Paint Pigment on Epigraphic Squeezes: A Case-Study from the Etruscological Collections of Olof August Danielsson

: The practice of wet squeezing as a means of creating epigraphic facsimiles was widespread during the 19th and 20th centuries, but often negatively impacted the original inscriptions, particularly in cases where paint pigment was present. Depending on the precise method used, the pigment may transfer from the original to the squeeze. This article presents a case-study of four such squeezes made by Swedish Etruscologist Olof August Danielsson in 1886 of a funerary tile from Chiusi, as part of the preparatory work for Corpus Inscriptionum Etruscarum. This material makes it possible to study epigraphic polychromy in cases where pigment on the original has been damaged through squeezing. In addition, it illustrates how squeezes can giveinsight into the history of individual objects, asthe squeezes, which show both the carving and the paint, cast light on the various proposed readings of the tile ’ s Etruscan inscription (CIE 1973/ET Cl 1.1478).


Introduction
Squeezing (sometimes referred to as wet squeezing) is a method by which a negative facsimile of a carving is created in paper.Likely developed already in the sixteenth century, it became widely used by epigraphists during the nineteenth century. 1 Still today, squeezing is taught to students of epigraphy.2While there are still scholars who make use of it, squeezing has generally fallen out of favor. 3In the 1940s, photography was seen as a complement to squeezing,4 but the development of more reliable cameras (in particular digital cameras which cut out the cost of development and, later, the advent of smartphones, which makes it possible to take high-quality photographs without having to carry heavy camera equipment) likely contributed to squeezes being phased out.However, the impact of squeezing on the inscriptions has also been a concern.Many institutions housing inscriptions, such as the Athens Epigraphical Museum and the British Museum, no longer allow squeezes to be made of the inscriptions in their collections. 5 squeeze is made by placing a piece of wet paper over the inscription, which is then beaten into the carving with a special brush.6W.M.F.Petrie's description of the process captures how vigorous it could be: [The paper is] gently beaten with a spoke-brush until it is pushed into the hollows [of the stone].
[…] Finally, a severe beating is given to the whole, as violent as can be done without tearing the paper.The paper should be pulped on the stone, and driven into every crack and porosity; using a second, and even a third, sheet to bind it together.The pulp in the hollows should be kneaded in with the sharp edge of the brush-back, using the whole weight of the body to force it home. 7 is no surprise, then, that originals could incur damage through the process.One of the main issues is the impact squeezing has on any paint present on the inscription.The risks to pigment have been known for some time.When describing the damage done to Egyptian wall-paintings in her 1878 travelogue, Amelia Edwards describes how "the student of Egyptology, by taking wet paper 'squeezes', sponges away every vestige of the original colour". 8Petrie points out that "on all coloured work, and many kinds of tender stone, wet squeezing is a crime […].Fatuous tourists and brazen students have wrecked innumerable monuments by wet squeezing".9At the time Petrie was writing, wet squeezing had been prohibited in British-controlled Egypt. 10 It is not a coincidence that these observations came from writers whose focus was ancient Egypt, where polychromy was not viewed with the same suspicion as in relation to Greece and Rome. 11 This chromophobia, as it has sometimes been called, 12 goes back to Johann Joachim Winckelmann, who argued that "the essence of beauty consists not in colour but in shape" and that "a beautiful body will […] be the more beautiful the whiter it is." 13Remnants of paint on Greek and Roman marble sculptures and reliefs would often be removed, sometimes with chisels. 14The same aversion to color has not been as present in the field of Etruscology, but much of the research, like the recent study of polychromy in Greco-Roman sculpture, has focused on figurative art. 15Polychromy in epigraphy is overall still under-researched. 16here are two stages of the squeezing process which pose a risk to pigments.The one that most critics focus on is the initial cleaning of the stone.Petrie describes that before a squeeze can be made, "the stone must be thoroughly cleaned and soaked". 17Almost 80 years later, Gordon instructed that "one cleans the inscribed surface with a small cleaning brush, with or without water, so as to remove all dirt, incrustations of lichen, etc., and (modern) paint". 18He goes on to claim that this will not damage any ancient miniumred leadpresent on the stone.Numerous scholars see this as the point at which the worst damage is done, envisioning any pigment being brushed off the inscription or discarded along with the water used for washing. 19lthough the risk to pigment through initial cleaning is considerable, there is also evidence that the actual act of squeezing impacts it.When this occurs, the pigment is indeed removed from the original, but it is not lost.It is instead transferred to the paper squeeze.While this has been mentioned in passing in relation to squeezes of material from Egypt, Mesopotamia and Persia, it has never been the subject of any in-depth research. 20his article considers pigment transfer during the squeezing process through a case-study of four squeezes of an Etruscan funerary tile.The squeezes were made by the Swedish philologist Olof August Danielsson in 1886, and are currently kept in his archive at the Uppsala University Library as part of a squeeze collection comprising 46 large boxes.21

The Four Squeezes
In 1884, Danielsson and his colleague Carl Pauli, a German scholar and schoolmaster, agreed to start work on a new Etruscan corpus.Pauli had had the ambition to create a corpus for Etruscan inscriptions in the mould of the CIL for several years.It had initially been planned as a joint venture between him and Wilhelm Deecke, with whom Pauli had become acquainted through his championing of Etruscan as a non-Indo-European isolate.Together, they submitted a request for funding to the Berlin Academy in 1881, but it was denied.A year later, their collaboration and friendship came to an abrupt end when Deecke retracted his statements on the classification of Etruscan.The following year, Pauli came into contact with Danielsson, who had been awarded his doctoral degree only three years prior.After some correspondence, they agreed to start work on what would eventually become the Corpus Inscriptionum Etruscarum. 22Pauli had already conducted autopsies of Etruscan inscriptions in Leiden in 1883, but for the project to be truly successful, the editors would have to travel to Italy.Pauli and Danielsson arrived in Italy in November 1885.Pauli returned to Germany before the end of the year.Danielsson continued his epigraphic fieldwork alone until July 1886, travelling around Italy and making notes, sketches and, most importantly, squeezes.
The collection is estimated to contain around 9000 squeezes.As itemization of the collection is still ongoing, the first nine boxes of 46, containing almost 1900 squeezes of inscriptions with CIE numbers up to 2200, will be used as a sample. 23Of the inscriptions CIE 1-2200, a total of 1365, around 62%, are represented by facsimiles in the collection.Wet squeezing is by far the most common technique, representing 90% of the collection, but other methods like tracing and rubbing were also used.Most of the facsimiles, over three quarters of the sample, were made by Danielsson. 24 The facsimiles collected as part of the CIE project were carefully marked up to facilitate organization.Those made by Danielsson, both in 1886 and during later trips, contain two types of annotation: notes written in pencil on the squeeze and a series of labels with relevant information written in ink.Based on analysis of the sample squeezes, it is possible to make an informed guess how the process occurred.
Danielsson would autopsy an inscription, make notes and make one or more squeezes.At some point after the squeeze was dry, he would add information in pencil along the edges where the paper had not been subject to the squeeze brush.On some squeezes, these notes appear to have been written against an uneven surface.As this would not be the case if he had a desk or even a book to lean the squeeze against, it seems likely that he made these while still in the field, resting the squeeze against some nearby object that was reasonably (but not completely) flat, possibly one of the objects he had just autopsied.
The first of the four squeezes that will be discussed further on in this article, 1393,25 is marked in the typical way: "Chiusi 100 16/IV 86"the city where the squeeze was made, a number and the date of the squeeze's creation.The number 100 is in reference to the inscription rather than the squeeze, as can be seen in squeeze 1396, where it reoccurs, modified by a Roman numeral II. 26Often, though not in this case, Danielsson would add the letter "D." after the date, to identify himself as the squeezemaker.This goes back to when Pauli and Danielsson made squeezes together, and would mark them with their final initial.Danielsson would often add a reference to the Corpus Inscriptionum Italicarum (CII), the most commonly used corpus for Etruscan at the time.These references are formatted in the same way as those found in CIE, with "Fa." used to refer to Ariodante Fabretti's main volume of CII and three supplements, and "Ga." used to refer to Gian Francesco Gamurrini's appendix.This implies that Danielsson had these volumes with him on his trip.As these volumes were quartos, they were far more portable than many other corpora.However, the uneven quality in the handwriting is generally not seen in the CII references, indicating that these were not added in the field.
The squeezes were sent to Pauli, who equipped them with labels on the obverse of the squeeze. 27This sequence of events is strongly implied by the fact that it is only ever Pauli's handwriting on the labels. 28As chief editor, he was the one who organised and used the squeezes, making it likely that he designed the system himself.Five different labels were used, with specific designs and inks reserved for each piece of information.The labels recorded the following: -Ancient or modern location, corresponding to the chapter or subheading under which the inscription would appear in CIE.-Information about the creation of the squeeze, including when, where and by whom it was made.-References to previous editions.
-Inscription number, frequently (but not always) the same as the CIE number.
-Squeeze number.Squeezes were numbered in the order they were made, based on field-notes where necessary. 29anks to the information on the squeezes as well as Danielsson's notes, we can piece together the process of autopsy, note-taking and squeezing.
Danielsson arrived in Chiusi on the 13th of April, 1886.At this point, he would have been an experienced squeeze-maker, having spent almost six months with this as his main task.If Danielsson kept a diary, it has not been preserved, but the notes for much of the CIE field-work are kept in the Uppsala University Library.On the title page for the notes of Chiusi, Danielsson made a note about the weather conditions.It had been windy, so the squeezes of inscriptions displayed outdoors in the Succursale had etwas schlechter geworden, "turned out a little worse". 30He also complained of the humidity inside the Chiusi museum, which had complicated the squeezing work.It was inside the museum that the tile at the heart of this casestudy was located.
Funerary tiles were in use in the Clusine territory from the third century BCE to the second half of the first century BCE.Their purpose was to seal the niches where the burial urns were kept in dromos graves. 31The provenance of this particular funerary tile (Figure 1; museum inventory number 176, national inventory number SA FI65808, TM 150650/CIE 1973/ET2 Cl 1.1478) is unknown.It is made from kneaded clay and measures 61.5 cm in height and 45 in width at its widest point. 32The vertical sides of the tile are raised, providing a border for the epigraphic space.The tile bears two lines of text in the Clusine alphabet (the reading of which will be discussed below), seemingly written by a layperson, something that was standard for funerary tiles.While most funerary tiles are unadorned, this tile features a carving of two anchors. 33Each anchor has a ring both on the stock and at the crown, a feature it shares with the mosaic anchor in the House of the Anchor in Pompeii (VI 10.7) as well as the iron anchor found in Lake Nemi. 34Gamurrini (CII App.181) gave the following reading: Generally, Danielsson would dedicate only a few lines to each inscription in his notebook, often covering five or more inscriptions on each page.The notes on this tile, however, takes up two thirds of a page.The reason for this is that Danielsson could not reconcile Gamurrini's reading with what he was seeing on the tile.
Danielsson made four squeezes, one covering the full inscription and three featuring various details (referred to by his notes as Specialabklatsche).The areas 30 Danielsson 1886, 9. 31 See Benelli 2010, 123-4.32 Soprintendenza alle antichità d'Etruria, catalogue card 65808.I am indebted to Fabrizio Vallelonga for providing me with documentation concerning the tile as well as numerous photographs.33 One other tile (CIE 2864/ET 2 Cl 1.2476) with a similar carving survives, though featuring only one anchor.The anchor on this tile is very similar to the anchors on the tile under discussion in this article, down to details such as the way the shank extends into the upper ring.It is tempting to posit some relationship between these two tiles, but with no known provenance, this would only be speculation.34 Speziale 1931, 311, pl.i-ii.See Nautical Archeology Society 2008, 200 on terminology relating to anchors.
covered by each squeeze are given in Figure 2. Generally, Danielsson would only make one squeeze per inscription.Analysis of the sample shows that of the 867 inscriptions represented, only 22.8% had more than one facsimile made.Inscriptions that were copied more than twice make up only 6.2%.Making four squeezes, particularly where the original was small enough to capture with only one squeeze, was therefore out of the ordinary.It appears that Danielsson made notes and squeezes simultaneously, as the notes contain one sentence in Swedish which is now sandwiched between two lines that follow on one another, implying that he wrote this below what he had written before, and then had to write around it.Considering  35 It is possible that Pauli could read Swedish, as he contacted Danielsson after reading his review of Pauli's Altitalische Studien in the Swedish-language Nordisk Revy (Danielsson 1883, see Wikander and  Wikander 2003, 19).However, he may simply have been aware of the review and the fact that it was positive.Considering that the sentence in Swedish in Danielsson's note would be of little use for Pauli in Germany, the more reasonable explanation is that it was written as a reminder to himself.36 Danielsson 1886, 90.37 Danielsson 1886, 90.I could have been spared all this back and forth, several special squeezes and quite the headache if I'd just had the sense and courage to wash off the remains of the red paint as well as the worst of the dirt.

[It is] now entirely clear that the second row [reads] laθi cauśine
The reading on which Danielsson settled is the one that appears in CIE (number 1973): larθi : cauśtine laθi cauśine The inscription was not autopsied in preparation for either edition of Etruskische Texte, and Rix and Meiser give largely the same reading as Pauli and Danielsson (ET Cl 1.1478).The only difference is that they read the word-divider as a single punct (rather than a dicolon) and put the second line in curled brackets to indicate that it is superfluous.They also indicate that cauśtine in the first line is a mistake for cauśline, a gentilicium that occurs elsewhere in the epigraphic record. 38aving looked at Danielsson's process, let us now turn to the squeezes themselves.
Squeeze 1393 (Figure 3) 39 covers the whole inscribed area, measuring 26.6 cm in height and 44.5 cm in width.The reverse of the off-white squeeze is of particular interest.A considerable amount of red pigment from the inscribed text and the anchors has transferred onto the squeeze paper.
It is worth asking how we can know that the red pigment was indeed transferred from the original.There are, however, numerous reasons to be certain that this is a case of paint transfer.While some epigraphists were in the habit of touching up their squeezes using crayons, tracing the letters on either the reverse and the obverse, Danielsson appears not to have done this. 40His notes mention the presence of red paint as well as dirt on the tile.Two other details concerning squeeze 1393 show that the red is definitely transferred from the original.
Firstly, the reverse of the squeeze bears yellowish discoloration, particularly around the carvings, where Danielsson would likely have directed most of his blows with the squeeze-brush.In the museum catalogue, the clay of the tile is described as giallino 'light yellow', which indicates that this discoloration is in fact residue from the object itself.
More sediment, the color of yellow ochre, was caught in the adhesive Danielsson used to secure the paper to the tile, visible in both upper corners of the squeeze.
Secondly, there is a second instance of paint transfer.In the middle of the squeeze, just above the second word of the first line, one can make out the mirrored numeral "28".The evenness of the numerals indicate that they were made with a stencil or stamp.The orientation is due to the fact that squeezes will always give a mirror image of the original (as can be seen by the fact that the Etruscan writing, originally right to left, runs from left to right in all the squeezes).Danielsson's notes again provide an explanation.Along with his own numbering of the squeeze, 100, he has added "Mus.28", indicating that this was the museum catalogue number.It is no longer in use as the museum collection was recatalogued and renumbered when the museum was moved to its current location in 1901. 41A similar way of marking the objects was used after the renumbering, as can be seen by the current catalogue number 176 on the squeeze in Figure 1.
Squeeze 1394 (Figure 4), 42 which, like squeeze 1393, was made with Danielsson's standard paper, focuses on the beginning of line 1 and the whole of line 2.There are faint traces of red pigment in many of the letters, particularly in the letters <θi : ca> in line 1.It can also be seen in <laθi> in line 2.There are splotches of yellowish discoloration, particularly around the anchor, but they are far less noticeable than on squeeze 1393.In the top-right corner, the curve of the numeral <8> can be made out, but with difficulty.
Squeeze 1395 (Figure 5) 43 was made using two pieces of very thin paper placed on top of one another, rather than a single sheet of thicker paper as with the other squeezes.This method makes smaller details more obvious.This squeeze is only 5 cm high but 29 cm wide and covers the second line and parts of both anchors.Although the yellowish tone of this paper makes it harder to see, upon close inspection evidence of paint transfer can be found.There are small, isolated patches of  Paint Pigment on Epigraphic Squeezes discoloration from the tile above the <l>, below the <i> and to the right of the anchor ring.There is red pigment in the letters, particularly the sequence <laθ> and the <u>, as well as the shank of the anchor, but it is far more scattered than in previous squeezes.
The final squeeze, 1396 (Figure 6), 44 measures 38.5 cm in width and 15.5 cm in height.Although it covers a smaller area than the first squeeze, it depicts both lines of the inscription, as well as the upper part of both anchors.In one of the three points of adhesive used to hold up the squeeze, some sediment has become stuck, but the squeeze itself has almost none of the discoloration the other three have.Yellow remnants of the clay are seen primarily under the second line, in an area which was not covered by squeeze 1395 and has therefore been subjected to Danielsson's ministrations less than the rest of the epigraphic space.Elsewhere, the discoloration only appears as specks, such as just under the <u> and in the area above <larθi>.The red pigment that was so prevalent on squeeze 1393 is all but gone.The only visible remnant is a speck of red in the <θ> in line one.All that remains of the museum number is the top curve of the <2>, just above the san in line one.
During the course of the creation of these four squeezes, we can see how less and less pigment is being transferred, as the amount of pigment on the original is decreasing.Considering how squeeze 1393 looks, as well as Danielsson's comment about not having had "common sense and courage" to clean the inscription, it is unlikely that the tile was washed before the first squeeze was made.It is naturally possible that Danielsson washed or otherwise cleaned the tile between making the squeezes, but even if that was the case, we can see how the squeezing removes more and more pigment.
Very little, if any, red pigment remains on the original.Microscopic remnants may still survive, but this would require a more in-depth study.The black numerals <28> are very faint, only visible if one knows exactly where to look for them.The new catalogue number, 176, was painted onto an area of the tile that does not bear any carvings.
These squeezes show the detrimental effect that the process has on paint pigments, but they also provide valuable insights.In the words of Booth: "Considering pigment transference is one of the reasons that squeezes have such a poor reputation as a recording method, it is time to turn that negative into a positive academic feature". 45When taken together with Danielsson's notes, the squeezes show the process by which the CIE editors conducted their work, as well as the work of previous corpus editors.
Squeeze 1393, which contains most of the paint pigment, shines a new light on Gamurrini's reading and line-drawing (Figure 7).Although Gamurrini says he autopsied most of the inscriptions included in the Appendice, it is unclear whether this tile was one of them. 46Unlike in other entries, he does not give a date of autopsy.Whatever the case, the line-drawing is obviously based on the painted words.The <a> in larθi, which is written above the line, was not filled in with paint in squeeze 1393, making it harder to make out.It is completely missing in both Gamurrini's linedrawing and reading.The name cauśine in line two has been painted as caunei, which is again what Gamurrini gives. 47his raises the question when the paint was added.Possible options are: (1) The paint is ancient.
(a) It was added to correct a mistake in the carving, i.e. the carver mistakenly carved cauśine when the intended name was caunei, and caunei was added using paint.(b) It was added by someone who did not know the name of the deceased (and therefore misread the name), but who was still familiar with the Etruscan script.
(2) The paint is early modern or modern and was added at some point after the tile was found (made by someone with enough experience of Etruscan epigraphy to know that women's gentilicia end in -ei).
45 Booth 2018, 206.46 Gamurrini 1880, vi.47 The person who made the line-drawing has drawn the <c> in line 1 as partially damaged, and when presented this way it could be read as a <f>, as Gamurrini does.However, there is no basis, whether in the painted letters or the line-drawing, for Gamurrini's reading lati in line 2.Both show a clear <θ>.
Paint Pigment on Epigraphic Squeezes (3) The paint is a combination of ancient pigment and early modern/modern touchups.
Option (1a) is made less likely by the fact that cauśtine is not corrected to cauśline.It also requires that the tile commemorates two people, rather than just one. 48onsidering the carver's tendency of leaving out letters, e.g.laθi for larθi, it seems likely that both lines refer to the same person.Option (1b), that the paint was added in antiquity but not in conjunction with the carving, is more likely than (1a).We do not know what type of upkeep was done on Etruscan graves, and whether the tiles may have been (re)painted by later generations.If this occurred, it is possible that the mistake was made then.As Etruscan women kept their gentilicia and their children were given the gentilicium of the father, descendants would not have had the same name as a female ancestor.Descendants may also not have been certain whether the tile commemorated one 48 Gamurrini (1880, 23) believes the tile commemorates two women.The addition of curled brackets in both editions of Etruskische Texte around the second line, indicating it is superfluous, indicate that Rix and Meiser believe it commemorates a single woman.person (whose name was repeated twice) or two people (with the same praenomen but different gentilicia).
The fact that the misreading in line two, which led to the carved cauśine being painted as caunei, involves a san may point to the person responsible not being familiar with the Etruscan script.Early modern scholars would often assume < > was the equivalent of a Roman <M>, such as when Passeri, writing in 1767, transcribed < cauślini : auleś (CIE 970/ET2 Cl 1.1029), as "CAIMLINI • AVLEM". 49This could support option (2).However, the same mistake could have been done in antiquity after the local community had started speaking Latin.Misreading cauśine for caunei could have been done by someone who could read the Etruscan script because of the form of the carved letters.A comparison of the painted version from squeeze 1393 and the carved version from squeeze 1396 are given in Figure 8.In the carving, the san is squashed, with the point closer to the <i> being considerably lower than usual.It is thus possible to interpret the san as part of an <n>, with the <i> making up the final vertical stroke.The <n> is in turn misinterpreted as a partial <e>.

Paint Pigment on Epigraphic Squeezes
As the third bar of the <e> is very close to the anchor ring, it would be easy to interpret the vertical stroke of the <e> as an <i>.
Finally, as mentioned under option (3), the tile may have been both in antiquity and after being taken from its original context.The pigment present on the squeezes could thus be a combination of ancient and early modern/modern paint.
In my opinion, the reading of the carved inscription should be: larθi : cauśtine la(r)θi cauś(t)ine The word-divider in line 1 is a dicolon, as given in CIE, rather than a single punct, as in ET.While the top-most punct is not included in the line-drawing in CIE and is clearly shallow, it is visible (both on the photograph of the tile, Figure 1, and Danielsson's squeezes, particularly Figure 6).As mentioned above, I judge it most likely that the inscription is in reference to one person, but this does not necessitate rejecting the second line as superfluous, as Rix and Meiser do.While the vast majority of the 893 funerary tiles from Clusium listed in ET2 commemorate only one person, there exist 18 other examples which bear more than one name, 11 of which repeat the name of one person. 50hat being said, inscriptions where the carving and pigment give different texts present a challenge to editors, and it is worth asking why the carved letters are taken as more primary.It is clear that inscriptions were painted in antiquity, and paint may have been used as a way to correct mistakes in the carving.However, paint was also added to inscriptions in early modern or modern times.The same day as Danielsson made squeezes of CIE 1973/ET2 Cl 1.1478, he autopsied another tile, CIE 1922/ET2 Cl 1.1411, which is described in the CIE as having been neglegentissimo et stultissimo modo et colore et scalpro restauratus "restored in a most neglectful and foolish way using both paint and chisel". 51Here too the carved text and the painted text differ.In most cases, ancient objects are not preserved as time-capsules, but have an afterlife, where they have been handled and potentially manipulated by both devious and 51 Three squeezes of this tile were done (Uppsala University Library, O.A. Danielsson 37:3:2, 37:3:2/  1922/1412, 37:3:2/1922/1413 and 37:3:2/1922/1414), and show a similar pattern to the squeezes of the casestudy, with the paint wearing off more with each squeeze.
well-meaning actors.Without conducting laboratory tests on the paint, we cannot say with certainty what the age of the paint is.

Conclusion
The four squeezes discussed in this article are by no means the only examples of squeezes paint pigments in Danielsson's squeeze collection.In the first nine boxes, there are 162 examples of squeezes bearing paint pigment.Of these, 148 squeezes are of Etruscan inscriptions, 13 are of early Latin inscriptions from Etruria, and one is of a Latin-Etruscan bilingual inscription (CIE 1729/ET2 Cl 1.1221).Despite Danielsson's frustration at himself for not washing away the red paint on the tile, he did not adopt this as part of his regular preparations, as there are squeezes with paint transfer made by him as late as July 1890. 52More examples of pigment transfer are bound to be found in the collection as itemization continues, and it is very likely that other squeeze collections focusing on Classical epigraphy contain squeezes with pigments.This makes squeezes a valuable resource in the study of ancient pigment and its use in epigraphy.

Figure 1 :
Figure 1: Funerary tile kept in the Museo Nazionale Archeologico in Chiusi.Photograph by Fabrizio Vallelonga.Used with permission from Ministero della Cultura -Direzione regionale Musei della Toscana -Firenze.

Figure 2 :
Figure 2: Line-drawing of the squeezes, showing the area covered by each squeeze.The line-drawing is based on the squeezes, meaning the inscription is mirrored.The broken line indicates the extent of the epigraphic field on the original tile.By the author.

Figure 8 :
Figure 8: Line-drawings of the second word on line 2, representing the painted caunei from squeeze 1393 and the carved cauśine from squeeze 1396.By the author.