The goals of the Hora Apotheke Reorganization Project (HARP) are to reorganize, conserve, restudy and publish material from Carl Blegen's excavations at the Palace of Nestor in Pylos. The 2001 season was funded entirely by the Classics Department of the University of Cincinnati. In 2001, Stocker and Erin Lope of the University of Cincinnati spent 10 days transferring the excavated material that is stored in apotheke 2 from cardboard boxes into the new wooden containers that were manufactured with funds provided by INSTAP in August of 2000. All information from each cardboard container was transcribed directly onto the wooden box in permanent magic marker and entered into the HARP database. Well over half of the material was curated in this way. Much of this season's work focused on the back portion of Storeroom (Apotheke) #2 where there still appears to be considerable water seepage. We were able to rebox most of the material from the "pantries" of the Palace of Nestor, i.e., Rooms 18-20. Our plan is to finish the transfer of the remaining ceramic finds from Blegen's excavations that are still in cardboard cartons to these wooden boxes in the summer of 2002. In 2001 the following material was transferred:
New Room 18
For several days (9/6-9/14) Stocker and Davis examined ceramics from deposits excavated by Blegen, from which the bones currently being examined by Paul Halstead and Valasia Isaakidou, were retained. Our objective in doing so was to clarify the dates and contextual associations of each of these, in preparation for a full publication of the faunal remains from the Palace of Nestor. More than one-half of the relevant ceramic deposits were examined. It is anticipated that the entire task can be completed in one more study season in 2002.
Much of the material was very fragile, and had suffered more or less severe fragmentation and erosion during excavation, while attempts to join fresh breaks suggested very incomplete recovery during excavation. Some of the most intensively calcined material was also severely distorted. As a result of such partial retrieval and survival, these groups posed serious problems of both identification and quantification, but it remains clear that they are the remains of highly structures and significant deposits.
Five of the groups were apparently recovered (judging from adhering traces) from an earth matrix and so had evidently been deposited in their archaeological findspots after burning. The sixth group was apparently recovered from a grey-black matrix of what may have been 'burnt earth', and so may not have been redeposited after burning, although no associated fragments of fuel (e.g. wood charcoal) were observed. Most/each of these groups was 'contaminated' by only a few fragments of unburnt bone, which clearly differed from the burnt material in terms of anatomical and/or taxonomic composition (see below). The burnt material, therefore, had perhaps been deposited with some care, rather than being discarded with mixed refuse, and had also remained more or less undisturbed thereafter until excavation in the mid-20c AD.
All six groups displayed greater or lesser heterogeneity in the colour and texture of bone, suggesting exposure to a range of temperatures (as might be found between the core and periphery of an open fire). On the other hand, the groups displayed distinctively different ranges of colors and textures, suggesting that each group (with one possible exception) was derived from a different and, probably, single fire.
Conditions for the observation of butchery traces were far from ideal: the surfaces of the most lightly burnt bone (and of the 'contaminating' unburnt fragments) were heavily pitted by soil acidity, while the more heavily burnt bone was often subject to such severe cracking that cut marks could not be distinguished with confidence. Nonetheless, many of the bones bear clear traces of knife marks consistent with either dismembering or filleting, suggesting that some of the material, at least, had been stripped of meat before burning. In most cases, surviving traces of earth matrix plainly adhered directly to the external bone surface and close examination found no intervening charred organic matter (except in a few cases where two or more bones had fused together during burning). It is inferred, therefore, that most, and probably all, of the bones had been stripped of meat before burning.
Although the bone was more or less heavily fragmented, much of the breakage was evidently fresh (post-excavation) and many old breaks displayed a prismatic edge suggesting fracturing after burning. Deep chop marks or fracture patterns characteristic of deliberate breakage of the bone, for example in marrow extraction (cf. Binford 1981), were not observed and, in one bone group, one near-complete bovine humerus (including both proximal and distal epiphyses and much of the intervening shaft) could be reconstructed from the available fragments. In addition, (almost) no convincing traces of gnawing were observed, while body parts vulnerable to carnivore attrition (the articular ends of proximal and distal femur and, especially, proximal humerus) are well represented. It seems likely, therefore, that this material was subjected to burning as whole, bare bones, without any prior attempt by humans or dogs to extract within-bone nutrients.
The burnt bone is (almost) exclusively of cattle and is also highly selective in terms of anatomical composition: humerus, femur and mandible (with rare fragments). In this respect, preliminary scanning of the, as yet unstudied, mass of unburnt bone from the Palace of Nestor, which includes a balanced mixture of body parts of the common domestic animals, underlines the highly structured nature of the burnt bone groups.
Despite the fragmented state of the bone groups, there is fairly abundant information on age at death from the state of epiphyseal fusion of humeri and femora. All distal humeri appear to be fused, suggesting deaths after several months, although a few small-sized specimens (details) indicate some deaths among juveniles, well below adult body size. The epiphyses of proximal humerus, proximal femur and distal femur include fused, fusing and unfused specimens, implying deaths after, during and before, respectively, the achievement of skeletal maturity, conventionally attributed to an age of ca. 3-4 years. Most unfused specimens appear to be more or less full-sized, however, and so almost all the animals represented were probably slaughtered on the threshold of adulthood or later. In no cases, were the bony growths typical of elderly animals observed, nor were any examples noted of eburnation of the proximal femur, as may occur in animals used for traction.
Unfortunately, burning of the mandibles has totally destroyed the crowns of almost all teeth, but in several mandibles the third molar had evidently erupted, implying an age of 2 years or more, and in some cases the development of the roots of the third molar suggested a relatively advanced state of eruption and wear, hinting that the adult deaths may have included some fairly mature adults. There was no mandibular evidence (e.g. roots or sockets of deciduous teeth, mandibles with unerupted third molar) for animals dying before 2 years of age.
Determination of the sex of the cattle represented in these deposits is extremely difficult, as metrical analysis is impeded by fragmentation and by the shrinkage and distortion often resulting from severe burning. Nonetheless, several of the distal humeri (the best preserved body part) seem rather large, given their later Bronze Age date and burnt state, and so might plausibly be attributed to male cattle. More tentatively, it was noted that (apart from a few obviously juvenile specimens) the burnt deposits appear not to represent such a wide range of cattle sizes as some of the scanned, unburnt deposits, raising the possibility that most or all of the burnt bones might be derived from males.
The deliberate burning of defleshed bones, followed by their apparently careful redeposition elsewhere, suggests important ritual acts and interpretation in terms of burnt animal sacrifice would be tempting even without the later Homeric accounts of burnt offerings to the gods of thigh bones wrapped in fat. The significance of these rituals is made clear by the contrast between these highly structured groups of burnt cow bones and the very mixed assemblages of unburnt bone from other parts of the Palace of Nestor. This importance is further underlined by the scale of the sacrifice. It has been argued that each burnt offering included the remains of several individuals and so it is also quite plausible that several animals had been slaughtered at the same time; certainly, the absence of traces of gnawing is consistent with the burning of fresh bones. The traces of dismembering and filleting also suggest that the burnt offerings were accompanied or preceded by feasting on substantial quantities of meat, an activity for which there are abundant hints in the Linear B texts from Pylos (Killen 1994).
The burnt bone groups from the Palace of Nestor thus confirm Linear B evidence for palace-sponsored consumption of livestock in both feasts and religious offerings and suggests that such offerings broadly followed the pattern described in the Iliad and Odyssey. The body parts represented in these groups do not precisely match the Homeric description and no interpretation is offered here for the symbolism of the recurrent sacrifice of mandible, humerus and femur.
This analysis also reveals much which cannot be gleaned from the Linear B texts. First, it underlines the particular importance of cattle (perhaps male cattle) in palatial feasting and sacrifice. Such an emphasis is perhaps implied by contemporary iconography, but is not clear in Linear B records, which detail the mobilization of several species for such purposes. Secondly, it suggests that the palace itself was a major focus of feasting and sacrifice, which again is not apparent in the texts, although implied by a range of other archaeological evidence.
Thirdly, it has been tentatively suggested that the cattle slaughtered for these sacrificial feasts were mostly adults or young adults and may primarily have been males. Some at least of these animals seem to have been rather young to represent discarded plough oxen (or breeding females) and no pathological evidence for elderly or traction animals was noted. On the other hand, given their age, they will have consumed both pasture and herding labour for several years prior to slaughter and thus represent significant 'sacrifices' in an economic sense. How they were reared is as yet a matter for speculation. In 20c AD Central Macedonia, young bulls donated to the church for slaughter at the kourbá ni festival of the patron saint were allowed to graze freely on local fields and pastures, while concurrently serving as stud animals for villagers' cows. As they were also allowed to raid cereal crops, these bulls grew quite large, as seems to have been the case with the cattle sacrificed at the Palace of Nestor. The scenes of of bull netting shown in the Vapheio cup might represent the capture of bulls allowed to roam freely in this way. Alternatively, animal reared for palatial feasts and sacrifices may have been herded singly or in small numbers, to ensure access to restricted patches of rich pasture, for example on field edges or along streams. The latter scenario might provide a basis for explaining the large numbers of apparently unemployed ox-herds listed in the Linear B texts from Pylos, although the texts also imply that sacrificial bulls were acquired from outside the palace's own livestock and so it is not clear why herders assigned to their management should appear in palatial records.
In 2000, after completing analysis of the six, now widely reported deposits of sacrificially burnt cattle bone, we began study of a large group of bones from 'Belvedere (W13) from street between walls'. This group was selected for study because it appeared to be dominated by lower limb bones of cattle and thus to represent some form of specialized dump, anatomically and perhaps functionally complementary to the burnt bone groups. In 2001, we completed study of this group and of four further groups from roughly the same area:
We found numerous joins (both fresh and old breaks) between 'Belvedere (W13) from street between walls' and 'MB Belvedere Box 1 animal bones', and also one join between the former and 'MB Belvedere W13 box 3 from street, drain and miscellaneous areas'. It now seems that these groups are not biased to lower limb bones, but rather that the lower limb bones of cattle are less fragmented than the upper limbs; these more complete bones had presumably been separated from the more fragmentary upper limb material by Nobis.
Several preliminary observations of some interest can be made, strictly speaking valid only for these MB groups but also partly confirmed by rapid scanning of the remainder of the assemblage in 1998. (1) In contrast to the highly selective composition of the burnt bone groups (mandible, humerus and femur, overwhelmingly of cattle [probably male] with a few deer), these groups contain almost all body parts from a range of domestic species (sheep and pig, cattle, goat, and dog, in probable descending order of abundance), together with smaller quantities of bones of wild animals (red deer, boar, fallow and roe deer, again in probable descending order of abundance). Among Linear B texts from Pylos, a clear contrast is apparent between 'livestock'/herd management records (dominated by sheep) and 'deadstock'/consumption records (with a more balanced mixture of species) (Halstead in press). Encouragingly, the composition of the faunal groups (the remains of meat consumption - below) is closer to that of the deadstock/consumption records. (2) The cattle bones are of very variable size and almost certainly include both male and female specimens (in probable contrast with the burnt bone groups). Some of the older cattle bones (including probable males) bear pathological traces often attributed to the effects of traction. (3) Among the ovicaprid bones, sheep heavily outnumber goats (contrary to the earlier results of Nobis, based on identification to species of only very few ovicaprid specimens). (4) The remains of deer are dominated by red deer (Cervus elaphus), but also include a few specimens of roe deer (Capreolus capreolus) and almost certainly fallow deer (Dama dama). This raises the question of which species is/are represented by the Linear B DEER ideogram. In the case of red deer, at least, sample size is large enough to demonstrate that the head and feet are largely unrepresented - suggesting primary butchery at a distant kill site and so compatible with a conventional 'hunting' pattern of exploitation. (5) Although bone surfaces are more or less severely pitted by root and soil etching and, to a lesser extent, by carnivore gnawing, surviving butchery marks (mostly knife marks, with a few possible cleaver marks) document the skinning, dismembering and filleting of carcasses, and at least some cracking of bones probably for marrow extraction. More significantly, despite biases of poor preservation, surviving butchery marks are highly repetitive, perhaps indicating performance by specialist butchers. All butchery seems, not surprisingly, to have been performed with metal tools. Widespread traces of gnawing and, occasionally, of digestion suggest relatively easy access to these bone dumps by animals such as dogs.
We worked for 2.5 weeks in 2001 and were allowed to work every day except one during this time. Progress is slow, however, for three principal reasons: (1) the permitted working day is short and a significant amount of time is lost at the beginning and end of each daily session moving study material in and out of the storeroom; (2) the careful examination of cut marks, which has yielded huge dividends with the burnt bone groups and appears to be productive also with the unburnt material, is very difficult because of poor preservation of bone surfaces; (3) the search for joins between contexts is very time-consuming but seems justified by the potential to shed light on the provenance of bone groups. Our first two study seasons have probably done no more than keep pace with discoveries of additional faunal assemblage during the ongoing reorganization of the storerooms by S. Stocker. On the other hand, work on the storerooms seems now to have reached the stage where significant additions of material are unlikely.
Analysis of the human skeletal material from the Palace of Nestor and associated chamber and tholos tombs was conducted from August 27 through September 15. The objectives of the season were 1) detailed analysis and description of the Tsakalis Tomb E-6 sample, and 2) continued cleaning and preparation of the materials from other Tsakalis tombs.
The analysis of Tsakalis Tomb E-6 was completed. Each specimen was fully described and measured. Individuals A, B and C were associated with the last use of the chamber, and are thought to be contemporary with the height of the palace occupation. The following is a list of the burials:
Tomb E6 contained a minimum of 5 children, 6 adult females, 6 adult males, and two other adults, or a total of 19 individuals --based on crania and dentition. Using the divisions of individuals made by the excavators and incorporating all the available skeletal data, there were 5 children, 10 adult females, 9 adult males, and 4 adults of unknown gender--or a total of 28 individuals.
Analysis of Tsakalis Tomb E8 was begun. According to the excavation description, the tomb was used over an extended period and contained at least 16 recognizable individuals designated Burials A-P. These remains, excavated in 1966, were not previously studied by J.L. Angel. Unfortunately, only portions of Burials A, B, D, G, H and J were found in the Hora Museum. All were fragmentary, and some groupings were commingled remains of many adults. I am doubtful that the skeletal material from this tomb can be sorted as thoroughly as the material from the other Tsakalis tombs. Analysis of A, B and J was completed.
The remaining material from Tomb E-8 consists of a minimum of 3 individuals based on cranial fragments, and potentially many more based on postcranial representation. Again, it should be emphasized that the original numbers described in the excavation report are not represented in the currently known collection.
The individuals from the Tsakalis tombs were mostly young adults, with the oldest individuals probably dying before age 40. This is based on the amount of dental attrition, evaluated in conjunction with the amount of sutural closure. Neither of these age methods is as reliable as other methods based upon skeletal elements that are not preserved in the Palace of Nestor collection. In addition, several individuals exhibit heavier wear on their anterior teeth. This may bias some of the dental wear aging estimates, although differential wear was factored into age estimation when possible.
Most of the individuals were healthy. There was no evidence for osteoarthritis or other age-related changes. Two or three of the adult cranial vaults had expansion of the diploe that may be associated with anemia (although other bone changes characteristic of anemias such as vault porosity and cribra orbitalia were not present), while other individuals have thickened compact bone in some areas of the vault. Many individuals have numerous caries and premortem loss of molars from these infections. Dental hypoplasia, a sign of childhood disease or dietary stress, is commonly seen on both males and females. No particularly severe cases were observed.
In total, thirty-one Tsakalis burials were analyzed in 2001. Nine were analyzed in 2000. When Tomb E8 is finished the entire Tsakalis tomb sample will be completed. In addition to the work on the Tsakalis burials, the individual from the tomb excavated in 1939 was cleaned and prepared. It was possible to reconstruct most of the cranial vault. This individual will be studied in 2002.
The work completed in 2001 involved analysis of approximately 1/4 of the sample housed in the Hora Museum. The work progresses slowly due to both the poor curation and mixing of specimens through secondary burial in the tombs. I estimate that a minimum of three months will be required to complete this analytical phase of the project. Following that, I will 1) analyze the skeletal data in conjunction with the archaeological data on burial treatment, and 2) compare the Palace of Nestor sample with other Mycenaean and Mediterranean skeletal series. This may involve use of published data or direct study of other collections.
by Hariclia Brecoulaki
Hariclia Brecoulaki, Wiener Laboratory, ASCSA; Université de Paris I Panthé on, Sorbonne
Luigi Musella, Archaeological Museum of Naples
Caroline Zaitoun, Université de Paris X Nanterre
The major objective for the 2001 season’s project was to remove all fresco fragments from the wooden boxes where they had been piled, wrapped in acid newspaper, since the 1960’s (fig. 1-2), and then to place them in the new drawers that were designed last year to ensure better storage of the fresco fragments. This work had a threefold benefit. First, it ensured considerably improved storage conditions for the painted plaster fragments; secondly it was possible for us to control the state of conservation of a representative number of fresco fragments prior to placing them in the new drawers and to further consider restoration possibilities; thirdly, by removing the old boxes from the narrow storage room (apotheke 1), access was made possible to all of the drawers, making every single piece of wall painting attainable; additionally, air circulation and water evaporation are further enhanced, ensuring better preservation of the frescoes by avoiding damage risks, mainly due to moisture.
The environmental conditions in apotheke 1 and of the state of preservation of the frescoes compared to last year’s situation has been drastically improved. Damage to the frescoes from the infiltration of moisture was not detected this year and it appears that the problem of excessive moisture in the storeroom has been eliminated due to repairs that prevent rain-water from entering the apotheke. Additionally, signs of excessive dampness detected last year i.e., the presence of dark stains on the pavement on the lowest parts of the walls, the formation of efflorescence, etc., were not visible on the floor and the walls of the apotheke this year. This is principally due to improved air circulation through the removal from the apotheke of various objects (including furniture) that covered most of the floor and thus prevented the evaporation of moisture, and to the proper airing of the space by leaving the windows open for an extended period of time. Furthermore, the new drawers that were installed last year were build as free standing pieces of furniture and space was left behind them for air circulation (if drawers were in direct contact with the walls the moisture exchange between the atmosphere and the walls would not be possible and moisture would accumulate and exacerbate various problems, principally the growth of micro-organisms, which were detected in the old drawers).
Dust, soot, and deposits produced by insects and other animals are other external factors that are responsible for damage to frescoes and can cause darkening and impair the legibility of the painting. The area where the new drawers have been installed remained unaffected by these factors, while the accumulation of dust and deposits from insects were observed in the rest of the apotheke. The overall condition of the frescoes stored in the new drawers is dramatically improved compared to the condition of those in the old drawers, since neither moisture nor dust could affect them. Small holes on the back of the new drawers and the fact that they were not in immediate contact with the walls and the pavement, have enhanced air circulation which is one of the main requisites for maintaining the frescoes in a good state of preservation.
About one quarter of the total number of fresco fragments remaining in the original boxes were examined carefully before being placed in the new drawers. The rest of the fragments were stored in the drawers so that the original order in which they had been placed in the boxes was preserved. They were left wrapped in the newspapers as a temporary storage measure. The work of properly unpacking these bundles will be completed during the 2002 field season and will consist of the same procedures that were used this season. These procedures are as follows:
For a limited number of fragments further operations were carried out that included the gluing together two or three joining pieces of the same fragment (fig. 11-12, 13-14) using UHU Hart (a reversible cellulose nitrate based glue) and the superficial fixing of powdering paint layers using an acrylic emulsion PRIMAL AC33, diluted in distilled water (10%). In addition, a preliminary group of 20 samples was collected during the last day of work in Hora under the supervision of a member of the Ephoreia of Olympia (a copy of the samples taken this year is attached to the report).
The following remarks can be made about the fresco fragments that were unpacked and transferred this season: although the majority of the fragments did not present serious decohesion problems, some of them did exhibit the following types of alteration: a) a small number of clay and gypsum renderings (analysis will show the exact nature of the renderings) were weakened by moisture and mechanical causes (fig. 5-6); b) large pieces where superimposition of intonaco and paint layers are attested, probably due to re-decoration operations of the same walls, presented delamination between different layers; c) in a few fragments intonaco was detached from the arriccio by the action of moisture and in some cases renderings were particularly fragile with color alterations due to fire; d) paint layers in elaborate decoration motifs where impasto effects were attested (fig. 7-8) seem to be lacking a binder considering their powdery state, which is due to the action of humidity depriving it of calcium carbonate which binds the particles together; e) in many cases the paint layer was hardened by the formation of carbonates with the formation of craters and blisters filled with carbonates.
"We went through all four boxes of charcoal and bagged two lots as having possible dendro potential. Most of the rest of the stuff is either Pinus brutia or Pinus halepensis, and a 8-centimeter diameter chunk might have only 5-8 rings preserved. Now that is clearly useless for dendro, but it is excellent for radiocarbon. The timbers cannot have lost more than half a dozen or so rings."
Under the terms of a permit granted to ASCSA (UPPO/GDA/ARC/A3/F2,F64/37233) all samples of wood retained by Blegen were removed from apotheke 2 under the supervision of a representative from the Olympia Ephoria and transported to Athens by Davis and Stocker. They are now being temporarily stored in the repositories of the Wiener Laboratory. Because of the events of 9/11 it was impossible to transport the samples immediately to the United States for analysis at the Dendrochronology Laboratory of Cornell University and for radiocarbon dating. We plan to have these samples transported to the U.S. as soon as conditions once again become such that we can ensure their absolute safety. The wood samples will be returned to Hora once the analyses are completed. The following wood samples were taken to Athens on the authority of the Olympia Ephoreia and their removal from Hora is documented by a letter of transmission, signed by Mr. Anastasios Vasilaras on 7/9/01. A copy is in the files of ASCSA: