Surface Coatings on Terracotta Objects from Boeotia and Taranto 400 - 200 BC - the influence of composition on the susceptibility for detachment

(1)

Surface Coatings on Terracotta Objects

from Boeotia and Taranto 400 - 200 BC

- the influence of composition on the

susceptibility for detachment

Fiep Korstanje BA 11741686 Fiep_korstanje@hotmail.com MA Thesis

Conservation and Restoration of Cultural Heritage Specialisation: Glass, Ceramics and Stone

University of Amsterdam, Amsterdam

Supervisor: Kate van Lookeren Campagne – Nuttall 21-06-2019

(2)

(3)

Introduction

From the Bronze Age, terracotta objects were produced on a large scale in Ancient Greece. These objects were decorated with different organic and mineral pigments that were applied on top of a light-coloured surface coating.1_{The composition of these surface}

coatings often remains unclear, being referred to for instance as slip layers or as surface coatings containing, or mixed with, calcium carbonate.2_{Because it is unknown what type}

of coating is present, it has been decided in this research to refer to this layer as a ‘coating’. The application of a surface coating on terracotta was done from the fifth century BC onwards.3_{At the Allard Pierson Museum, an important collection of these Greek}

archaeological terracotta sculptures is stored. On some objects, a much larger percentage of the surface coating remains than on others. The exact factors that have led to this difference in loss of surface coating are unknown. Probable causes for the loss of surface coating are the composition of the coating, burial conditions as well as how the objects were treated on or after excavation.

This thesis discusses the research undertaken into the possible reasons for the variation in percentages of loss of the coatings and the way in which loss occurs on these terracotta objects. The main research question is: ‘In how far does the composition of the surface coating on ancient Greek terracotta figures influence the susceptibility of the coating layer?’, with focus on a specific region and period.

For this research, six terracotta objects with varying conditions of surface decoration were selected from the collection of the Allard Pierson Museum. Three originate from Boeotia (in Greece) and three are from Taranto (now Italy). All objects date between the fourth and the second century BC.4

It is hoped that the results of this research will aid the preservation of Greek terracotta figures, by giving greater insight in susceptibility of the coating layer for damage.

1_{Robert A. Lunsingh Scheurleer, Grieken in Het Klein: 100 Antieke Terracottas (Amsterdam: Allard Pierson}

Museum, 1986), 9. & Paula G. Leyenaar-Plaisier, Griekse Terracotta's: Uit De Collectie Van Het Haags

Gemeentemuseum (Den Haag: Haags Gemeentemuseum, 1986), 12-21. & Herman Brijder, "Griekse

Godinnen En Meisjes in Het Klein," in Kleur! Bij Grieken En Etrusken, ed. Vincenz Brinkmann and Herman Brijder (Amsterdam: Allard Pierson Museum, 2006), 57. & R. A. Higgins, "The Polychrome Decoration Of Greek Terracottas," Studies in Conservation 15, no. 4 (1970): 273.

2_{Anne Rinuy and Francois Schweizer, "Analysis of the White "ground" and Ancient Adhesives Found on}

Canosa Vases (south Italy) of the Third Century B.C.," in Proceedings of the 18th International Symposium on

Archaeometry and Archaeological Prospection, Bonn, 14-17 March 1978; Archaeo-Physika, Band 10, Rheinisches Landesmuseum Bonn (Köln: Rheinland-Verlag GmbH, 1978), 255. & Riemer R. Knoop, Antefixa Satricana: Sixth-century Architectural Terracottas from the Sanctuary of Mater Matuta at Satricum (Le Ferriere) (Assen/Maastricht: Van Gorcum, 1987), 21.

3_{Gloria S. Merker, "Corinthian Terracotta Figurines: The Development of an Industry," Corinth 20 (2003):}

234.

4_{Allard Pierson Museum Archeologische Collectie, December 2004, accessed February 1, 2019,}

(8)

(9)

Chapter 1. The Case-Study Objects

For this research an assemblage of six terracotta objects from the Allard Pierson Museum in Amsterdam, The Netherlands, was analysed (appendix I fig. I.1-I.21). The group was selected based on several criteria. The main criterion for selection was a variation in the condition of the white coating layers to enable comparison of objects. In addition, the provenance and date of the object needed to be known and literature about the specific objects available. In order to enable comparisons between the objects, the object’s dates needed to be as close as possible. Two different places of origin were chosen to see whether a correlation between a production region and coating composition could be made. Based on these characteristics, a pre-selection of fifteen objects was made and handheld XRF analysis (see chapter 3) was carried out to get an overview of the type of coating present on the objects in order to make sure that the objects would provide a balanced range. The selection of six suitable objects was made based on the percentage of aluminium, silica or calcium in the coating which suggested the presence of either clay minerals or chalk.

In this chapter the archaeological context of the objects will be discussed. In addition, an overview of the six objects is included and the condition of the objects is described, including a detailed description of the surface coating (table 1 & 2).

1.1 Context of the Objects: From Excavation to Museum

Three of the objects were produced in Taranto. This city, which is currently in Italy, was a Greek colony during the fourth and third century BC. The other three objects originate from different cities in Boeotia, a region in central Greece (fig. 1). The place of production has been based on the stylistic characters of these objects.

Five out of six objects used in this research were part of the collection of C.W. Lunsingh Scheurleer, who was professor of classical archaeology at the University of Leiden. Lunsingh Scheurleer was a collector of ancient artefacts who bought objects from dealers and collectors often at international art fairs. Most items were acquired by Lunsingh Scheurleer between 1900 and 1932. Between 1912 and 1920 many objects from Boeotia and Taranto were acquired, which make up a large part of the collection nowadays. The objects were acquired by the Allard Pierson Museum in Amsterdam when the collection was up for sale and at risk of being separated.5

One object, the altar/incense burner (APM14207), was donated to the Allard Pierson Museum in 1998. No further information is known about who donated this object.

5_{Leyenaar-Plaisier, Griekse Terracotta's, 3. & Lunsingh Scheurleer, Algemeene Gids, ed. L.J. Elferink}

(10)

Furthermore, the date of production and place of production is known as defined by the museum conservator or curator.6

In the beginning of the twentieth century, little emphasis was put on preserving information regarding the context and history of such objects in collections. Often hardly any information exists about where objects were found or excavated and who the objects were sold to. For this research, only objects of which the provenance and a production date are determined by the conservator or curator, are included.

Some of the objects have been photographed for publications for museum databases. Pictures taken of the objects previous to this research have been used to try to detect possible changes over time in the condition of the surface coating that is present on the objects (appendix II fig. II.1-II.9). If there was a clear difference between the amount of surface coating in the old pictures and the objects now, this may be an indication that the coatings were very friable and this would assume that a problem with the attachment of the coatings was present, although this is considered unlikely as they have been sitting in storage since their acquisition. Even though the quality of some of the pictures is not very good and some pictures are in black and white, it appears that between now and when the pictures were taken hardly any signs of loss of surface coating can be detected.

6_{Allard Pierson Museum Archeologische Collectie, http://dpc.uba.uva.nl/archeologischecollectie.}

Figure 1. Map depicting Taranto (modern Italy) and Boeotia (Greece). Source: Adapted from Google Maps. “Taranto and Boeotia” Accessed June 5, 2019.

Boeotia

Taranto

(11)

1.2 Object Description

7_{A chiton is “a long woollen tunic worn in ancient Greece.” Oxford English Dictionary, “Chiton,” accessed June 5, 2019, https://en.oxforddictionaries.com/definition/chiton.} 8_{A himation is “an outer garment worn by the ancient Greeks over the left shoulder and under the right.” Oxford English Dictionary, “Himation,” accessed June 5, 2019,}

https://en.oxforddictionaries.com/definition/himation.

Inventory Number

Date Place of Origin Dimensions (mm) Colour of Ceramic Body Description APM00257 (fig. 2) 400-200 BC Livadhia, Boeotia (Greece)

175h x 60w Yellow-red Standing woman wearing a chiton7_{and over that a himation}8_{until the}

knees that covers the head and arms. The object has a rectangular fire hole at the back.

APM00277 (fig. 3)

400-200 BC Thebe, Boeotia (Greece)

90h x 58w Red-brown Head of a woman with long hair over the shoulders, a tiara and hair ribbon. APM00394 (fig. 4) 400-200 BC Tanagra, Boeotia (Greece)

215h x 85w Red Standing woman wearing a chiton and over that a sideways draped himation, also over the head. The right hand is placed on the waist. The object has a rectangular fire hole at the back.

APM01145 (fig. 5)

400-300 BC Taranto (modern Southern Italy)

190h x 125w Yellow-red The lower part of a woman who is walking to the left. The figure is wearing a flowing himation with plunges that was held tight at the waist and depicted a girl bringing sacrificial gifts.

APM01161 (fig. 6)

300-275 BC Taranto (modern Southern Italy)

210h x 85w Light red Figure of naked Eros with spread wings and short hair. The body is robust and the wings are large relative to the body. The quills of the feathers are marked with ridges.

APM14207 (fig. 7) 300-200 BC Taranto (modern Southern Italy) 80h x 68w Light beige with a grey core

The object is a rectangular altar or incense burner with four decorated sides: A. Leto leaning on a rod, cornucopia, Apollo with lyre; B.

Poseidon with trident, Amymone; C. Dionysus hugging Ariadne, satyr; D. Woman garlanding a tropaion on stone. The top of the object is closed with the exception of a small slot. A profile is visible on top which is partly decorated with an egg-and-dart motif. On two corners of the top cover, palmette and volute shaped ornaments are present, which are connected by an edge with ramage motifs.

(12)

Figure 2. Standing woman with chiton and himation (APM00257)

175h x 60w

Figure 3. Head of woman (APM00277) 90h x 58w

Figure 4. Standing woman with hand on side (APM00394) 215h x 85w

Figure 5. Lower part of walking woman (APM01145) 190h x 125w

(13)

Figure 6. Eros (APM01161) 210h x 85w

Figure 7. Incense burner (APM14207) Side A 80h x 68w

(14)

Table 2. Object Condition and Surface Coating

1.3 Object Condition and Surface Coating

Inventory Number

Location Surface Coating Description Surface Coating Surface Decoration Deterioration Surface Coating Remaining

Coating

APM00257 Coating present on outside of object. No coating visible on inside, possibly due to encrustation layer.

Encrustation layer visible on top of the coating.

Coating: thin and evenly

applied. Only visible in places where encrustation layer is missing.

Colour: bright white.

On the legs, blue and pink pigments are present.

Layer seems worn on protruding areas such as nose.

70%

APM00277 White coating on the front and on part of the back.

applied. Along the lower areas, like around the nose, a thicker layer is present.

Red-brown paint/surface coating is present on the hair and traces of black paint are present on the eyes. A light pink paint seems to have been applied on the face.

Layer is worn on the more protruding areas.

70%

APM00394 Coating present on outside of object. No coating visible on inside, possibly due to encrustation layer.

Encrustation layer visible on top of the coating.

applied.

Black pigment is present near the eyes and hair. On the lips red pigment is present, and red and light blue pigments are present on the clothing. On the face some traces of light pink pigment are visible.

Layer seems worn. 90%

APM01145 Coating present on outside of object, not on inside.

applied.

Yellow-brown pigment is present on the himation.

Only small parts of coating remain in the form of slivers on top of the object as well as a thin and evenly applied surface in between the ridges.

10%

APM01161 White coating on the object, excluding the hair region, the inside and the back as well as part of the front of the wings.

applied.

Red, brown and pink traces of paint are present on the body. On the wings, also traces of green/blue and yellow can be detected.

Layer seems worn. No fresh breaks along missing coating visible.

60%

APM14207 Coating present on outside of object, not on inside.

applied.

Pink traces are visible on all sides as well as on top of the object.

Breaks along missing coating indicate that coating has detached and is not worn.

(15)

1.3.1 Standing Woman with Chiton and Himation (APM00257)

The ceramic of the object is yellow-red in colour. A coating layer is visible all over the front and the back of the object. On top of the coating layer, a grey encrustation layer is present, which is characterised by a white or off-white deposit on top of the coating layer. Also, on the inside an encrustation layer is present. The coating can be detected due to its very white colour, which can be seen on the face, mainly along the lower areas such as on the side of the nose, chiton as well as on the legs of the object. The coating is very smooth and seems to have been applied evenly (fig. III.2, III.4). Whether a coating was also applied on the inside is not visible due to the encrustation layer. On the legs blue and pink pigments are present (fig. III.3).

The coating is missing along the ridges, on high areas, as well as on the face (fig. III.2). Furthermore, on the back of the object an area is present where surface coating is missing (fig. III.5). In the ridges of the ceramic in this area, some coating is still present, showing that coating was applied here. About thirty per cent of the substrate area is exposed. No fresh breaks are present around the areas where surface coating is missing. Instead the layer seems to gradually become thinner up to areas where coating is missing, which would suggest that the layer is removed mechanically. Loss of the encrustation layer can however be detected on the back of the object, at the same place where the coating is missing. This may suggest that when the encrustation layer was removed, the coating was removed as well. Furthermore, no encrustation layer is present along the ridges and high areas.

1.3.2 Head of Woman (APM00277)

The ceramic has a red-brown colour. A white coating is present on the facial area of the object, mainly in the lower areas. Coating is missing in the more protruding areas (fig. III.7). A brown/red coating layer, which is similar in structure to the white coating is present on the hair and partly on the back of the object (fig. III.11). On the rest of the back of the object no coating is present. It seems that this area has never been coated since the dirt that is present on the coating has the same colour and amount of soiling as is present on the areas where no coating can be detected. This dirt looks like a very thin grey wash and is only present on the back of the object. The white coating is in some areas present underneath the red coating, around the hair. No white coating can be detected underneath the rest of the hair colour. The coating is very smooth and seems to have been applied thinly (fig. III.8, III.9). Traces of black are present on the eyes. A light pink paint seems to have been applied on the face.

Only the head of the object remains. The body is missing. The coating is missing along the more protruding areas such as the nose, chin, cheek and the hairband. No fresh breaks are present around the areas where surface coating is missing. Instead the layer seems to

(16)

gradually become thinner around areas where coating is missing, which would suggest that the layer is worn or was removed mechanically (fig. III.7, III.10). About seventy percent of the surface coating remains.

1.3.3 Standing Woman with Hand on Side (APM00394)

The object has a very bright red ceramic colour and an encrustation layer is present all over the object. In areas where the layer is missing, a bright white coating is visible (fig. III.13-15). The coating seems to have been applied all over the object since in most areas where the encrustation layer is missing, which is evenly over the object, coating is present underneath. Assuming that the coating is present under the encrustation layer, about ninety per cent of the surface coating remains. Even though it is not visible everywhere due to the encrustation layer, the coating seems to have been applied thinly. Black pigment is present near the eyes and hair. On the lips, red pigment is present, and red and light blue pigments are present in the ridges of the clothing along the legs. On the face, some traces of light pink pigment are visible.

The object has been broken into four pieces on the side. The pieces have been adhered during a previous restoration and one area in the back has been filled (fig. III.12, III.16). The surface coating seems to have been kept intact due to the encrustation layer on top.

1.3.4 Lower Part of Walking Woman (APM01145)

The ceramic has a yellow-red colour. Traces of a very white coating are present all over the object, as well as traces of an encrustation layer. The traces are mainly present between the ridges of the object (fig. III.18, III.20). The white coating appears to have been applied thinly and evenly over the object (fig. III.18). Yellow-brown pigment is present on the himation.

The top part and the feet of the object are missing. There are brushstrokes present on the leg of the object, assuming that the object was severely cleaned by a brush (fig. III.21, III.22). About ten per cent of the surface coating is still present on the object. Some fresh breaks are present around the areas where surface coating is missing, in combination with the layer gradually becoming thinner around areas where coating is missing (fig. III.19, III.21). This would suggest that the layer was removed mechanically, but could also have detached.

(17)

1.3.5 Eros (APM01161)

The ceramic has a light red colour and a bright white coating is present on the object, excluding the hair region, the inside and the back as well as part of the front of the wings. The coating seems to have been applied thinly and evenly. The top of the wing does not seem to have had a white coating applied, since no traces of coating are present here at all, in contrast to areas where coating is missing but traces are still present (fig. III.24). Possibly a different type of decoration was present on these parts of the wing. Pinktraces of paint are present all over the body. Traces of brown are visible in the hair region. On the wings, also traces of green/blue and yellow can be detected, mostly in between the ridges of the wing (fig. III.29).

The object is missing part of one of the wings as well as one of the legs and the back of the other foot. The foot has been adhered together. About sixty per cent of the surface coating remains. Furthermore, no fresh breaks are present around the areas where surface coating is missing. Instead the layer seems to gradually become thinner around areas where coating is missing, which would suggest that the layer is worn or was removed mechanically (fig. III.25-27).

1.3.6 Incense Burner (APM14207)

The ceramic is light beige and has a grey core.9_{A bright white surface coating has most}

likely been present all over the outside of the object. The coating seems to have been applied evenly and is thin (fig.III.36-37). Some traces of the coating are still present along the upper decoration. Pink traces are visible on all sides as well as on top of the object (fig. III.32-33). Also, a grey encrustation layer is present on the object (fig. III.32-33).

Some traces of a previous restoration are present in the form of adhesion of the object and a fill. One upper corner of the object is missing. About forty per cent of the surface coating remains. Along the depictions the coating seems to be missing from the object (fig. III.31, III.35). On the background of the reliefs, the coating is still attached. On the more protruding areas of the depictions, however, the coating is often missing. Also, on the top of the object as well as on the decorations on the upper region, most coating is missing. Where the surface coating is missing, a break edge is visible (fig. III.36-37). On handling during conduction of the XRD analysis, a very small part (about 2mm) of the coating with terracotta detached which must already have been loose (fig. III.31). This flake came off attached to a small layer of terracotta. The pigments on top of the object seem to have broken in smaller pieces, but are still attached to the object (fig. III.40-41).

9_{A grey core appears when the firing cycle is too short so that the organic material is not fully burnt out of}

(18)

(19)

Chapter 2. State of the Art: Greek Terracotta Figurines

This section discusses what is written in literature about the production process of terracotta objects and the application of a coating (table 3). Four main sources have been found that describe forming and firing of the object. Three of these sources, namely Lunsingh Scheurleer (1986), Leyenaar-Plaisier (1986), and Kaufmann (1915) give a general overview of the production of ancient Greek terracotta figurines and discuss the stylistic differences of several archaeological finds. The authors give a general description of the steps that were undertaken to produce these objects, including collecting the clay, forming the objects, and how the objects were decorated and fired. These sources were written using few references to other sources. The reason for this is probably that no or few earlier published sources for the production of these objects were available. It has proved impossible to track-down where the information given by these authors comes from. Also, the information is mostly art-historical. This is due to the unavailability of instrumental analysis in that period but also because the authors were not material scientists. Brijder (2006), who focusses on the coloured decorations on ancient Greek terracotta objects only gives short, descriptive statements on how objects were produced and what type of coating was applied.

In addition, Richter (1948), Oliver (1986), Knoop (1978) and Lulof (1991) described ancient Greek terracotta objects in which research was done into the provenance of the object or its style. In the article by Richter (1948), a terracotta head from the Metropolitan Museum is described, Oliver (1968) discusses ceramic objects from two Apulian tombs and Knoop (1987) and Lulof (1991) discuss architectural and monumental terracotta from Satricum, Italy. Each of these sources have included short statements on the coating that is present on the terracotta objects discussed. The scope of their research focussed on the description of the objects and the use, meaning and style of these objects. These statements, even though very useful to get an overview of what could have been applied on terracotta objects, does not give enough background information to compare the type of coating that is present and their deterioration. The reason that little chemical analysis has been implemented in earlier research is probably due to the limited analytical facilities as well as the research focus on the composition and deterioration processes within archaeology. More recent literature, as well as a few earlier sources, include chemical analysis of the composition of the clay or the surface coating. The first available sources that have discussed instrumental analysis of surface coatings on terracotta objects, are Higgins in 1970 and Rinuy and Schweizer (1978). Higgins describes the use of XRD analysis conducted on different ancient Greek terracotta objects. How this analysis was conducted or the exact results have not been included. The same counts for Rinuy and Schweizer, who have researched 3rd_{century BC Canosa vases, which are in provenance and production date very}

close to the figurines discussed in this research. Even though no methodology or results were included, this source has provided the first chemical analysis of objects comparable

(20)

to the objects in this research. Unfortunately, no deterioration patterns were discussed in this source, making it impossible to make comparisons between the objects.

Three more recent sources have included an instrumental analysis. First of all, Bordignon

et al. (2008) researched potsherds and polychrome terracotta from Cerveteri, Italy. In their

research, micro-Raman spectroscopy, X-ray Diffraction (XRD) and Fourier-Transform Infrared Spectroscopy (FTIR) were used to analyse the surface coatings. Furthermore, Costello and Klausmeyer (2013) analysed the surface coatings on 3rd_{century BC terracotta}

statues from Canosa. FTIR, SEM, and TGA were applied in this research. Finally, Kakoulli (2017) researched objects with a similar provenance and production date. This research applied XRF analysis to determine the composition of the surface coatings. The chemical analyses discussed in these sources include information that can help to interpret the results of instrumental analysis in this research. The results of these analyses have been included in table 3 and in paragraph 2.1.2, which describes the different coating materials that have been discussed in the literature.

To conclude, even though many sources are available in which the presence of a surface coating on terracotta has been mentioned, few of these sources included instrumental analysis. The articles that have included chemical analysis, though useful to get an idea of the composition of surface coatings on ancient terracotta, have mostly either not included the exact methodology (Higgins (1970) and Rinuy and Schweizer (1978)) or have discussed objects that have a different provenance or production date, such as Bordignon. Two articles, however have included extensive analysis and results on the composition of surface coatings on terracotta objects, namely Costello and Klausmeyer (2013) and Kakoulli

et al. (2017), who both discussed the composition of 3rd_{century BC Canosa (Italy) terracotta}

(21)

2.1 Production of Ancient Terracotta Objects

Suitable potters’ clay is found all over Greece and Italy and was used to create objects such as figurines.10_{The objects made of the iron-rich earthenware clay that was found are}

referred to as “terracottas”, based on the Italian word “terra” which means clay or earth.11

Terracotta figurines were made from the Bronze Age onwards. The objects, as well as the knowledge of how to produce them, spread from the Greek mainland along the Mediterranean Sea. Different types of objects can be traced back to specific production centres and can be dated based on stylistic elements.12

During the Classical Period (480-323 BC) and the beginning of the Hellenistic/Roman Period (323BC-300AD), to which the objects used in this research date, some of the largest production centres of terracotta objects were in Boeotia (Greece) as well as Athens (Greece) and Taranto (modern Italy).13

In the Classical Period, such figurines mainly had a religious function, and would either be used as votive statues or during funerary practices. However, terracotta objects were also used as toys. During the fourth century BC, terracotta statues also started to be used as decorations in for instances living areas. An example of this are incense burners which were solely made as decorations. During the Hellenistic/Roman Period, the function of terracottas as art objects increased dramatically. Terracotta objects were mainly found in graves, houses, sanctuaries and temples.14

From the fifth century BC onwards all terracotta figurines were decorated with pigments that were added on top of a white surface coating.15_{This coating would be applied for}

several reasons. First of all, a surface coating would provide a light-coloured base for other colours that would not be visible on a red-fired ceramic and coatings could create surface structure or can even out the surface of an object.16_{Moreover, the application of a slip}

coating decreases the porosity of an object.17_{Table 3 provides an overview of the sources}

found.

10_{Lunsingh Scheurleer, Grieken in Het Klein, 9.} 11_{Lunsingh Scheurleer, Grieken in Het Klein, 9.} 12_{Leyenaar-Plaisier, Griekse Terracotta's, 12-21.}

13_{Leyenaar-Plaisier, Griekse Terracotta's, 14. & Lunsingh Scheurleer, Grieken in Het Klein, 17-18.} 14_{Lunsingh Scheurleer, Grieken in Het Klein, 14. & Leyenaar-Plaisier, Griekse Terracotta's, 15.}

15_{Lunsingh Scheurleer, Grieken in Het Klein, 9. & Leyenaar-Plaisier, Griekse Terracotta's, 12-21. & Herman}

Brijder, "Griekse Godinnen En Meisjes in Het Klein," 57. & Higgins, "The Polychrome Decoration Of Greek Terracottas," 273.

16_{Sabine Fourrier, "East Greek and Cypriote Ceramics of the Archaic Period," Cyprus and the East Aegean,}

October 2008, 134. & Erophile Kolia, "Archaic Terracotta Reliefs from Ancient Helike," Hesperia: The Journal

of the American School of Classical Studies at Athens 83, no. 3 (September 2014): 431. & Brijder, "Griekse

Godinnen En Meisjes in Het Klein,” 57. & Owen S. Rye, Pottery Technology: Principles and Reconstruction (Washington, D.C.: Taraxacum, 2002), 41.

17_{Harry Fraser, Ceramic Faults and Their Remedies (London: A. & C. Black, 2005), 118. & Prudence M.}

(22)

Author & Date Object(s) studied Region of study Date of object(s)

Aim/area of study Analysis used to

study surface decoration Description/ composition of surface coating Production details Kaufmann 1915 Terracotta objects Greek-Roman from different find sources Greek-Roman (509-27BC) and Coptic period (300-700) History of koroplastic production and function of objects

None given ‘Kalkmilch’ (chalk

milk)

Coating applied after firing. Background decorated with black, red or yellow or completely painted (p.20) Richter 1948 A terracotta head from the Metropolitan Museum.

Greek End of the Archaic Period (800-480BC)

Study of one object None given ‘yellowish white slip’ (p332) ‘fine whitish slip’ on all Greek and Etruscan terracotta objects (p.335)

Coating applied by brush before firing. Painted colours also applied before firing (p. 335) Oliver 1968 Ceramic objects from two Apulian tomb groups

Canosa, Italy 3rd_{and 4}th

century BC

Reconstruction of two Apulian tomb groups from Canosa. Focus on pottery and the meaning of the objects within their context.

None given White lime wash, white ground or whitewash

Coating was applied after firing

Higgins 1970

Terracotta objects

Greece Not specified Analytical research of the polychrome decoration

XRD White clay containing gypsum (calcium sulphate dihydrate) (p.275) None given Rinuy & Schweizer 1978 Several terracotta vases

Canosa, Italy 3rd_{century BC} _Production

techniques of these vases and terracotta figurines

AES, XRD, TGA Slip made of pure kaolinite

Coating applied after firing

(23)

Lunsingh Scheurleer 1986 100 terracotta objects from the collection of the Allard Pierson Museum

Greece Not specified Catalogue including stylistic research of terracotta objects

None given White clay-slip Coating applied solely to the front of the object before firing

Leyenaar-Plaisier 1986

Terracotta objects

Greece Not specified Exhibition catalogue of objects from the Allard Pierson Museum

None given White paint on clay basis, called an engobe or ‘deklaag’ Coating applied before firing. Maximum firing temperature possible in the kiln was 1100˚C Knoop

1987

Architectural terracotta

Satricum, Italy 6th_century _{Description, context}

and historical meaning of architectural terracotta

None given Slip containing calcite. The term slip is used since this ‘refers to simply any surface coating’ (p.196)

Coating was applied by brush before firing. Firing temperature of maximum 890 degrees Celsius, since this is the critical decomposition temperature of the calcite (p. 196) Lulof 1991 Monumental terracotta statues

Satricum, Italy Late Archaic period (800-480BC) Description, context and historical meaning of monumental terracotta

None given Slip: ‘a fluid

suspension of clay in water, sometimes mixed with calcite to produce a white colour’ (p.57)

Coating was applied on a damp surface Merker 2003 Terracotta figurines Corinth, Greece Ancient times (1200BC- 600AD) The development of Corinthian terracotta figurines

None given White slip is present from the 5th_century

BC onwards (p.324) None given Brijder 2006 Terracotta objects Mediterranean area

Not specified Colour used on terracotta objects, including the production of terracotta objects in general

None given Kaolin (aluminium silicate) (p.57)

Coating applied before firing

(24)

Bordignon 2008 White-on-red potsherds & high-value polychrome terracotta Cerveteri area, Italy Orientalising period (800-600BC) & Archaic period (800-480BC) Spectroscopic identification of kaolin used in pigments on terracotta and kaolin found in the Monte Sughereto quarry Micro-Raman spectroscopy, XRD, FTIR High-purity kaolin which contains quartz, dickite is dominant, anatase as minority mineral (p.27) None given Fourrier 2008 Ceramic vessels East Greece and Cyprus Archaic Period (800-480BC)

The relation between Cyprus and East Greece based on ceramic vessel circulation

None given Slip layer in red and white None given Costello & Klausmeyer 2013 Two terracotta statues

Canosa, Italy 3rd_{century BC} _{Identification of}

materials and methods used during manufacturing and restoration

SEM-EDX and TGA White slip made of kaolin (p.380)

Coating applied after firing Kolia 2014 Several terracotta reliefs

Helike, Greece Archaic Period (800-480BC)

Stylistical research and research into the manufacturing process

None given Mostly slip and sometimes a clay wash None given Kakoulli 2017 Two polychrome terracotta vases

Canosa, Italy 3rd_{century BC} _Production

techniques of vases and terracotta figurines

XRF Kaolinite (p.109) Coating applied after firing (p.109)

(25)

2.1.1 Forming the Object

Collecting Clay

The clay that was used for the production of terracotta objects would have been locally collected. Most clay that is available and can be extracted is secondary clay which is formed from degraded stone and transported by rivers or wind. Most iron-rich earthenware clays are suitable for making low-fired terracotta figurines. This clay is earthenware clay and contains impurities such as iron oxide, and often coarse inclusions such as small stones or sand.18_{According to Lunsingh Scheurleer and Leyenaar-Plaisier the clay would be washed}

to remove coarse material that was present in the clay, a process that is called levigation.19

The clay would then be kneaded to remove air bubbles and make the clay homogenous. Larger particles would have been separated from the clay by mixing it with water and waiting for the heavy particles to settle.20_{Both Lunsingh Scheurleer and Leyenaar-Plaisier}

mention that after washing, sand, mica or crushed pottery was added to the clay to limit its shrinkage and to prevent cracking during firing.21

Moulding the Clay

Terracotta objects were made by a terracotta potter, also called a koroplast.22_{During the}

Bronze Age, mainly solid terracotta objects were created, formed by hand. From the eighth century BC onwards, the technique developed and objects started to be made in moulds. These objects, which were hollow, require less clay than solid objects and are easier to fire.23_{To produce these moulds, first a clay object would be made, or an already existing}

object would be used. Over this matrix, a layer of clay, the patrix, would be folded. After drying, the patrix could be used as a mould for the production of multiple objects. These moulds would be filled with one piece of clay, or first a thin layer of clay would be added after which the rest of the clay was pressed into the mould. The extra layer of clay was

18_{Daniel Rhodes, Clay and Glazes for the Potter (Mansfield Centre, CT: Martino Publishing, 2015), 11. &}

Lunsingh Scheurleer, Grieken in Het Klein, 9. & Grimshaw and Searle, The Chemistry and Physics of Clays (London: Ernest Benn Limited, 1960), 333.

19_{Joseph V. Noble, "The Technique of Attic Vase-Painting," American Journal of Archaeology 64, no. 4}

(1960): 313.

20_{Noble, "The Technique of Attic Vase-Painting," 313.} 21_{Leyenaar-Plaisier, Griekse Terracotta's, 4.}

22_{Carl Maria Kaufmann, Graeco-ägyptische Koroplastik: Terrakotten Der Griechisch-römischen Und}

Koptischen Epoche Aus Der Faijûm-Oase Und Andren Fundstätten (Leipzig, 1915), 19-20. & Eleni Hasaki,

"Ceramic Kilns in Ancient Greece: Technology and Organization of Ceramic Workshops," PhD diss., University of Cincinnati, Diss, 2006, abstract in 73.

(26)

sometimes used to create a smooth layer over the object.24_{Often only the front of the}

object was made using a mould. The back would be formed by hand.25

From the Hellenistic period onwards, plaster moulds were also in use. During the Hellenistic period also more complex figures were created, often made out of several pieces. These pieces were adhered together with clay slip before firing.26

Finishing the Object

After the object was moulded, an air hole was cut in the back to prevent the ceramic from cracking during firing due to moisture that was trapped in the object. Furthermore, this air hole allowed heat to reach the inside of the object. Afterwards, final touches would be made by the koroplast and the object was dried. At this stage an object would shrink up to ten per cent.27

Firing

Figurines were fired before or after the application of a coating (table 3). Firing was done in three stages in a kiln. While Kaufman states that firing was in kilns in which temperatures could go up to a maximum of 1150 degrees Celsius,28_{Rice claims that terracotta was mostly}

fired to a temperature between 800 and 900 degrees Celsius.29_{Objects could be placed a}

kiln by using kiln supports.30_{First, the objects would be fired in an oxidising state in a}

temperature between 750 and 950 degrees Celsius. At this stage, air was allowed in the kiln. Secondly, the air flow would be locked in a reducing stage. Finally, the air would be let back in, giving the ceramic core a red colour.31_{After firing, a coating layer would be applied,}

if this had not already been applied before firing, and organic or mineral pigments would be added as decoration.32

24_{Leyenaar-Plaisier, Griekse Terracotta's, 4. & Lunsingh Scheurleer, Grieken in Het Klein, 11.}

25_{Leyenaar-Plaisier, Griekse Terracotta's, 12. & Brijder, "Griekse Godinnen En Meisjes in Het Klein, 57.} 26_{Leyenaar-Plaisier, Griekse Terracotta's, 5-6.}

27_{Lunsingh Scheurleer, Grieken in Het Klein, 9 & 11. & Brijder, "Griekse Godinnen En Meisjes in Het Klein,"}

57.

28_{Kaufmann, Graeco-ägyptische Koroplastik, 19-20. & Hasaki, "Ceramic Kilns in Ancient Greece.”} 29_{Rice, Pottery Analysis, 5.}

30_{Hasaki, "Ceramic Kilns in Ancient Greece."} 31_{Lunsingh Scheurleer, Grieken in Het Klein, 12.}

(27)

2.1.2 Coating Materials

Different types of coating on ancient Greek terracotta objects are discussed in literature, which can be divided into clay-based and chalk-based coatings (table 3).

Slip Coating

The first, and most often discussed process, is the application of a slip coating. Slip, also known as engobe or barbotine, is a suspension of clay in water. Coatings of slip consist of clay minerals to which colorants or pigments can be added.33_{According to Versluys (1993),}

who discusses modern pottery techniques, slip can be made by either adding water to clay or by using clay powder, which is obtained from clay that has completely dried and mixed with water.34_{This slip could be made of the same clay used for the ceramic body, to which}

water and colorants are added. However, if lighter colours were needed as a background, for instance when colours would be applied, white clay for instance containing kaolinite (Al2Si2O5(H2O)4), could be used in which no colouring impurities are present. Kaolinite is a

white or grey clay mineral which is the chief constituent of kaolin.35_{Kaolin is defined as “a}

fine soft white clay, resulting from the natural decomposition of other clays or feldspar”.36

Kaolin mines are present both in Italy as well as in Greece. When writing about Greek figurines, Brijder claimed that white colourants could be added to these coatings but has not specified what exactly.37_{The viscosity of this slip coating can be adjusted by adding or}

extracting water from the mixture.38_{After application, the object would be dried and fired.}

Another possibility is that the coating would be applied after firing.

The presence of a slip coating is discussed by several authors. Richter, Lunsingh Scheurleer Leyenaar-Plaisier and Brijder39_{for instance, mention the presence of a slip coating that was}

applied before firing. Brijder specifies the presence of aluminium silicate (kaolinite) in this slip coating. These four authors, who have written about ancient Greek terracotta objects in general, did not support their claims with chemical analysis and focus on the historical and stylistic aspects of ancient Greek terracottas.

33_{Rhodes, Clay and Glazes for the Potter, 250. & Luk Versluys, Het Kleiboek: Oude Technieken En Nieuwe}

Mogelijkheden (Leuven: Kritak, 1993), 231. & Grimshaw and Searle, The Chemistry and Physics of Clays,

56-57.

34_{Versluys, Het Kleiboek, 232.}

35_{Oxford English Dictionary, “Kaolinite,” accessed June 6, 2019,}

https://en.oxforddictionaries.com/definition/kaolinite.

36_{Oxford English Dictionary, “Kaolin,” accessed June 6, 2019,}

https://en.oxforddictionaries.com/definition/kaolin.

37_{Brijder, "Griekse Godinnen En Meisjes in Het Klein," 57. & Rye, Pottery Technology, 41.} 38_{Rye, Pottery Technology, 20 & 41. & Versluys, Het Kleiboek, 231-232.}

39_{Leyenaar-Plaisier, Griekse Terracotta's, 6. & Lunsingh Scheurleer, Grieken in Het Klein, 13. & Brijder,}

"Griekse Godinnen En Meisjes in Het Klein," 57. & Gisela M. A. Richter, "A Greek Terracotta Head and the "Corinthian" School of Terracotta Sculpture," American Journal of Archaeology 52, no. 3 (1948).

(28)

A slip layer applied after firing is also discussed. Rinuy and Schweizer,40_{who researched}

terracotta vases from 3rd_{century BC Canosa applying Atomic Emission Spectrography (AES),}

X-ray Diffraction (XRD) and thermogravimetry (TGA), concluded that a slip made of kaolinite had been applied to the vases after firing. The results of this research have however not been published and merely the presence of a non-fired kaolin slip is mentioned. Furthermore, Costello and Klausmeyer41_{have researched two terracotta}

statues from Canosa, dating to the 3rd_{century BC, using Scanning Electron Microscopy with}

Energy Dispersive X-ray analysis (SEM-EDX) and TGA. The results from SEM-EDX analysis showed the presence of aluminium silicate, indicating that the surface coating consisted of kaolin. TGA analysis showed that the clay layer had not been fired. Finally, Kakoulli,42_who

discussed similar terracotta vases, also from Canosa and dating to the 3rd_{century BC,}

applied X-ray fluorescence (XRF). This research also concluded that the coating was made of kaolinite and was applied after firing.

In addition, Merker, Fourrier, Kolia, Bordignon and Higgins43_{discuss the presence of a slip}

layer, not mentioning the time of application. In the article by Merker, the focus was on the development of Corinthian Archaic terracotta figurines, Fourrier has researched ceramic vessels from East Greece and Cyprus in archaeological research and Kolia has discussed terracotta reliefs from Helike (Greece), dating to the Archaic Period (800-480BC). Bordignon, who researched white-on-red potsherds from Cerveteri from the Orientalising Period (800-600BC) and Archaic Period (800-480BC), analysed the surface coating that is present on the shards.44_{The analytical techniques applied included micro-Raman}

spectroscopy, XRD and Fourier Transform Infrared Spectroscopy (FTIR). XRD showed the presence of dickite, which is a polytype of kaolin. Even though calcite was also detected in the terracotta body, it was believed that this calcite was probably present due to impurities in the clay or the preparation method of the paint. Higgins used XRD analysis on the polychrome decoration and surface coatings of several Greek terracotta objects dating to different periods. The white coating from one of these objects, a terracotta figurine from Tanagra (Italy) dating to 380BC, was found to have a clay layer containing a small amount of gypsum. The exact results of this analysis were not included in the publication.

40_{Rinuy and Schweizer, "Analysis of the White "ground", 255.}

41_{Susan D. Costello and Philip Klausmeyer, "A Re-united Pair: The Conservation, Technical Study, and Ethical}

Decisions Involved in Exhibiting Two Terracotta Orante Statues from Canosa," Studies in Conservation 59, no. 6 (2013).

42_{Ioanna Kakoulli et al., "Application of Forensic Photography for the Detection and Mapping of Egyptian}

Blue and Madder Lake in Hellenistic Polychrome Terracottas Based on Their Photophysical Properties," Dyes and Pigments 136 (2017): 107.

43_{Merker, "Corinthian Terracotta Figurines," 234. & Fourrier, "East Greek and Cypriote Ceramics," 134. &}

Kolia, "Archaic Terracotta Reliefs from Ancient Helike," 431. & Francesca Bordignon et al., "The White Colour in Etruscan Polychromes on Terracotta: Spectroscopic Identification of Kaolin," Journal of Cultural

Heritage 9, no. 1 (2008): 27. & Higgins, "The Polychrome Decoration Of Greek Terracottas," 273.

44_{Francesca Bordignon et al., "The White Colour in Etruscan Polychromes on Terracotta: Spectroscopic}

(29)

Coating Containing Chalk or Calcite

Calcite, also known as chalk (CaCO3) has also been mentioned as a component of surface

coatings on ancient Greek terracotta figurines. Calcium carbonate can occur in various crystal forms including calcite or aragonite or can be present assedimentary rocks such as chalk and limestone. It is manufactured to produce different materials, such as plaster.45

Kaufmann, discusses the history of koroplastic, in which the production of these objects is been described, and mentions the presence of a layer of Kalkmilch (chalk milk) on the objects, which was applied after firing.46_{Oliver discusses the presence of a ‘white lime}

wash’, ‘white ground’, or ‘whitewash’ that was applied after firing in his research into ceramic objects from 3rd_{and 4}th_{century BC Canosa, Italy.}47_{It is unknown whether both}

Kaufmann and Oliver believed this or if they were just describing what they saw, as their research did not include any chemical analysis.

The same counts for Knoop and Lulof, who state that the coating present on 6th_century

architectural terracottas from Satricum (Italy) was white because it contained large quantities of calcite.48_{The coating they both discussed has been described as a ‘slip’ and}

according to the authors contained clay which was mixed with calcite. It is stated that these coatings were applied after firing. Neither authors used instrumental analysis in their research.

In conclusion, most literature mentions the presence of slip coatings that have been applied to terracotta objects. The majority of the authors that discuss the application of slip before firing did not undertake chemical analysis. This is in contrast to the sources that mention the application after firing who did apply instrumental analysis. However, since this group only consists of three research projects, one cannot assume that clay coatings were always applied after firing. Furthermore, the sources that mention the presence of chalk or calcite have not conducted chemical analysis or have not included this research in their publication. This makes it difficult to confirm whether they mean that chalk and/or calcite are indeed found as a surface coating on Greek terracotta figurines.

45_{Grimshaw and Searle, The Chemistry and Physics of Clays, 281. & W. David Kingery, Pamela B. Vandiver,}

and Martha Prickett, "The Beginnings of Pyrotechnology, Part II: Production and Use of Lime and Gypsum Plaster in the Pre-Pottery Neolithic near East," Journal of Field Archaeology 15, no. 2 (1988): 219 & 221. & R. J. Gettens, E. W. Fitzhugh, and R. L. Feller, "Identification Of The Materials Of Paintings: Calcium Carbonate Whites," Studies in Conservation 19 (1974): 157.

46_{Kaufmann, Graeco-ägyptische Koroplastik, 19-20.}

47_{Andrew Oliver, The Reconstruction of Two Apulian Tomb Groups, 9-23.}

48_{Patricia S. Lulof, "Monumental Terracotta Statues from Satricum: A Late Archaic Group of Gods and}

Giants," PhD diss., Proefschrift Amsterdam, Universiteit Van Amsterdam, 1991, abstract in 118. & Knoop, Antefixa Satricana, 21.

(30)

2.1.3 Coating Application

Some authors include information on the way the surface coating would have been applied. Richter, Higgins and Knoop state that the slip coating could have been applied by using a brush or by dipping the object in the slip.49_{Richter mentions the application of two ‘white}

coatings’ on top of each other.50

None of these sources, however, give clear reasons for why they believe that these application methods were applied and many of these authors discuss terracotta that is different either in date or in origin from the objects discussed in this research. Therefore, it cannot be said categorically which methods were used on the objects used for the application of surface coatings on Greek terracotta figurines from the 4th_{and 3}rd_century

BC.

Thermogravimetry (TGA) analysis was conducted by Rinuy and Schweizer and Costello and Klausmeyer who researched 3rd_{century BC Canosa vases. This analysis concluded that}

unfired kaolinite was applied onto the ceramic after firing. No other sources discussing the application of a slip coating have included instrumental analysis in their research focussing on the moment in production that it was applied.

All sources discussing a coating with chalk or calcite have mentioned the application after firing. When heated between 750 and 850 degrees Celsius, calcium carbonate will decompose, making it impossible for a layer of chalk or calcite to still be present on ceramic after firing.51_{This would be the same if chalk was mixed in clay in high proportions.}

49_{Richter, "A Greek Terracotta Head," 332. & Higgins, "The Polychrome Decoration Of Greek}

Terracottas," 273. & Knoop, Antefixa Satricana, 21.

50_{Richter, "A Greek Terracotta Head," 335. & Knoop, Antefixa Satricana, 21.}

51_{Harry Fraser, Ceramic Faults and Their Remedies (London: A. & C. Black, 2005), 17-19. & Prudence M.}

Rice, Pottery Analysis: A Sourcebook (Chicago: University of Chicago Press, 2015), 97-98 & 103. & Owen S. Rye, Pottery Technology: Principles and Reconstruction (Washington, D.C.: Taraxacum, 2002), 32.

(31)

2.2 Factors Influencing the Loss of Surface Coatings

Several causes for the detachment or loss of surface coatings from Greek terracotta sculptures have been discussed in the literature. A combination of the chemical composition of the coating and ceramic, the production process, and cleaning seem to affect the loss of surface coatings.

2.2.1 Production

Application

The moment in the production process at which a surface coating is applied and the viscosity of the coating have an influence on the susceptibility of any coating to detachment. These factors can result in differential shrinkage. For example, a slip layer which is a mixture of clay and water that is applied before firing, will have a higher water content than the clay body. The slip will dry faster than the clay object itself, which contains less water. This results in a faster shrinkage rate of the slip than that of the object.52_This

can cause stresses at the interface and lead to cracking and a weakened interface.53_{If the}

ceramic object is completely dry when a slip is applied, as is the case with coatings applied after firing, the object will extract water from the applied slip, so that the slip cracks due to loss of water.54

However, when slip is applied to partly-dried clay that is ‘leather-hard’, the first stage of water evaporation will have finished so minimal differential shrinkage occurs, and the object absorbs enough, but not too much, water.55

If the clay is too wet, a slip coating will have difficulty to attach to the object because of the lack of water absorption and might run off the surface.56_{Other factors that can influence}

differential shrinkage in slip decoration is if the slip is not stirred properly before application. This will lead to larger particles that may have settled, causing differences in the shrinkage of the slip in certain areas.57_{If thick layers of slip are applied, the slip may}

crack due to the uneven drying, forming a network of more or less hexagonal lines.58 It has

been observed that flaking during firing occurs most often on convex areas as well as on edges of an object.59_{The reason for this could be that the coating layer has been applied}

52_{Fraser, Ceramic Faults and Their Remedies, 8-9.} 53_{Rhodes, Clay and Glazes for the Potter, 288.} 54_{Versluys, Het Kleiboek, 233.}

55_{Rhodes, Clay and Glazes for the Potter, 288. & Versluys, Het Kleiboek, 240.} 56_{Rye, Pottery Technology, 24.}

57_{Fraser, Ceramic Faults and Their Remedies, 29.}

58_{Rye, Pottery Technology, 41 & 43. & Fraser, Ceramic Faults and Their Remedies, 8-9 & 29.} 59_{Fraser, Ceramic Faults and Their Remedies, 30.}

(32)

too thinly or that the evaporation rate of the clay is too fast in these areas due to the thin clay layer.

Slip coatings, which consist of clay mixed with water, could be applied before or after firing. The difference between fired ceramic and unfired clay is that ceramic becomes non-plastic, hard and brittle due to firing, forming a vitrified layer.60_{Not only would a fired slip layer}

therefore be expected to be harder and less vulnerable, for instance to abrasion than an unfired slip layer, the way these layers are bonded differs. Namely, an unfired slip will have slightly been absorbed into the ceramic body, but has after that simply dried and is susceptible to dissolution in water. The clay in a fired layer, on the other hand, will have gone through chemical changes in the kiln, irreversibly fixing the structure in the clay particles and resulting in a physiochemical bond with the terracotta. A surface coating containing chalk would be applied after firing and would therefore be very susceptible to abrasion and water.

2.2.2 Morphology and Chemical Composition of Ceramic and Coatings

Differential Shrinkage

During the production of terracotta objects, several stages occur during which shrinkage of the clay object and a fired-on clay slip coating occurs. While shrinkage will occur during firing, it is greatest during the drying process. Shrinkage of the clay object occurs when the water that is present in the clay to make the clay plastic (malleable) evaporates. This happens in two stages during drying. First of all, water inside the pores of the clay will migrate to the surface of the object due to capillary action. Secondly, the water that is present at the surface will evaporate out of the object.61_{The evaporation of water from}

the clay causes an object to shrink. Shrinkage will occur mainly in the first stage of the drying process, until the object is considered ‘leather-hard’. During the second stage, shrinkage will only occur at a minimal range.62

When a surface coating is applied onto a ceramic object, differential shrinkage can occur between the coating and the ceramic. This happens when the shrinkage rate of the surface coating and the ceramic object differ from each other. The shrinkage rate of clay or a slip coating is based on the size of the clay particles and inclusions. Another reason is differential packing of clay particles in an object. Areas where the particles are packed in a

60_{Rice, Pottery Analysis, 473.}

61_{Fraser, Ceramic Faults and Their Remedies, 65.}

(33)

flocculated way,63_{it will dry faster than in areas where the particles are deflocculated}64_.65

The clay particles in slip coatings are mostly oriented in a deflocculated way, meaning that they have the same charge and occur in dispersed suspensions instead of all having the same orientation.66_{Different clay types have different shrinkage rates. For instance, little}

shrinkage occurs in clay that is rich in kaolinite.67

Due to differences in shrinkage rates, stresses can occur between a slip coating and the object, which will cause cracks in the coating as well as stresses at the interface of the slip and ceramic.68_{In general, the most suitable slip would have a composition similar to that}

of the clay body. The grain size of a slip should, however, be finer to prevent differential shrinkage.

During firing problems with differential coefficient of expansion can occur which leads to the shrinkage and expansion of the ceramic and slip at different rates when the composition of the coating and ceramic are different. At five hundred degrees Celsius, mechanically-bound pore water will evaporate from the object, causing the clay to shrink about one per cent. At 573 degrees Celsius quartz that is present in the clay will rearrange in a different order so it increases in volume, causing the object to expand. This point of ‘quartz inversion’ has to happen slowly to avoid cracking.69

Porosity

The susceptibility of a surface coating to loss, whether containing chalk or not, will be different depending on the porosity of the clay. Namely, the higher the porosity of a slip, the more friable it is and therefore it is more susceptible to mechanical ware and abrasion.70_{The porosity of ceramic is related to the clay composition and morphology as}

well as the temperature at which an object has been fired. The higher the firing temperature, the lower the porosity of a clay body and the more the clay will sinter. Sintering occurs when clay is fired between 800 to 1000 degrees Celsius when the clay particles fuse.71

63_{“Flocculation is the agglomeration or coming together of particles in a suspension, such as a slip, forming}

"floes" and causing the suspension to thicken or settle.” Rice, Pottery Analysis, 476.

64_{To deflocculate means “to disperse a fine clay suspension so that particles repel each other and the}

substance becomes more fluid.” Rice, Pottery Analysis, 475.

65_{Rhodes, Clay and Glazes for the Potter, 251.} 66_{Rice, Pottery Analysis, 77.}

67_{Rice, Pottery Analysis, 67.}

68_{Rhodes, Clay and Glazes for the Potter, 13.}

69_{Prudence M. Rice, Pottery Analysis: A Sourcebook (Chicago: University of Chicago Press, 2015), 103. &}

Rhodes, Clay and Glazes for the Potter, 17-18.

70_{Susan Buys and Victoria Oakley, The Conservation and Restoration of Ceramics (London: Routledge,}

2011), 19 & 22.

(34)

In a ceramic with a large number of open pores, liquids will be transported faster by capillary action, meaning that when in contact with water, more will be drawn into the object, which can lead to staining or saturation of the clay which leads to expansion.72

2.2.3 Coating Composition

The composition of a surface coating does not only influence the porosity of a surface coating and the shrinkage rate, also the other characteristics related to stability of a surface coating differ depending on its chemical composition.

When a layer of unfired slip comes in contact with water, the bonds between the clay particles will break, leading to loss of coating.73

_{When chalk or calcite is added to a slip, the}

calcium reduces clay shrinkage during drying.74_{Moreover, when calcium carbonate is}

applied to clay, the amount of water surrounding clay particles is reduced. This causes the clay particles to move around less easily, forming cement-like compounds. The addition of lime would therefore make a slip layer harder and more stable.75_{This coating layer, even}

though similarly attached to the surface as an unfired low-calcium slip layer, would be expected to dissolve less quickly in water and be harder and therefore less vulnerable to abrasion.

Furthermore, all clay and slip coatings contain fluxes, which are elements that reduce the melting temperature of silica-rich materials. These fluxes can occur in different quantities.76

This means that when firing, the temperature and speed at which clay starts to sinter is influenced by the type and amount of fluxes that are present in the clay. Calcium-rich clay, for instance, shrinks more on firing than low-calcium clays because of its higher thermal expansion coefficient.77_{If the type and proportion of fluxes in the ceramic body and the}

surface coating differ, tension at the interface can develop during firing.

72_{Susan Buys and Victoria Oakley, The Conservation and Restoration of Ceramics (London: Routledge,}

2011), 19.

73_{Sebastian Teir et al., "Stability of Calcium Carbonate and Magnesium Carbonate in Rainwater and Nitric}

Acid Solutions," Energy Conversion and Management 47, no. 18-19 (2006): 3060-3061.

74_{Grimshaw and Searle, The Chemistry and Physics of Clays, 282.}

75_{Marek Lichtarowicz, "Calcium Carbonate," The Essential Chemical Industry Online, accessed May 10,}

2019, http://www.essentialchemicalindustry.org/chemicals/calcium-carbonate.html.

76_{Rhodes, Clay and Glazes for the Potter, 251.}

77_{M.s. Tite, "The Production Technology of Italian Maiolica: A Reassessment," Journal of Archaeological}

Surface Coatings on Terracotta Objects from Boeotia and Taranto 400 - 200 BC - the influence of composition on the susceptibility for detachment