Zooarchaeology 2006: Recovering animal bone at the house of the high priest Panehsy

Phillipa Payne


Aims of the study

Aims of the study

As part of a long-term study of the faunal remains from Amarna (begun in 1982 by Howard Hecker) an investigation has been started in the immediate environs of the house of Panehsy (who was both High Priest of the Aten and ‘Keeper of the Cattle of the Aten’) which lies to the south of the enclosure wall of the Great Aten Temple. Excavating the house in 1926, Frankfort (1927, 212) noted that:

"Remains of cattle, horns and bones, cropped up everywhere, and it thus seems probable that these rooms served for the preparation of the sacrificial animals, for which Panehsy, as superintendent of the cattle of the Aten, was responsible."

Frankfort did not say what became of the faunal material, making it likely that it had become incorporated into the spoil heaps adjacent to the house. The close association with the temple and its priesthood gives this material sufficient importance to justify an attempt at locating at least a sample of it for proper study.

I undertook a one-day surface survey in March 2004, designed to assess the feasibility of full-scale recovery of the excavated bones from the spoil heaps. Several of the heaps around the temple were sampled to check that the concentration of cattle bones noted by Frankfort was not repeated at other places near to the temple. It was not, the only other significant find of animal remains belonging to dog, the preservation and articulation of the bone (and its discovery almost on the surface of the debris) suggesting it was modern. A further day was spent on the Panehsy heaps in April 2005. In March 2006 permission was granted for a more thorough investigation and ten days were spent on site with the assistance of five local workmen.

The investigation had four main objectives:

As a first step the spoil heaps were surveyed and related to the remaining walls of the house, to produce a composite plan (Figure 1). Four trenches were then dug in sequence into the heaps. Three of them were on the large heap on the north side of the house; the fourth was on top of the shallower heap on the south.


The procedure was to begin each trench by taking off the top 5 cm of the deposit, as possibly modern or wind blown, ensuring that a minimum of modern contaminants were included. This precaution was especially important in the light of the proximity of the dumps to a modern drove road. This uppermost context was retained but is not included in the following results. The context number was changed where there was a discernable change in the composition of the spoil, except in Trench 4 where a single context was followed throughout. Trenches 1 and 4 were the most profitable, and were completely excavated by trowel; trenches 2 and 3 were begun by trowel and completed by touria once the volume of bone had been established as minimal. Trench size was largely dictated by the gradient of the area under excavation. All material was sieved with a 1 cm by 1 cm mesh in an agricultural sieve.

Table showing the recovered faunal material from all 4 trenches


Number of animal bones

Percentage of cattle

Number of bird bones

Number of fish bones

1 161 89% 83 4
2 11 100% 3 0
3 16 5% 13 1
4 570 91% 89 12

The initial two-day investigation in 2004/5 in the vicinity of Panehsy’s house had provided a small sample of which 70% came from cattle. The results from the 2006 season show an even higher proportion, 90% overall. They bear out the general observation of Frankfort that cattle remains were a noteworthy element in the fill of the house. We found few complete horn cores, however: just 13 fragments. There was one significant bovine example with a circumference of 109 millimetres and a total length of 169 millimetres. Its curve was similar to those seen in processions to the temple, for example, in the tomb of Meryra II (Davies 1903, pls XA, XIV), indicative of domestic rather than wild individuals. Some horn also survived which, in conjunction with marks on the bases of cores, suggests the horn-sheaths were also removed at this point. These non-meat products may have had a different destination after butchery (the few pieces that were recovered here had been perhaps deemed unusable or accidentally damaged and so discarded with the bone waste).

Pottery was not retained unless it had been used for burning resin or bore elements of hieratic labels. Small finds included a carved stone fragment showing the tail of a garment and the wearer’s shin, both incised and coloured red; also gold leaf, painted mud plaster and, appropriately, a piece of blue-painted pottery showing a cow. Also recovered were pieces of flint, both worked tools and flakes, which help to corroborate that we are dealing here with butchery remains.


There was a considerable weight (4.7 kilograms) of skull bone and several examples of mandibles with partially complete dental arcades. The whole skeleton was represented, including parts underrepresented in other areas of the site, such as tarsals, carpals and phalanges. In the residential area excavated in 2004/5 (Grid 12), just 1% of the sample was carpal or tarsal bone and only 6% were toes. In the Panehsy collection, although a significantly smaller sample overall, the proportions were much higher: 10% carpal bone and 15% toes. These differences are illustrative of the differences between butchery and consumption/disposal sites.

The Minimum Number of Individuals (MNI) was calculated using the femur as one of the most robust and easily identifiable bones, thus limiting any biases from the preservation or recording stages. The age of the animals was also taken into consideration. Across all four trenches the cattle bones recorded represent a MNI of 36 animals.

Haring has calculated that it would take 119 animals a year to fulfil the cattle requirement for festival presentation at Medinet Habu during the later New Kingdom (Haring 1997, 253). In just five years, on the scale of the Medinet Habu lists, 595 animals should be represented, providing the number of feast days remained the same. To present an ox every day, the figure is a substantial 1820 animals over the same period. With results of just 36 animals is this significant enough to discuss the role of cattle at the Aten temple?
The building belonging to the High Priest was certainly a residence during the occupation of the city, if only part-time, and it is therefore unlikely that vast amounts of bone would be allowed to accumulate over long periods. The area in the vicinity of the living quarters must have been swept out frequently, in order to avoid noxious smell, and to prevent contamination of the area with decomposing organic matter. Therefore, although the MNI seems reasonably low compared to the festival count for Medinet Habu, it is reasonably large if it represents activity over a short period of time, such as a number of days or at most weeks rather than over months or years. We have also excavated under a quarter of the available spoil-heap area. The density was not uniform over the four trenches that we opened, so it is not simply a matter of multiplying this number. However, it might not be too conservative an estimate to consider that the spoil heaps hold the remains of upwards of 100 animals. 

An unpublished fragment from excavation at the butcher’s yard adjacent to the Raneferef complex at Abu Sir records that 13 oxen were butchered daily for the duration of a 10-day festival. Verner comments that this would be the slaughterhouse running at full capacity, providing for the very intensive needs of the festival period (Verner 1986). Unfortunately, the stela that once stood in the Aten temple listing the offerings was smashed in antiquity. Small fragments of uninscribed purple quartzite still lie on the sand, and a few small inscribed pieces were excavated by Pendlebury but remain unpublished (pers. comm. B. Kemp). These tantalisingly show cattle but with no indications of numbers, these parts being too damaged or missing. Similarly when Akhenaten announced the hetacomb for the inauguration of the temple in his early proclamation on boundary stelae M, X and K, he makes no mention of quantity: ‘There was presented a great oblation to the Father, The Aten, consisting of bread, beer, long- and short-horned cattle, calves, fowl, wine, fruits, incense, all sorts of fresh green plants, and everything good in front of the mountain of “Horizon of the Orb” (Murnane and Van Siclen 1993, 37). The typical offering-scene, replicated in many of the tombs at Amarna (e.g. that of Panehsy himself), shows heads and forelimbs of cattle being offered to the Aten on tables. If the remains under study were animals disposed of after they had been offered in the temple we would expect a reasonable bias towards heads and forelimbs.

The elements of the carcass that were present in the sample record that these special parts are not over-represented in the sample. Overall, the sample favours back legs. The Number of Individual Specimens Present (NISP) is 154 from the forelimb and 204 from the back. If we exclude the carpals the ratio is even higher – NISP 82 for the forelimb and 148 for the hind. That there is neither a dearth nor a heavy favouring bias makes it impossible to speculate further.

The main bone that was conspicuous by its absence was the scapula. Only four fragments from an animal large enough to be a bovid were recorded, and none could be confidently assigned to cattle. These four fragments all came from the same context in trench 4. Ikram (1995, 116) notes that when the forelimb is removed it is shown cut away at the scapula. The complete lack of scapula in the north dump and the scarcity of its representation in the south, could suggest that when we see the offerings of forelimbs depicted in tomb scenes, we are actually seeing only the upper part of that limb donated in reality. The remainder of the limb was treated as the rest of the body and only this small token part extracted as the offering accompanying the head. In terms of meat yield this upper forelimb, known in modern terms as the ‘chuck’ down to the ‘clod, is usually sold at a higher price in modern Egypt and is considered an ‘expensive cut’ (Ikram 1995, 125). Illustrations of this part alone being offered are rare, however, as artists usually show the complete forelimb.

It is also possible that the animals were hung-up by the shoulder joint at some point after death, the scapulae being left behind when the carcass was returned to the temple for secondary butchery. The rest of the animal, including the ribs, vertebrae and pelvis are present which makes the supposition that the carcass was moved from site unlikely. This is also attested by the presence of the large number of carpals, tarsals and phalanges recovered. These small bones are easily disarticulated when a carcass is moved; their presence suggests that this was not the case, and that we are looking at the actual site of the butchery.

Once the shoulder blade was removed, the remainder of the carcass was thoroughly utilised. Many of the long-bone fragments had been split along the ‘grain’ of the bone, from the top to the bottom. This kind of fracture signature is widely associated with marrow extraction. The medullary cavity in the diaphysis of the shaft is a rich source of marrow, especially in juvenile animals. This style of damage was recorded on bone belonging to sub-adult and adult animals, but the majority of this bone retained no diagnostic indicators of age at death, being too broken up by the process. It can also indicate a secondary usage for the bone itself. Whilst bone tools are found at Amarna, it is unlikely that they would have a use in the butchering process among the debris of which these bones are found discarded. None of the fractures showed use-wear smoothing or pitting, also suggesting that the marrow or grease of the medullary cavity, not reshaping of the bone itself, was the object of the activity. One piece of worked bone was recorded, but its function as a tool is far from clear (unit [11363].408 from trench 4).

We also recovered four examples of articulated joints:

Navicular/cuboid and first tarsal (context [11363] no. 210 and [11363] no. 241),
Tibia, astragals and calcanius ([11363] no. 488 [11363] no. 515 and [11363] no. 501)
Tibia and astragals [11363] no. 302 and [11363] no. 367)
Tibia, astragals and possibly femur [(11363] no. 303, [11363] no. 356).

These groupings illustrate that the limb was articulated when it was deposited. It is also important from the point of view of reconstructing associations within the debris. As the remains are not found within sealed contexts, the fact that bones from the same animal occur together suggests that they have not been disassociated too widely by the original excavation, and associations in the spoil heap might be conjectured to have been associations in the original deposit.

Painted plaster fragments were recovered in trench 4. The original excavation discovered painted plaster in both the entrance hall and the central room (Pendlebury 1951, 26). The association of this plaster with the remains may point to a possible original location for the deposit but no clear relationship could be established.


These observations have implications for our understanding of the flint blades that were found with some deposits of bone. These tools could have been involved in preparing the meat and discarded when the job was done or they became blunt with use. At the Fifth Dynasty mortuary temple of Raneferef at Abu Sir, flints and bone were found in association within the butchers’ complex. This building is, however, separated from the house of Panehsy by over a thousand years. The utilisation of flint in the preparation of animal carcasses, should leave traces on the bone itself.

Many of the marks on the bones were from heavy tools, such as cleavers or axes, leaving deep V-shaped incisions. A bronze axe might have been the ideal tool (a fragment of bronze axe-head, measuring 2.2 x 1.8 cms, was found in the 1926 excavation of Panehsy’s house: Pendlebury 1951, 26, object 26/80. The object card describes it as possible a model axe-head). There were some examples of damage caused by lighter tools, which may have been made by the kind of flints recovered. These marks are shallow and often over-cut each other as the user inserts the blade into a joint, for example the axis [11363] no. 397, where the atlas has been cut away on the cranial side. Here the blade has been repeatedly dragged over the surface of the bone to attempt to separate the adjoining cervical vertebrae.

This suggests that a large implement was used for the primary dismembering and jointing, but that flints were employed later in the process. Some bones have even been damaged by both kinds of tool although this is unusual in the group overall (for example: axis [11361] 3, from trench 1). It is tempting to suggest that these small blades and scrapers were used in the skinning of the animal and for salvaging useful products like tendons, perhaps in the hands of a second individual. However, among the large number of phalanges recovered none showed signs of damage attributable to the flints, and one specimen had been repeatedly struck with the same heavy tool, as evidenced by the long bones, vertebrae and pelvis. Marks on phalanges (toes) are often associated with skinning (Lyman 1994, 298). That this task was being accomplished with such an unsuitable tool goes some way to explaining the damage done to these parts of the carcass.

Small flint blades would be unsuitable for cutting into large muscle groups; their size might suggest a role in the removal of marrow or possibly subcutaneous fat from the hide. Their presence with the discarded butchered bone suggests that all these activities went on closely together. It is impossible to tell if there was a significant delay between the primary processing of the carcass for meat and hide and then the tasks employing the smaller tools.

Fish bone can suffer from detrimental taphonomic conditions, and we are fortunate at Amarna that this has not been the case. The negligible amount of fish bone recovered is in marked contrast with the amounts that were found in the residential areas, where they imply that fish was virtually a staple food:

Area of Amarna Weight of recovered fish-bone
N49.18    (Ranefer’s House) 76.1 grams
South Suburb (#12) 972.6 grams
Panehsy 2006 1.2 grams

Fish feature very infrequently in offering-lists; their absence here adds further conviction to the supposition that the bones we are dealing with are those of the temple sacrifices rather than simply reflecting the wealthy priest’s menu. This is also true of the small number of pig bones recovered. This animal was reared and consumed at Amarna in daily life (Hecker 1984, 156–8), but once again does not feature in tomb scenes depicting offerings at the city. From these resources alone, the food of the gods would be reconstructed in the formula ‘bread, beer, oxen and fowl’, or, as Akhenaten himself stated, ‘long and short horned cattle, calves and fowl’ (see above). Just 5 definitely identified examples of pig remains were recovered, with a further group of 16 bones possibly belonging to this taxa. 

Most if the caprid remains were identified as sheep/goat; an animal that exists for the convenience of zoo-archaeologists. This is also in keeping with the way that the ancient Egyptians classified these species, referring to them with a single term, conveniently translated as ‘small cattle’. Although some could be recorded with more accuracy, the small reference collection at the dig house does not contain sufficient material to rule out other similar wild animals, such as ibex and desert-dwelling small antelope. The loan of half a gazelle has helped to rule this species out, but it cannot be conclusively stated that all the recorded caprids come from domesticated animals. A depiction of the two youngest princesses with their pet gazelle (Davies 1905, pls XXXVII, XXXVIII; also the antelope on the feeding trough from the North Palace: Newton 1924, 296, pl. XXX.3) even questions the simple divide between domestic and wild by introducing the category of the tame.

So far, results match Akhenaten’s early proclamation and the tomb depictions well: 90% of the recorded mammal bones were from cattle and 78% of the 188 bird bones were of goose. What we see from the marrow extraction and the removal of horn-sheaths, evidenced in this debris, is work going on ‘behind the scenes’. Representations in the tombs show live cattle being led into the temple, and then their flayed hides and decapitated, bound carcasses ready for presentation. This obviously leaves a significant part of the process unrecorded in the art, but it is recorded in the remains of the animals themselves.

Having seen some economy at work in the presentation of the forelimb, might the same have been true of the head? In contrast to Ranefer, the mandible/cranial ratio suggests that the jaws were remaining with the body while the rest of the skull went elsewhere. The fragmentary nature of the ascending and hinge areas of many of the mandibles excludes the possibility of the preservation of marks for removal of the tongue. If this were to be considered the case, it is another example of the butchers at the house of Panehsy retaining as much of the fleshy parts of the animal as possible, while sending only the minimum of heads and shoulders back to the offering-tables. There are no examples of tongue being among the offerings at Amarna, but this does not preclude that it was not consumed. Many of the labels from meat-jars simply state meat is contained within. Occasionally record will be made of the quality, but the cut or even the animal preserved is only rarely noted (Pendlebury 1951, 169–75). Tongue might be just such meat to be treated in this anonymous way.

Brain was probably consumed; another candidate for those unspecified jars. At Malkata, although there was no evidence of splitting the skull, head bones were only found in an area associated with the main palace, not distributed across the site (Ikram 1995, 118). Skull fragments in the South Suburb had generally suffered as the rest of the carcass, suggesting full utilisation of available resources.

The extraction of marrow, fat and possibly tongue and organs at this site necessitates a nearby facility for processing these commodities. Building S40.1 provides just such an ideal location According to the original excavators the building was divided into four areas and a great deal of paving and stone facing had been employed in its construction. This suggested purification to the excavators (Pendlebury 1951, 28) but might also be connected with a task like preparing meat for preservation; a job requiring water on hand for cleaning instruments, hands and meat. Ox bones were discovered in the ‘third division’ (Pendlebury 1951, 29) but no comment was made as to their frequency, whether they exhibited butchery marks or from which part of the animal they originated. Ikram (1995, 103) also notes that the building has no roof and is aligned with a break in the temple wall, arguing for its use in the cattle butchery process.

Here then, not within the temple proper, the actual business of dismemberment took place. Panehsy’s role as ‘Keeper of the Cattle of the Aten’ involved him in the preparation and processing of animals beside the temple as well as possibly within it.


Davies, N. de G. 1903, Rock Tombs of El Amarna. Part I. – The Tomb of Meryra, London: Egypt Exploration Fund.

Davies, N. de G. 1905, The Rock Tombs of El Amarna Part II.– The tombs of Panehesy and Meryra II, London: Egypt Exploration Fund.

Frankfort, H. 1927, ‘Preliminary report on the excavations at Tell el-‘Amarnah, 1926–7.’ Journal of Egyptian Archaeology 13, 209–18.

Harring, B. J. J. 1997, Divine Households: Administrative and economic aspects of the New Kingdom Royal Memorial Temples in Western Thebes, Leiden: Nederlands Instituut voor het Nabije Oosten te Leiden.

Hecker, H. M. 1984, ‘Preliminary report on the faunal remains from the Workmen’s Village.’ In B.J. Kemp, ed., Amarna Reports I, London: Egypt Exploration Society, 154–64.

Ikram, S. 1995, Choice cuts: meat production in ancient Egypt, Leuven: Peeters.

Lyman, R. L. 1994, Vertebrate Taphonomy, Cambridge: University Press.

Murnane, W. J. and Van Siclen III, C. C. 1993 The Boundary Stelae of Akhenaten, London: Kegan Paul International.

Newton, F. G. 1924, ‘Excavations at El-‘Amarnah, 1923–24.’ Journal of Egyptian Archaeology 10, 289–98.

Pendlebury, J. D. S. 1951, The City of Akhenaten III, London: Egypt Exploration Society.

Verner, M. 1986, ‘A slaughterhouse from the Old Kingdom’ MDAIK 42, 184–9.


For the loan of a reference gazelle I am indebted to Dr Salima Ikram at The American Research Center in Cairo.



Website first posted September 2000; last updated October 2017 | enquiries concerning website: email bjk2@cam.ac.uk