THE PYLOS REGIONAL ARCHAEOLOGICAL PROJECT: 14th Season Preliminary Report to the 7th Ephoreia of Prehistoric and Classical Antiquities, Olympia, on the Results of Museum Study, September 2003-October 2004

1.1. Abstract for the 2004 Paleopathology Association Meeting, Durham (UK)

For inclusion in the session “Bioarchaeology in Greece: Current Aspects and Future Perspectives” organized by Chryssi Bourbou

Temporal Perspectives on the Mycenaeans of Pylos: The Palace of Nestor Grave Circle Burials and Neighboring Tombs

Lynne A. Schepartz, University of Cincinnati

Sari Miller-Antonio, California State University Stanislaus

Keywords: Pylos, Greece; Palace of Nestor; Mycenaean; Bronze Age; dental pathology; paleodemography

Mycenaean burials near the Palace of Nestor were excavated by Blegen and colleagues between 1939-1966. They span the pre-palatial and palatial periods, making a diachronic perspective on the health and burial practices of the population possible. The earliest graves, from the ‘so-called’ Grave Circle, date to the end of the Middle and beginning of the Late Bronze Age prior to construction of the Palace. These individuals (N=31), found with prestigious grave goods, probably represent the founding elites of Pylos. The later Pylian Mycenaean population (N=83), from the 14th-13th Centuries BC, is represented by multiple and secondary burials from the Tsakalis, Kondou, and Kokkevis tomb groups.

Few Pylians lived beyond age 35, and there is limited evidence of osteoarthritis or other age-related pathologies. Some individuals display diplöe expansion and marked limb shaft cortical thickening that may have a genetic or stress-response basis. When dental pathology (caries, premortem tooth loss, hypoplasia) is assessed, the Grave Circle individuals differ from the later burials. The overall frequency of dental pathology is greater in the later burials. Females have the highest rates of dental decay and tooth loss--potentially linked to gender specific dietary stresses or the depletion of resources during reproduction.

1.2. Abstract for the 2005 Archaeological Institute of America Meeting in Boston, MA

The Later Mycenaeans of Pylos: Skeletal Biology of the Tsakalis, Kondou and Kokkevis Tomb Populations

Lynne A. Schepartz, University of Cincinnati

Sari Miller-Antonio, California State University Stanislaus

Mycenaean tombs in the vicinity of the Palace of Nestor were excavated by Blegen and his colleagues between 1939 and 1966. The later Mycenaean population (N=83), dating from the 14th-13th Centuries BC, is represented by primary and secondary burials from the Tsakalis, Kondou, and Kokkevis tombs. Aside from brief references in Blegen et al., The Palace of Nestor III (1973), these burials are unpublished.

The importance of the later skeletal sample stems from the direct perspective on the changing lives of Pylians that it provides through comparison between these remains and those from the earlier so-called Grave Circle (N=31). While few Pylians, at any time, appear to have lived beyond age 35, a number of indicators associated with mild infectious disease or slight functional impairment during childhood and early adult life suggest that health was worse for those buried in the later tombs. The differences between the earlier and later burials seem most marked when dentition (caries, premortem tooth loss, hypoplasia) is assessed. The overall frequency of dental pathologies is greater in the later burials, and females have the highest rates of tooth decay and loss-- possibly linked to gender specific dietary stresses or the depletion of nutritional resources during reproduction. Such differences in skeletal biology must be interpreted from the perspective of archaeological evidence for changes in resource base and potential variation in status between the individuals interred in the Grave Circle and those buried in the later tombs.

2.1. Organization and Inventory of Fresco Fragments

Significant progress was achieved this year towards the definitive organization of the wall paintings and their permanent storage. Every single fragment of the wall painting corpus of the palace is now properly stored in newly built drawers that are numbered from 1 to 208. The organization of the material in the drawers is based on their context, with reference to revised room numbers as they appear in Blegen’s final publication. For example, fragments belonging to room 1 are grouped together in drawers from 1 to 7, followed by fragments found between rooms 1 and 2 (see the table attached to this report, where there is a list of drawers).

Although this system of storage required more time because we had to retrieve the maximum amount of information about each fragment in order to identify correctly its provenance correctly, it has proven to be a great help in the re-study of the material and in the relocation of particular fragments. One of the hardest parts of the procedure, however, consisted of the accurate evaluation and cross-checking of information written on Blegen's original wooden tags that accompanied most of the fragments, notes written on paper by M. Lang, and notes from the diaries of various excavators. Very often only the number of the trench was recorded, and not the room in which the trench was situated. Lang sometimes also included in parentheses room numbers assigned at the time of excavation, while in other cases parentheses were used for the new room numbers assigned in the final publication. In addition, many of Lang’s notes on tags had either faded or were badly damaged by insects. Glaubius helped to decipher excavation diaries and in doing so significantly clarified many issues.

An inventory number was written on the back of each fragment. This code will serve as a reference for all future researchers, as well as provide links to a database with information about each fragments and illustrations. Individual fragments were previously unnumbered. Inventory numbers are composed of three numbers. The first number of each fragment corresponds to the number of the drawer that contains it and the second is unique to each fragment within the given drawer. The third number indicates the archaeological context of the wall paintings, when sufficient evidence allowed us to define it; in most cases this number corresponds to the room or area from which the fragments were collected.

It is important to stress, however, that on certain occasions we have found joins between fragments that were found in different rooms. For example, a number of fragments collected on the floor of room 20 (old room 4) belong to the pictorial program of room 6 (the Throne room); a fragment found in room 31, stored together with other fragments from room 27, belonged to the same scene as fragments from the ne fresco dump; a tiny fragment from room 27 joined with a large fragment from outside room 32. In such cases we decided to store joining pieces, or pieces belonging to the same iconographical program, together inside the same drawer; the third part of their inventory number preserves information about the area where they were found.

Such information about joining fragments is of great value for the reconstruction of the original location of the wall paintings in the various rooms of the palace and for studying the way in which the walls that supported them collapsed. For example, in last year’s report, two new joins between fragments from room 20 (old 4) and a fragment from the Throne room (19C6) were mentioned. This year, examination of the entire group of fragments collected on the floor of room 20 (southwest of the Throne room), demonstrated that other fragments also seem to belong to the decorative program of the Throne room (6); these presumably had fallen from the upper storey’s southwest wall.

The practical problems related to attaching an inventory number to the back of the fragments were solved by testing the behaviour of various materials with regards to the compatibility between materials used to attach the number and their permanence. We tried different kinds of "supports" on which inventory numbers could be written, some of these natural (white plaster, dental plaster, calcium carbonate), others synthetic (acrylic stucco); we also experimented with various kinds of permanent markers and pencils of variable hardness. We finally decided on the following procedure:

a. A thin layer of synthetic resin Paraloid B72 (in concentration 10% diluted in acetone) was applied in order to create an isolating film between the original mortar of the artefact and the layer of the "support";

b. After the Paraloid B72 film had dried, a layer of dental plaster was applied with a tiny spatula, so as to create a smooth surface on which an inventory number could suitably be written. The decision to use dental plaster was based both on its better mechanical properties and quicker drying time, in comparison to ordinary white plaster, as well as on its easier manipulation, in comparison to acrylic stucco.

c. After the layer of dental plaster had dried, the inventory number was written on it, using a pencil of 2H hardness. Pencil proved to be much more permanent than any other kind of marker available.

d. Finally, another coat of Paraloid B72 (in concentration 5% diluted in acetone) was applied in order to protect the written surface.

All materials used in this operation are reversible and we tried to constrain the dimensions of the plaster "support" to a minimum. Almost all individuals who participated in this year’s project helped restorers to assign numbers. Numbers were assigned to a total of 129 drawers; these contained more than 8000 fragments. The inventorying of individual fragments is proving to be extremely important for the re-study and restoration of the iconography. Fragments can be more quickly rearranged for study, with immediate reference to their excavated context. The search for joins was greatly facilitated, since numerous fragments with different provenances could be studied without fear of losing vital contextual information.

2.2. Photographical Documentation

The systematic digital recording of the wall painting fragments that started last year was continued following the same process: each drawer with all artefacts was photographed in its entirety; subsequently each unique fragment was recorded and labelled with its inventory number. Julie Hruby, with the assistance of Kori Duncan, Jen Glaubius, and Yuki Furuya, photographed 4250 fragments contained in 67 drawers; these images were then stored in a database.

2.3. Restoration Operations

Significant progress was made this year in restoration of the wall paintings, owing to the participation of four restorers: two students (Fardi and Kottoula) and two professionals (Musella and Kapizionis). In addition to the usual treatments that have been employed for most fragments (superficial brushing; consolidation of flaking pictorial layers or powdery mortars using an acrylic emulsion, Primal AC33, diluted in distilled water 3-7%; reinforcing of the pictorial layer’s cohesion using Paraloid B72, a copolymer of acrylates and methacrlylates diluted in acetone 3%), more advanced operations were performed in some instances as described here:

a) The so-called White Goddess, a well known painting preserving most of the head and neck of a life-size female in left profile, recovered from the North West Slope of the palace and published by Lang in 1969 (49a H nws), has been re-restored. This piece, part of a larger composition, is composed of thirty-three joining fragments that were glued together roughly, probably soon after their removal from the ground; the surface had been cleaned preliminarily, removing only partially a layer of saline incrustation.

All fragments were first unglued. Joining surfaces were then properly cleaned (earth was still attached to them) using purified acetone (99%) and distilled water. In a second stage, each single fragment was examined under a stereomicroscope in order to define areas that needed further cleaning and to document the state of preservation of the pictorial layer. The striking blue background of the composition is made of Egyptian blue (as was indicated through XRF analysis performed in situ); the coarse grain of this sandy pigment is often responsible for its pulverisation when applied in thick layers, although its chemical composition makes it one of the most stable pigments in the ancient world. Consequently, the original colour of the pigment remained unaltered chemically, even though superficial incrustations had created an uneven surface with spots of different shades, and its physical properties had been affected by external conditions, causing the powdering of its surface in many areas. The whitish surface of the Goddess’s face, composed of pure calcite, presented a more homogeneous and compact surface, but here, too, saline incrustations had created grey spots that detracted from a full appreciation of the fine execution of the painting.

We decided not to attempt a more intensive cleaning of the blue background, except in restricted areas, because the risk of removing grains of the original pictorial layer was high. We did, however, begin a meticulous cleaning of the face, by removing the saline incrustations. This operation was conducted mostly chemically, using a mixture composed of distilled water and slightly basic salts (ammonium and sodium bicarbonate), reinforced by surfactants applied with an organic gel (carboxymethyl cellulose) in the form of compresses. This mixture acts as a non-polar system disrupting the molecules of the salts so that solution is then possible. We also used a cationic surfactant (Amber SH), applied with carboxymethyl cellulose. The remaining salts were eliminated mechanically at a second stage using sharp scalpels, always under a microscope (a delicate operation conducted by Fardi). After the cleaning was completed, the precise and clean joins of the fragments were again glued together using an acrylic adhesive HMG product (Paraloid B72). On the reverse of the restored piece, a new lime based mortar was used to enhance adherence of the fragments and to fill cavities.

The benefit of this restoration was twofold: on the one hand, old restoration materials (organic glues and consolidants oxidize with ageing) were removed and fragments were glued correctly, thus reducing mechanical tension between the bonds; on the other hand, we had the opportunity to appreciate significant details in the face of the woman that revealed previously hidden aspects of the original pictorial technique.

The faded black lines of the eye and eyebrow are now more visible and it is clear that there were pink highlights on the nose and the cheek of the face, presumably composed of an organic lake, applied in diluted and fine brush strokes. This face has always been considered as white, and Lang in her description (Lang 1969, pp. 83-84) nowhere mentioned the superposition of pink highlights since these were almost entirely veiled by saline incrustations. The artist who created this face was much more concerned with the rendering of plastic effects than was initially thought, breaking with the bi-dimensionality of a uniform white paint.

A preliminary attempt was made to fix the restored fragment on a Plexiglas mat with aluminium attachments, for possible future display on a vertical surface.

b) Modern gesso coatings were removed from the back of two large fragments of fresco that joined (21D46 and 9Fnws). This operation required the use of big scalpels, pincers, and an electric auger, and was performed by the two professional restorers, Musella and Kapizionis. The fragments were liberated from the thick plaster; each was cleaned, and then precise joins glued together. Finally new mortar (composed of two parts of sand and one part of lime) was applied to fill gaps and cavities on the backs of the fragments, in order to reinforce their adherence.

New mortar was applied as well to other fragments with deep cracks, holes on their surface, and/or with badly damaged edges.

2.4. Technical Examination of the Wall Paintings in Situ (X-Ray Fluorescence, X-Ray Diffraction, PIXE) and the Results of Laboratory Analyses (Gas-Chromatography/Mass Spectrometry, High Pressure Liquid Chromatography, FTIR).

Measurements were taken in situ for the characterisation of inorganic materials, pigments, and mortars by means of XRF, XRD and PIXE.

Analyses were performed on wall-painting fragments by means of the portable X-Ray Fluorescence (XRF) equipment of the Institute of Nuclear Physics “DEMOKRITOS”, and the portable XRD and PIXE-ALPHA equipment of Catania University and the Laboratory of Non Destructive Analysis of the Istituto Nazionale di Fisica Nucleare. Our goal was to identify the chemical composition of mineral and artificial pigments used for the creation of the paintings. 84 "positions" (i.e., points on individual fragments of frescoes) were analysed this year by means of XRF, 26 positions by means of PIXE-ALPHA and 3 by XRD. Major objectives of this year’s analytical program were the application of complementary non-destructive techniques in order to confirm measurements obtained in previous years, and to obtain more quantitative data by estimating, for instance, the presence of elements low in the periodic table, such as Si and Al.

These analyses also allowed us to investigate quantitative variations in the composition of various hues of the same pigment, in the case of Egyptian blue. PIXE-ALPHA measurements taken on 11 fragments of Egyptian blue from different rooms/areas of the palace confirmed that lighter tones of blue contain a lower amount of copper. XRF had in previous years pointed to the presence of Si (silica) on white paint layers, but it was impossible to determine with certainty if this element was contained in the mortar or if it constituted a major element in the pictorial layer. The situation was clarified this year by the complementary use of XRD and PIXE. It is now obvious that silica and aluminium belong to the composition of the mortar (sand); the pictorial layer in most cases proved to be pure calcite.

Analysis of new positions by means of XRF further enriched our data-base, which is now composed of ca. 300 analysed pigments. In general, these confirmed the palette that has been reconstructed in previous years. Differences in the choice of blacks provided further evidence for distinguishing groups of fragments and for associating them with different painters and/or workshops. It is interesting to note at this point that an identical checkerboard pattern attested on two fragments from different contexts was executed by means of two different black pigments, in one case with an organic black and in the other with the mineral black pyrolusite; such differences suggest different workshops were involved and remind us that stylistic affinities may not always constitute a solid criterion for associating fragments from different contexts.

Binding materials and organic dyes were further characterized by means of Gas-Chromatography/Mass Spectrometry, High Pressure Liquid Chromatography, and FTIR. Already last year there had been stressed the importance and the particular interest of the abundant presence of pinkish hues, ranging from dark purple and violet to pale pink; it was suggested that their origin should be organic, since no other elements belonging to inorganic pigments were detected by XRF analysis. This hypothesis was confirmed this year by analyses performed under the supervision of Colombini at the University of Pisa: organic lakes, both of vegetable and animal origin, had been employed.

In most cases, the particular organic dye could not be identified, either because it had deteriorated and its chemical formula thus was altered, because the composition of the plant used to produce the dye is no longer available today, or because it has not yet been documented. In any case, the pink colour is not derived from the well-known plant Rubia Tinctorum (alizarin red), the most common vegetable source for red dyes in Classical antiquity, and of a colour very close to the hues represented in our fragments.

Of great interest is the detection of murex purple on two samples taken from the northwestern wall of the Throne room; macroscopically these fragments presented the most intense hues. This fact demonstrates that there was a particular interest in the choice of colour employed in the most important room of the palace and suggests that the colour itself may have had symbolic and ritual meanings in addition to aesthetic value; it also enriches our knowledge of the use in the Aegean world of this expensive and rare dye. Dyes with a vegetable provenance were also employed in association with murex purple, but more analyses will be required to determine their precise character. Likewise, we suspect, but are not yet able conclusively to prove, that the use of murex purple was reserved for the Throne Room.

The identification of such fragile and precious materials as those used in the Throne Room of the Palace of Nestor has never been carried out in a systematic manner in the analysis of Aegean frescoes, with the only exception being the recent identification of murex purple on a wall painting from Thera. Consequently, results obtained from the analyses of paint samples shed entirely new light on the technology of Aegean painting.

Our analyses also are relevant to the long-standing debate over the manufacture of Aegean paintings: are they real “frescoes”? The characterisation of the binding materials used in the frescoes from the Palace of Nestor, performed by means of GC/MS, HLPC and FTIR, confirmed in the case of almost all samples, the use of organic binders, and, therefore, that a tempera technique had been employed. Egg and tragacanth gum, derived from the plant astragalus, native in Greece, were the major components used by Mycenaean painters to bind their pigments and to apply their colours. It is interesting to stress here that these materials were also the ones used by Classical painters, together with Arabic gum, as attested by recent analysis of late Classical and early Hellenistic paintings. It is thus clear that techniques used by Mycenaean painters in the LH IIIB period were extremely sophisticated.

2.5. New Joins and New "Discoveries"

The final organisation of the Apotheke 1 of the Hora Museum has facilitated enormously the study of the frescoes and their restoration. In addition, the use of a stereomicroscope, loaned to us by the director of the Wiener Laboratory of the ASCSA, has helped us to retrieve otherwise macroscopically invisible motifs, and to reconstruct in many cases the original polychromy of scantily preserved fragments. The most interesting information revealed from this year’s study may be summarized as follow:

• A new join was found between the “archer fragment” (see our report from 2003), uncovered outside the ashlar wall of room 32, and a tiny fragment stored together with other fragments from room 27. The interest of this new join is twofold: on the one hand it helped in reconstructing the missing part of the hand of the figure and the lowest edge of the object she is holding, confirming the suggestion in our previous report that it is a bow. On the other hand, this discovery allows us to associate the “archer fragment” definitively with room 27. Further research should be conducted on the context of these two joining pieces in order to clarify the original location of this interesting wall painting fragment (analysis on fragments from room 27 and the “archer fragment” conducted last year confirmed the use of a common palette).
• Analysis under the stereomicroscope confirmed remains of a green colour on large fragments of painted plaster from the Throne room: green fronds and flowers were not published in Lang’s catalogue. These new elements are parts of the iconographical program of the Throne room; they are rendered in a naturalist rather than in a stylised manner, and should be compared with floral and nature motifs attested elsewhere in Aegean wall paintings. The nature of the green pigment applied on these fragments is also of great interest, since it consists of a natural copper-based pigment and is not a mixture of blue and yellow as usually is the case for other fronds and floral motifs in the palace.
• The presence of a small human head was noticed, and confirmed under microscope, below the scene of “Two Men at Table” on fragment 44a H 6.
• A new fragment was found that belongs to the same scene as two other fragments previously identified as parts of a ship (19 M ne). This fragment was found together with others from room 31, and it amplifies the reconstruction proposed recently by Maria Shaw ("Symbols of Naval Power at the Palace at Pylos: The Evidence from the Frescoes," ITHAKI: Festschrift für Jörg Schäfer zum 75. Geburtstag am 25. April 2001, Böhm, Stephanie and Klaus-Valtin von Eickstedt, eds., Würzburg: Ergon Verlag, pp. 37-43).

The most interesting “discovery” of this year consisted of a group of unpublished fragments that presumably depict a naval theme: ship patterns, oars, and perhaps fish were recognized. The fragments were previously stored in an exposed location on an old shelf. During the cleaning of the pictorial surface unusual patterns were noticed that do not appear in Lang's publication. Surprisingly, the entire series of fragments (11 in total, 5 large and 6 small) not only has not been published, but is not mentioned in print.

Among these fragments with naval themes there was no written indication regarding their context, but they were stored together with other fragments from Hall 64 that depict battle scenes. In 2003 we found many other fragments from Hall 64 in old boxes, almost all of them poorly preserved. Contextual information was still preserved on four wooden tags (HS7- Ext. 5 plaster fragments from wall HSW or high up on wall S, R. H., p. 144, 26-6-53; HS7- Ext. 6, plaster from floor near wall HSW, R. H., p. 150-151, 27-6-53; HS7-Ext. 7, R. H., p. 181, 3-7-53), and we had stored these together with the published fragments found in Hall 64.

Affinities were immediately noticed between these fragments and those with ship patterns as regards the texture of the surfaces, the thickness of the mortars, and their state of preservation. After a preliminary cleaning of the surfaces of the fragments studied in 2003, we found a small piece with a zig-zag pattern that joined to an edge of one of the large fragments, discovered in 2004, that evinced the exact same pattern. This join allowed us to determine definitively that Hall 64 was the context of the unpublished fragments with naval themes.

Presumably the poor state of their preservation discouraged Lang from investing time in cleaning and further studying these. Their surfaces gave us the impression that they had never, in fact, been cleaned; earth still adhered to them and they had been removed from the floor of the room by applying plaster to their back. It is now a major priority that these pieces be properly restored, so that the maximum amount of information relevant to the iconography of these unique scenes can be recovered.

The background of the composition seems to have been originally purple, but fire altered dramatically the original polychromy. Poorly preserved areas with traces of purple colour recall the purple dye used as a background mixed with blue for the main battle scene of Hall 64. Purple background has elsewhere been attested so far only in the main palace building on the main compositions of the vestibule (Procession Scene) and the Throne room (Lyre Player and the so-called "Bull's Shoulder"), and, from the Southwest building, only in Hall 64.

Shaw's intuition that there should be in Hall 64 a naval theme in Hall 64 seems confirmed, although the fragments from the ne fresco dump on which she based her argument do not come from this room. The checkerboard patterns that they share with the battle scenes from Hall 64 cannot be taken to prove that they derive from the same composition. In the first place, this pattern also appears in room 5 and in room 20 (and also in the nws fresco dump). In the second place, and of greater importance, is the fact that a different black colour was employed in each case to trace the checkerboard pattern. The black used on Shaw's fragments from the ne fresco dump is made of an organic black, presumably carbon black; in contrast, the black used on fragments from the border from 64 is pyrolusite, a regional mineral that is also attested in other parts of the battle scene fragments published by Lang.

6.1. Objects in Apotheke 2:

8 large wooden boxes of material marked:

• Kokkevis
• Vayenas Small Finds
• Vayenas Pithos Covering
• PoN Vayenas 57’58 Ceramics
• PoN Vayenas Ceramics
• Misc. Tombs Ceramics and Small Finds
• Tsakalis Ceramics
• Tholos 1953.

6.2. Objects in the Chora Museum: