes were quadranted, the fill sampled, and the sections drawn at 1:2 (Fig. 8).

C10-21 were filled with a hard compact buff to yellowish gypsum with limestone fragments. Large pieces of chert and quartzite, with one showing traces of copper flecks, were taken from C19-21. Some holes, such as h6 and h5, appeared within this gypsum fill as purplish-grey spots of finely disintegrated granite (“granite dust”). Others were filled with a tan or brown clay, or a grey compound of clay, gypsum, sand, and granite dust; with limestone, granite, and dolerite chip inclusions (Fig. 8). One of the holes (not shown in Fig. 7) occurred at the bottom of a gypsum-filled depression. A dark band of carbon cut through the gypsum fill to the hole, suggesting a kind of post hole, perhaps for scaffolding. It seemed reasonable to expect some kind of pattern to emerge from the positions of the holes, yet this did not prove to be the case with this particular group of features.

In several of the holes sherds have been found, some of which are diagnostic (Fig. 8, h16). In one case the sherds joined to produce the wall of a vessel with a diameter and profile that matched that of the hole. In this case the hole seems to have served as a socket for a crude red-ware (water?) jar, which was later sheared off at the rim of the hole. In h15 (Fig. 8) a carefully laid cache of flint debitage was found, and under this a lens of yellow clay contained dehydrated seeds.

Our tentative conclusion is that these holes were cut and later filled in the Old Kingdom, during construction activity on the Sphinx and Sphinx Temple. The granite dust and dolerite chips found in several of them should derive from finish work on the nearby walls of the Sphinx Temple. The stratigraphic evidence from Mr. Hawass’ excavations (Fig. 6) goes far toward confirming this.

The rectangular cuttings shown in Fig. 7 also likely reflect Old Kingdom building activity. In this area they occur in patterns of groups of three. The profile across any group of three (Fig. 7 upper inset) reveals a pattern of shallow to deep cutting (the individual cuttings range in depth from a few to 50 cm.). All have the wedge-shaped graded bottoms shown for C2 (inset). The fact that they occur under and alongside in-situ core-blocks of the Sphinx Temple (cb 1 and cb 2 in Figs. 6, 7) and core-blocks left where they now lie during construction activities (cb B, cb C), indicates that these cuttings were sockets for large levers used in maneuvering the core-blocks. The width of each group of three is about 2 m. – as wide as, or slightly wider than, the ends of many of the core-blocks that form the walls of the Sphinx Temple: this suggests that three large levers were used at one time under the ends or sides of the core-blocks.

To be effective in moving a multi-tonned block of stone, a wooden lever would have to be about the size of a railroad tie – or about the width of one of these cuttings. Just such a beam may be represented in the lower register of the scene of moving a large colossus in the el-Bersheh tomb of Djehutihotep (Newberry 1893-94, I, PL. XV). Three men are needed to carry a large wooden beam with a series of notches cut into one side (Fig. 7 lower inset). The inscription above this group reads f3t htw n st3 jn …wt: “carrying beams for moving by the group of .…s.” [2] In the case of the Djehutihotep colossus, the technique of the lever and wedge-shaped cuttings may have been necessary for occasionally getting the transport sled over a rough spot on a rock surface. At the Sphinx, these cuttings and similar levers would be necessary for getting the large core-blocks off the floor, and for moving them short distances when it would have been less practical to mount them on rubble, sleds, and/or sleepers.

Similar cuttings are to be seen along the N front of the Great Pyramid and elsewhere at Giza. They were found at the ends of the boat pits at the S side of the Great Pyramid [3] The fact that they are occasional, and not a regular feature to be found associated with all large masonry works, suggests that they were used with large levers as an aid when occasion arose – particularly when large blocks had to be lifted and maneuvered off a rock floor. Clark and Engelbach (1930, 109) seem to refer to such cuttings when they state:

Though here and there holes can be seen in the pavement or in the courses [of the Pyramid] in which levers may have engaged when used as handspikes, or in which the fulcra of the levers may have been anchored, they must be disregarded in the inquiry, since they are exceptional… The most that can be said regarding the occasional holes in the pavements and courses is that they may have been brought into use if a block proved refractory – if something went wrong.

Nevertheless, the patterns of the cuttings, and their association with in situ and maneuvered core-blocks suggests how levers were generally used with the ends of blocks. The same system could have been used when the blocks were on a bed of rubble or sand, or with transport sleds riding over a bedding. In some cases, such as the bottom row of core-blocks forming the S wall of the Sphinx Temple, the notches are cut in reverse on the undersides of the blocks.

4. Excavation of Undisturbed Deposits

Excavations carried out in the NE corner of the Sphinx sanctuary under Mr. Zahi Hawass, Chief Inspector of the Giza Pyramids, in 1978 revealed undisturbed ancient stratification in the slope of debris connecting the 18th-Dynasty Temple of Amenhotep II, the NW corner of the Sphinx Temple, and the remains of Roman stairs in front of the forepaws (Fig. 6). A bottom level, partly sealed by the corner of Amenhotep’s Temple, appears to be construction debris deposited in the Old Kingdom, judging by the ceramic evidence. This deposit lies on the uppermost of the two terraces that form the Sphinx complex.

In order to add to the stratigraphic picture revealed in the course of Mr. Hawass’ project, our project excavated a small trench through a mound of ancient debris left by previous excavations under a toppled core-block leaning against the N wall of the Sphinx Temple in the colonnade. This offered some idea of the lowest levels of ancient fill, mostly removed by the massive clearing of Baraize in 1925-36. In contrast to the 1978 excavations, this deposit is lying on the lower terrace and inside the temple walls.

Our trench, designated R16 in conformity with the sequence from 1978, revealed that this small remaining deposit was stratified. Fig. 9 is the main section obtained from the trench, which ran to the limestone core-blocks of the N wall. Level 3b, on which the toppled core-block rests, is a packed humidified sand, which may signify a period of abandonment and drift. Levels 4-4e likely resulted from the robbing of the temple: these all contained many fragments of granite and alabaster, the costly stone with which the temple was finished. The bottom rubble level, 4d-e, contained several fragments of the corners of granite blocks. Level 6 stood out in the section as a pale blue band of concentrated, finely disintegrated granite (“granite dust”). This was likely deposited when the granite sheathing had been placed against the limestone core of the N wall and afterward finished and polished smooth on its outer face, the granite dust collecting in the slope leading down to the level socket for the base of the granite sheathing. Over this granite dust, the alabaster flooring was laid up to the granite sheathing of the wall. When the granite sheathing was robbed, the blocks were pulled out down to their socket cut into the floor. The alabaster floor slabs were likewise removed, leaving only a thin band of mortar (level 5) and the granite dust (level 6) as a stratigraphic note that the temple had been completed on the interior and later denuded. Unfortunately, diagnostic sherds from the trench totaled only a handful of very small pieces, giving little secure indication of the period of robbing.

5. Chronology: Observations and Hypotheses

During his study of the Sphinx in April 1980, Dr. Gauri offered some tentative observations of the structural geology of the site that appear to be important for an understanding of its quarrying and construction. We are now hoping to have Dr. Gauri back for an in-depth study of some of these aspects: mapping the fabric of the Sphinx, plotting faults, and identifying the significant, geological units. Until a study is done, the possible insights his observations generated must be considered as tentative.

Dr. Gauri pointed out that the Sphinx complex appears to have been quarried out of three principle geological beds: a hard limestone, in which the head was cut; an intermediate bed of softer limestone in which the core-body was fashioned; and a lower bed, again of harder grey limestone, from which the lower terrace for the Sphinx Temple was formed. We were pleased to find confirmation of this view in Rushdi Said’s Geology of Egypt (1962, 98) where he incidentally cites the Sphinx to exemplify the Eocene succession west of Cairo:

In the pyramids of Gizeh plateau the succession has a 25 m. thick grey to yellowish limestone bed at the base. This is rich in Nummulites gizehensis, Echinolampas africanus, Velates schmiedlii, Ostrea pharaonus, and others. This is followed by a 15 m. thick unit best exposed in the Sphinx ditch in the quarry along the Fayum road. It is dolomitic in places and contains a number of fossils, mainly pelecypods, gastropods, and occasional Nummulites gizehensis. Near the top of this unit there is a thin bed carrying Ostrea reili and Operculina pyramidum. The head and neck of the Sphinx are carved in this latter bed.

The contrast between the head and core-body of the Sphinx strikes even the casual observer. One reason for this is the better state of preservation of the head, probably due to the fact that it was fashioned from a harder limestone. The separation line between this, Bed 1, and Bed 2 of the body is clearly seen in back of the head where a thin slice of Bed I disappears into the top of the back (Fig. 10). The salient characteristic of Bed 2 is a succession of yellow bands, which may be due to limonite, running horizontally through the core-body of the Sphinx. Here, as with the side of the Khafre causeway just to the S, the yellowish bands of the softer layers have eroded much more severely than the intermediate harder layers in Bed 2, leaving a profile of successive rolls or sharp undulations.

The lower terrace of the Sphinx Temple is cut down into the lowest Bed 3, again a hard limestone like the head. The slope of the formation, including all three beds, is downward to the E and S. In the area of the Sphinx, Bed 1 has all but disappeared, with the exception of the statue’s head, due to quarrying. Ancient quarrying has left the surface of Bed 2, the softer intermediate unit, laid bare along its radical slope to the SE behind the Khafre Valley Temple. Bed 2 is seen in section as the upper ledge forming the W limit of the larger Sphinx “amphitheater”, sloping up to the modern paved road. In fact, much of the Old Kingdom necropolis arrayed along the S side of the Khafre causeway, and the causeway itself, appears to have been set into the exposed surface of Bed 2 as it slopes up westward to the foundation of the Khafre pyramid. Within the Sphinx “amphitheater”, on the north side, the surface of Bed 3 has been laid bare sloping down to form the N limit of the immediate Sphinx cavity from the quarried ledge to the N. A good portion of the paws and the base of the statue have been formed from Bed 3 as it slopes down eastward to run through the lower part of the core-body (where it is mostly covered by ancient stonework on the Sphinx) and to disappear in the floor about 10 m. S of the S forepaw. It appears that the ancient Egyptians purposely followed these natural strata as they quarried out the Sphinx complex and foundations for the adjacent Valley Temple, if not the entire necropolis.

Many or most of the massive limestone core-blocks of the Sphinx Temple walls are characterized by a yellow band running horizontally through them. It is clear that some have been extracted along the line of such a soft vein, as they often show the soft yellow limestone running just on their tops or bottoms. Yet an intermediate yellowish band is also seen running through the center of most of the blocks. To a remarkable extent, one can trace the thinning out and thickening of this band, and its undulations from one core-block to another for a considerable length of a series of adjacent core-blocks (Fig. 11). What this suggests is that the blocks were extracted and moved into place in sequence, so that the geological stratification runs remarkably continuous through the blocks as it did before they were extracted from parent rock. The construction of the Sphinx Temple must have been organized so that there was little opportunity for separating adjacent blocks en route from the quarry to the wall.

This geological stratification, running to varying extents continuously through the Sphinx Temple walls, serves as a kind of signature for the source of the blocks and the order of construction. Petrographic and geological study will be needed to confirm what direct observation wants to suggest: that the core-blocks forming the Sphinx Temple walls were mostly extracted from Bed 2, the soft bed with yellowish bands in which the major part of the Sphinx sanctuary and its core are formed. Their provenance might be defined more precisely as the quarry that formed the S side of the Sphinx sanctuary and the face of the higher ledge marking its W end, since this is the largest exposure of Bed 2 in the area before it slopes down to disappear behind the Valley Temple, or phases out over the rise of Bed 3 to the N side of the Sphinx. [4]

If these preliminary hypotheses can be maintained, then the Sphinx Temple and the Sphinx itself must have been formed during the same quarrying and construction process: the core-blocks for the temple walls were taken from the quarry that formed the sanctuary, leaving a core from which the body of the Sphinx was carved. Features in the bedrock floor of the sanctuary may offer a short note of confirmation. The general rough area shown in Fig. 7, in which the wedge-shaped cuttings and holes occur, leads over to the spot where quarry work was abandoned in the N ledge of the sanctuary (Fig. 6, R2). This suggests that some last few core-blocks (from Bed 3 here) were being taken from the cutting of this N ledge, up to the point at which the ledge becomes too thin for extracting core-blocks. To quarry out the Sphinx and to produce core-blocks for its temple in the same process would have been most economical of time, labor, and materials. The question arises, however, as to why the core-blocks for the Sphinx Temple were not taken immediately from the quarrying that formed its lower terrace – the stone there is of better quality – or why the walls of this temple were not entirely fashioned directly from parent rock, as is the case with the lower parts of the walls of the westernmost rooms. One possible answer – which needs further investigation – is that much of the stone extracted from what became the lower terrace was used in the construction of the adjacent Valley Temple. As a working hypothesis, then, we suggest that many of the core-blocks for the Valley Temple were extracted from the quarrying that formed the lower terrace of the Sphinx complex, and upon this terrace the Sphinx Temple was constructed of core-blocks taken from the quarrying that fashioned the Sphinx.

Certainly, a good deal more stone than that taken from the Sphinx cavity and the lower terrace was extracted from this area, since the original plateau rose here at least to the present height of the Sphinx’s head. The core work of the Valley temple also shows some continuous stratification running through adjacent core-blocks, and at least some of these might also be from Bed 2. The hypothesis here applied to the Sphinx complex and Valley Temple must also be set into the quarrying and construction sequence of the entire Khafre complex and the Bauge-schichte of the Giza Necropolis as a whole.

Other questions about the building history of the Sphinx have come up during our recording of the statue itself. To what extent was the Sphinx finished in the Old Kingdom? To what extent was it sculpted from parent rock to begin with? When were the earliest and subsequent layers of stonework added to the core?

It is clear that work was abandoned on the Sphinx sanctuary and temple before the intended plan was completed. Ricke (1970, 38-39) pointed out that the temple was left unfinished and may have never been put into use. Although the interior of the temple was likely finished off with granite sheathing on the walls and alabaster flooring, a number of factors indicate that the exterior of the temple, particularly in the NW corner, was abandoned before completion. In Figs. 6 and 7 core-blocks B and C have likely been left in their present position at the moment work stopped on this part of the temple – their ends have been brought together and cut parallel so that they would match when placed in the core work of the temple wall.

Simple observation of the sanctuary also reveals where quarry work was abandoned. The very straight cutting into parent rock (Bed 3) forming the N side of the N corridor of the Sphinx Temple continues under the foundations of the Amenhotep Temple and passes westward to form the N limit of the Sphinx sanctuary (Fig. 6). It was doubtless intended that this line should be cut straight for the entire perimeter of the sanctuary; yet the cutting of it was abandoned in the NE corner (Fig. 6, R2). The top of the ledge formed by this unfinished cutting contains a series of squarish removal channels, from a few to forty cm. deep, left as the top of the ledge was being worked down.

The S side of the sanctuary is formed by the face of the ascending Khafre causeway. For most of the length of the sanctuary the face has been shaved to a uniform batter (although the softer bands have eroded out). In the rear SW corner, a ramp of parent rock (Bed 3 surface) ascends from the floor level to the lower ledge immediately behind the rump of the Sphinx, indicating where work was abandoned on the S side. The ledge behind the Sphinx – entirely of Bed 3 – has been left completely unfinished as a huge massif of rock jutting out to within a meter or two of the statue.

It is generally supposed that the Sphinx statue was originally completed in the parent rock, which subsequently eroded and was later restored with various-sized blocks of stone. The very poor quality of Bed 2, which comprises the major part of the core-body, however has prompted us to question seriously whether a finished sculpture could have been executed in this rock to began with.

Judging by the way in which the surfaces of successive limestone beds were laid bare and followed by the Old Kingdom quarries in the area, the ancient Egyptians must have had an idea of the strength and quality of the various beds. There are indications that Bed 1, of the Sphinx’s head, had been exploited for the length of the plateau. By the time they arrived at the quarries for the Sphinx complex, the ancient workmen would have had a good idea of the softness of Bed 2, of which the body was formed. As they worked the stone down to form the body and sanctuary of the Sphinx they must have encountered the large fault that runs N-S and cuts the intended back of the Sphinx towards the rear, where it opens to a space a meter or more wide. This fault is clearly shown in Arch. Lacau photos. nos. CI 52 and 55, members of a series showing the condition of the Sphinx when excavated in 1925-6. This fault was recognized as a natural feature by Mariette, [5] who was the first to clear it; after it was blocked up along the sides with cement and limestone blocks, however, others have taken it to be an intrusive tomb shaft. The fault descends vertically through the entirety of the core-body, through Beds 2 and 3, and can be seen running through the causeway and on the floor of the sanctuary on the S side, and again along the floor on the N side where it opens at one point to a depth of 4-5 m. under the leveled floor How did the 4th Dynasty builders deal with this gaping fault in their plan for a finished statue? This and other questions led us to consider that the earliest level of masonry abutting the core-body, a layer of some of the largest blocks added to the Sphinx, may he in fact of Old Kingdom date, so that the body of the statue was finished at that period with the addition of large blocks and limestone-mortar packing. This could have been true at least for the portions of the core formed out of Bed 2. Arch. Lacau photo. no. CI 35 is a close-up of the innermost toe of the N hind paw. A single layer of small (late?) masonry has been denuded for half the height of the toe to reveal the original bedrock (since recovered). The nail or claw has clearly been carved in relief on the face of the parent rock toe, a cutting later closely matched in the added masonry on the lower part. This would seem to indicate that the paw had been finished from parent rock and detail added. This paw is formed of the harder rock of Bed 3 passing here through the lower part of the core-body. On the other hand, Arch. Lacau photo Cl 31 shows the S hind paw as it was being cleared of sand and rubble. It looks as though this paw was almost entirely formed of large blocks of limestone, which were found tumbled about in the sand and rubble. This paw is located about where the hard stone of Bed 3 disappears in a downward slope to the S.

The picture of the origin of the Sphinx that emerges, then is one of work stopped at various stages. At the time work was abandoned, quarrying had formed most of the Sphinx sanctuary and its core-body. The extracted blocks were taken directly to the E for the building of the Sphinx Temple on the lower terrace. Finishing work was completed for the interior of the temple, which was laid with an alabaster floor and granite sheathing. Parts of the exterior walls were still being worked, a last few core-blocks for the NW corner of the temple having been moved from the cutting for the N side of the sanctuary in its NE corner. Rough quarry work was still in progress to form the N line of the sanctuary and the back W end. Meanwhile, the head of the Sphinx was completed in the hard stone of Bed I. The softer core of the body was being packed with limestone and mortar – debris in some cases – and the contours of the statue were finished off to varying degrees by a casing of large slabs of stone added to the core.

There is, however, a serious impediment to this view. Except for the prominent boss on the chest, we have nowhere observed any kind of working marks on the core-body, either in the way of tool marks or of surfaces that would seem to have been left by rough quarrying activity. Neither have we found any profile on the core that would appear to be of finished sculpture. This might easily be explained by saying that the part of the core-body now showing – almost entirely of the very soft Bed 2 stone – has been eroded so badly that all such traces have disappeared. Even so in the cross-sections showing through the successive layers of masonry added to the core, one would expect such traces to show under the earliest level of stonework had it been added soon after the core was formed, thereafter protecting the profile of the parent rock. But on the face and profile of the core in such cases (Figs. 3, 4) there are no observable indications of parts of a finished profile or of working marks. Rather, the profile of the core seems in all cases to be one of severe erosion, leaving the softer yellowish bands and harder intermediate strata showing a profile of successive rolls and undulations. These considerations would seem to indicate that the core-body of the Sphinx was already severely eroded when the earliest level of large-block masonry was added to it.

This idea is supported by features toward the rear of the statue, on the upper part of the N hind flank (Fig. 12) and on the rump, where large boulder-like chunks of the core have nearly fallen only to be arrested by the earliest level of large stones. The most graphic of these features is located on the broad ledge of masonry toward the upper part of the rump. While at first sight it appears that the core of the rump has been cut back in a horizontally-stepped pattern of deep notches, closer examination reveals that the notches have been formed by large pieces of the core separating from fine transverse faults. One such piece, in the form of a large boulder, still lies at an angle of rest after it separated from the core, leaving a space of 25 cm. at the top and 2 cm. at the bottom. Abutting the separated boulder are the large slabs of the earliest masonry, such that it is clear that the masonry was laid after the boulder had separated. It seems necessary to conclude, therefore, that the core-body, of the Sphinx was already in a severe state of erosion when the earliest level of masonry was added, either completing sculpture left unfinished in the Old Kingdom, or restoring the original contours of a statue finished from the parent rock alone.

To seek agreement with known historical facts, we should probably expect the earliest restoration to have been done in the New Kingdom, more specifically in the 18th-Dynasty reigns of Amenhotep II, who built the mudbrick temple of Haroun-Horemakhet facing the head of the Sphinx in the NE corner of the sanctuary, and his son Thutmose IV, who records on his large granite stela erected at the base of the chest between the forepaws, that he excavated this “very great statue of Khepri which rests in this place” from the sands that had filled the sanctuary to the level of the Sphinx’s neck. We hope that a detailed architectural study of the chapel between the forepaws, of which Thutmose’s stela forms the W side, and its articulation with the masonry added to the core, will shed more light on this question. [6]

If the Sphinx was covered to the height of its neck until the reign of Amenhotep II or his son Thutmose IV, it must have been in this state for a considerable period prior to their time. Although there is no information on the cult of the Sphinx proper prior to the New Kingdom, we might hazard the speculation that the temple and sanctuary remained more or less open until the end of the Old Kingdom. We could further speculate that the temple and sanctuary were abandoned and began to fill up with debris and drift sand at some point in the First Intermediate Period or Middle. Kingdom, when, it seems, the Giza Necropolis was largely unattended. It was Ricke’s conclusion that there were two different periods of robbing which affected the Sphinx Temple, the earliest in the reign of Amenemhet I when the alabaster flooring and granite sheathing were ripped from the interior (1970, 25). If we assume that a sand covering would act more to protect than to erode the statue, this leaves less than a millennium, or perhaps half a millennium, for the core to have eroded to the condition shown by the profiles under the earliest added masonry.

In this regard, one other feature might be noted. Simple observation and the scale photogrammetric side elevations of the Sphinx suggest that the tail of the nemes headdress, long missing, should have brought the pleats of the headdress together and ended at a point about 2.5 to 3 m. above the present top of the back behind the head (Fig. 10). The top of the back is fairly flat and regular and shows the phasing out of Bed 1 in which the head was cut. Either two and a half or three meters of the top of the black have disappeared since the origin of the Sphinx – which seems unlikely, since this would have been entirely of hard bed 1 stone, as is the well-preserved head – or the top of the back was intended to have been built up with masonry from the beginning.

Notes * The American Research Center gratefully acknowledges the support of the Edgar Cayce Foundation for the work of the Sphinx Project.

1. The Sphinx Temple was “architecturally recorded and studied to a high standard, with an interpretive, reconstruction, by the Schweizerisches Institut für ägyptische Bauforschung and Altertumskunde in Kairo from 1965-67 under Ricke (1970). Because the principle focus of this study was the temple, the Sphinx itself and its sanctuary, while accurately planned, were published at the small scale of 1:000.

2. Clark and Engelbach (1930) have suggested this beam is a “sleeper” to be laid in front of the sled on which the colossus is being transported. If the notches are deliberate rather than merely indicating the rough side of a beam, they may have functioned in conjunction with wedge-shaped cuttings like those at the Sphinx.

3. Such cuttings and holes were noted as long ago as 1837 in the foundation of the Great Pyramid by Howard Vyse (1840-42, 205-6). Nour, et. al. (1960, 6, Pls. IX.A, XXIV B) recorded cuttings very similar to those at the Sphinx at the ends of the boat pits on the S side of the Great Pyramid and suggested they were for moving blocks of stone.

4. If the portions of the divine beard found in Caviglia’s 1818 excavation at the base of the chest are original to the core of the statue, as Ricke suggested (1970, 33), rather than a New Kingdom addition, this will be obvious in the presence of Bed 1 and Bed 2 strata – a point yet to be investigated.


Mais, malgré son apparence de puits funéraire, ce n’est qu’une fissure agrandie qui va en s’ élargissant, et qui se termine par un vide assez spacieux ménagé précisement dans le plein des cuisses (1882, 95).

6. Reisner (1942, 26), in reference to Baraize’s excavation of the Sphinx, stated:

It was definitely determined that the Sphinx was carved from a natural nodule of rock left by the old Cheops quarrymen, and was originally coated with white plaster and painted. It was also proved that the Sphinx was restored twice, with a coating of small masonry over the whole body, the first time by Thutmose IV… and again by a later king probably in the Ptolemaic Period.

While the head could have been formed from a nodule left by quarrying done under Khufu, there is little evidence for this, and the quarry that formed the core-body and sanctuary was likely begun with the purpose of creating the Sphinx. If Reisner meant to indicate that the Sphinx was coated with white plaster and painted in the old Kingdom, again there is no observable evidence that this is so or that the traces of red paint now observable on the face are original. Rather, the presence of red paint on the second layer of small stonework in many places suggests that the paint on the face, like the outer stonework, was added late in antiquity.

Mark E. Lehner in collaboration
with James P. Allen and
K. Lal Gauri

References Cited

Clark, S. and R. Engelbach
London: Oxford University Press.
Howard Vyse, Richard William Howard
Operations Carried on at the Pyramids of Gizeh in 1837
3 vols. London: J. Fraser. John Weale.
Mariette, A
Le Serapeum-de-Memphis Paris:
F. Viewleg
Newberry, P
El Bersheh:
London: Egypt Exploration Fund.
Nour, M., Z. Iskander, M.Osman, and A. Youssof
The-Cheops Boat. Cairo: Govt. Press.
Reisner, G.

A History of the Giza Necropolis
Cambridge, Mass Harvard University Press
Ricke, H.
Der Harmachistempel des Chefren in Giseh.
Beitrage zur agyptischen Bauforschung­and Altertumskunde 10 1-43
Said, Rushdi
The Geology of Egypt
Amsterdam: Elsevier.


Fig 1 Sphinx-Giza Front View
Fig 2 Plan of the forepaws
Fig 3 Stonework abutting the parent-rock core
on the N front side of the Sphinx
Fig 4 Section 1: stonework casing and rubble fill
at the base of the chest, N side; Section 33:
denuded cross-section to inner side of N forepaw.
Cross-hatched fill is modern reinforcement, large
block at top and masonry at bottom are ancient.
Fig 5 Preservation study: sampling profile at front of the Sphinx.
Fig 6 NE corner of the Sphinx Sanctuary
Fig 7 Detail plan of cuttings and holes in floor of NE corner of
the sanctuary. Top inset: sections of wedge-shaped cuttings.
Lower inset: detail of scene of moving colossus, from
el-Bersheh tomb of Djehutihotep
Fig 8 1:2 stratigraphic sections of bedrock floor holes
Fig 9 Main section from trench in Sphinx Temple
Fig 10Rear view of Sphinx’s head. Thin line just below neck
shows Bed 1 phasing out at top of back; ancient break
of the tail of the Nemes headdress located 3 m. above
present top of back
Fig 11 Sphinx Temple limestone core-blocks forming E niche
Fig 12 Large core boulder at upper N hind flank supported by
large stone slabs of earliest level of masonry added to
the core. These slabs have received a thin cover of
modern cement. Later layer of masonry can be seen
at lower right corner