Monument Future

References

Carvalho J., Lisboa V. 2018. Ornamental stone potential areas for land use planning: a case study in a limestone massif from Portugal. Environ Earth Sci 77:206. https://doi.org/10.1007/s12665-018-7382-x.

Dunham R. J. 1962. Classification of carbonate rocks according to depositional textures. In Ham, W. E. (ed.), Classification of Carbonate Rocks: American Association of Petroleum Geologists, Memoir 1, 108–121.

Flügel E. 2004. Microfacies of Carbonate Rocks. Analysis, Interpretation and Application: Germany, Springer, 976 pp.

Folk R. L. 1962. Spectral subdivision of limestones types. In Ham, W. E. (ed.), Classification of Carbonate Rocks: American Association of Petroleum Geologists, Memoir 1, 62–84.

Koch A., Siegesmund S. 2002. Bowing of marble panels: on-site damage analysis from the Oeconomicum building at Goettingen (Germany). Geological Society, London, Special Publications 205(1):299–314

Koch A., Siegesmund S. 2004. The combined effect of moisture and temperature on the anomalous expansion behaviour of marble. Environ Geol 46:350–363

150Menningen J., Siegesmund S., Lopes L., Martins, R., Sousa, L. 2018. The Estremoz marbles: an updated summary on the geological, mineralogical and rock physical characteristics. Environ Earth Sci 77: 191. https://doi.org/10.1007/s12665-018-7328-3

Shushakova V., Fuller E. R., Heidelbach F., Mainprice D., Siegesmund S. 2013. Marble decay induced by thermal strains: simulations and experiments. Environ Earth Sci 69(4):1281–1297

Siegesmund S. 2008. Neue Steine und alte Sorgen—Fassadenplatten aus Naturstein: Sicherheitsrisiken und Sanierungsstrategien. In: H. Venzmer (Hrsg.) 19. Hanseatische Sanierungstage Bauphysik und Bausanierung Heringsdorf 2008, S. 17–27, Beuth Verlag, Berlin.

Siegesmund S., Dürrast H. 2014. Physical and mechanical properties of the rocks. In: Siegesmund S, Snethlage R (eds) Stone in architecture. Properties, durability, 5th ed. Springer.

Siegesmund S., Rüdrich J., Koch A. 2008. Marble bowing: comparative studies of three different public building façades. Environ Geol 56(3–4):473–494.

Siegesmund S., Sousa L., Knell C. 2018. Thermal expansion of granitoids Environ Earth Sci 77: 41. https://doi.org/10.1007/s12665-017-7119-2.

Siegesmund S., Ullemeyer K., Weiss T., Tschegg E. K. 2000. Physical weathering of marbles caused by anisotropic thermal expansion. Int J Earth Sci 89(1):170–182

Silva Z. 2017. The Portuguese Lioz, a Monumental Limestone. Geophysical Research Abstracts. Vol. 19, EGU2017-9019.

151

THE FOUR SCULPTED COLUMNS OF THE ST. MARK BASILICA’ CIBORIUM, VENICE: MARBLES, POLYCHROMY, PAST TREATMENTS

Lorenzo Lazzarini, Elena Tesser*

IN: SIEGESMUND, S. & MIDDENDORF, B. (EDS.): MONUMENT FUTURE: DECAY AND CONSERVATION OF STONE.

– PROCEEDINGS OF THE 14TH INTERNATIONAL CONGRESS ON THE DETERIORATION AND CONSERVATION OF STONE –

VOLUME I AND VOLUME II. MITTELDEUTSCHER VERLAG 2020.

LAMA – Laboratory for Analysing Materials of Ancient origin Iuav University of Venice, San Polo 2468/B, 30125 Venice, Italy, lama@iuav.it – * etesser@iuav.it

Abstract

The famous four marble columns of the ciborium are among the most important artifacts preserved in St. Mark’s Basilica, Venice. Carved with scenes of Christ and Mary’s lives by two masters in Constantinople in the first half of the 6th century BC, they were subsequently pillaged from an unknown church by the Venetians soon after the 1204 Crusader’s sack of the city. Reworked and gilded in Venice, they were re-installed after the 1222 earthquake in the ciborium covering the sarcophagus of the evangelist and the main altar. Thought to be made of oriental alabaster, they have been subject to many treatments throughout the centuries and the subject of extensive archaeometric study with identification of their materials, previous gildings and protective treatments. Micro-samples taken from each column were analysed by minero-petrographic (XRD, OM, SEM+EDS on thin and polished cross sections) and geochemical (SIRA, microFTIR and Raman spectroscopy) techniques. The results indicate that the columns consist of Dokymaean marble originating from three different quarries.The original gold leaf gilding was applied to a yellowish preparatory layer of lead-white mixed with other pigments subsequently covered by two later gildings, the most recent of which was laid on a red minium-ground. A protective/consolidation film of dammar, now discoloured into a brown coating, was identified together with other organic media such as siccative oils and proteinaceous substances and were probably applied in the XIX century and later.

Introduction

Saint Mark’s Basilica is the most important monument of Venice. Built as the private Ducal Chapel, it became the centre of religious life in 1806 when it became the town cathedral. Since its construction, the main altar of the Basilica holds the relics of the body of the evangelist, transported in 828 AD to Venice, by two sailors, Bono of Malamocco and Rustico of Torcello. The receipt of these relics prompted the initial building of the first Basilica, which was inaugurated two years later under the doge Giovanni Parteciaco. At the end of the 11th century, this was lavishly decorated with polychrome marbles but increasingly after the Fourth Crusade (1204), and has been restored many times up to the present day, with the continuous addition of altars and sculptures. At some point in the first half of the XIII century, likely after a strong earthquake in 1222, a ciborium was erected above the main altar (Wolters 2014: 145-6) with marble spoliated after the Latin’s 1204 sack of Constantinople. The ciborium has a square plan and an 152elevation of rectangular shape forming a four-sided baldachin that stands isolated in the middle of the presbiterium (Fig. 1). The vault is cross-shaped with external walls faced with marmor thessalicum (verde antico) (Lazzarini 1997: 324; Lazzarini 2007: 223–244); the covering roof holds six small stone statues, some of which are re-used from other monuments. The whole structure is supported by four marble Corinthian-style capitals made in Venice in the XIII century imitating Roman originals of the II century A. D. These are also gilded similar to the corresponding magnificent columns (Fig. 2) that are the subject of this study. The circumferential carving covers the length of the columns with the story of Christ (in columns A, B, and D, infra) and Mary (only in column C) (Fig. 3): according to Weigel (2015) they are Byzantine masterpieces of two different sculptors of the VI century, originally installed in an important unknown basilica of Constantinople: the master chiselled the front columns B and D, and his pupils sculpted columns A and B on the back.

Figure 1: The ciborium of the Saint Mark’s Basilica.

Historically, these columns have been wrongly identified as of oriental alabaster (for ex. Lorenzetti 1994: 188), often confirmed in publications by several scholars. This misidentification may be justified by the gilding still partially present on their surface, which has been subject to repeated protective treatments with natural organic substances leading to a full covering of their surface. On accurate macroscopic observation by the current authors, marble appeared to be the more likely material of these columns, tentatively identified as Proconnesian (Lazzarini 1997). Proconnesian marble was quarried in ancient Proconnesos, the present day Island of Marmara (Turkey) and is the marble of Venetian monuments (Lazzarini 2015). On a monographic collective study of these columns, it was possible to sample three of them, and submit the marble to archaeometric studies (minero-petrographic and isotopic, see below) that allowed identification of two of the columns to be of Dokymaean and/or Pentelic (Lazzarini 2015b: 59) marble. To determine the origin of the fourth column as well as to investigate the 153nature and stratification of the gilding and past treatments a new sampling was made recently.

Figure 2: Details of the four carved marble column: a) column A at the back and B at the front; b) column C at the back and D at the front.

Sampling and experimental

The sampling was made by removing with a small, very sharp chisel micro fragments of the marble of the four columns (Fig. 3) from now on indicated as follows: column A = back, left (looking from the front of the ciborium); column B = front, left; column C = back, right; column D = front, right.

Also sampled were some of the gildings (including the relative red bole-preparation) and of the patinas. Polished cross-sections of the latter two components were prepared and later examined in polarized reflected light and in ultraviolet light under a LEITZ DM RXP microscope. The same sections were then studied under a SEM coupled with an EDS microanalysis (EVO+BRUKER) for the topographical and chemical analyses. Raman spectroscopy supported the identification of the pigments used for the preparation of gilding layers. The chemical nature of the gilding preparation and of eventual past treatments was determined by FTIR and µFTIR analysis using ThermoScientific instruments (iZ10 and iN10 Infrared Microscope). The µFTIR investigations of allowed the mapping of the distribution and penetration depth of the organic components, whereas FTIR analyses of microflakes on standard KBr pellets allowed better identification of the mixture-components.

To identify the marble provenance, a small portion of each marble sample was finely ground and the powder subjected to X-Ray diffraction (PANalytical Empyrean X-ray diffractometer, Cu-Kα radiation at 40 kV and 20 mA) to evaluate the possible presence and relative abundance of dolomite. O and C stable isotopes ratio analyses were performed on the same powders through a Gasbench II preparation line connected on-line to a Thermo Finnigan Five Plus mass spectrometer in a continuous flow mode. All δ13C and δ18O values were measured against a PDB standard: the results were then plotted in reference isotopic diagrams obtained from the most updated database (Antonelli and Lazzarini, 2015). The remaining larger fragment of each marble sample was used for the preparation of a thin section studied under a LEITZ DM RXP polarising microscope. The main petrographic features of marbles (fabric, boundary shapes of the carbonate crystals, maximum grain size (MGS) of the largest crystal of calcite expressed in mm, presence and relative quantity of accessory minerals) were compared with both the most recent published data and with reference samples taken from ancient quarries (the extensive thin section collection present in the LAMA, University Iuav of Venice).

Figure 3: Detail of columns A (left) and D (right).

Results and their discussion

The results of the minero-petrographic and isotopic analyses of the columns’ marbles are summarised in Table 1. There was considerable homogeneity of the results for the fabric parameters, hetero/homeoblastic mosaic type, embayed crystal boundaries and maximum grain size around 0.6–0.8 mm (Fig. 4), and mineral composition: rather pure calcite, with small amounts of accessories, mainly carbonaceous matter/graphite and traces of quartz and k-mica. The results of the isotopic analyses indicate that the columns clearly forms two groups, one with δ values around –0.8, and one with δ +1.6/–1.7, while for 3 columns gave similar δ values (around –9.4/–9.7), one totally different (–5.4). From the plot of such results in the reference diagram (Fig. 5) it may be deduced that columns A and D were cut from the same quarry locus, possibly the same block of marble, while columns 154B and C came from different quarries. Combining the results of the minero-petrographic and isotopic analyses, with reference to the most updated databases (Antonelli, Lazzarini 2015) and with direct comparison of thin sections of ancient quarry samples, it may be concluded that the marble of three columns is most probably from ancient Docimium (corresponding to the present-day village of Iscehisar, province of Afyon, Turkey), namely from two different loci, one for columns A and D, and one for column C. The isotopic ratio of the fourth column is outside all the known reference fields of the most important fine-grained marbles used in antiquity and therefore, cannot be assigned isotopically. Considering that its petrographic features are very close to those of the other three shafts, however, one may hypothesize that column B is also of the same marble, but of an unknown ancient quarry.

Figure 4: Photomicrograph of the thin sections of the marbles of columns A, C, B, D (left to right, top to button), N+, long side = 3.8 mm: all showing similar mosaic fabric formed by calcite crystals with curved-to-embayed boundaries.

In a thin cross section of column D, a superficial brown film of Ca-oxalates (Fig. 6a) was observed by optical microscopy, formed from the mineralization of an unknown organic treatment material. The presence of abundant phosphorus (P) (Fig. 6b), covered by a deposit of airborne quartz particles and gypsum (Fig. 6c and d) suggests the organic matter to be casein, corresponding to an ancient conservation treatment, with a weak sulphation process affecting the marble of the column.

Figure 5: Plot on the reference isotopic diagram of the most important fine-grained marbles (Antonelli, Lazzarini 2015) of the resulting ratios of the 4 columns.

Table 1: Summary of the minero-petrographic and isotopic analyses (HE, heteroblastic; HO, homeoblastic; M. G. S., maximum grain size; +++, very abundant; ++, abundant; +, present; ±, traces; –, absent).

155

Figure 6: Photomicrograph of the thin section of column D: a) showing the old brown treatment layer on top of the marble substratum, covered by a thick gypsum layer, N+, long side = 0.96 mm; b) SEM-EDS mapping of the phosphorous distribution; c) same for Si; d) same for S.

The microscopic study (OM and SEM) of several unmounted samples and of four cross sections has allowed us to conclude that the columns were originally most probably painted. A white-yellowish layer made of lead-white was mixed with a small amount of calcium carbonate, orpiment, minium, vermilion and yellow ochre (in decreasing order of abundance) identified by Raman spectroscopy and SEM-EDX (Fig. 7, 8). It is approximately 0.15 mm thick, and may be considered as a sort of imprimitura (priming) applied directly over the marble substratum, very much likely the one often found in variously dated panel paintings (Gettens et al. 1967: 123). Over this layer, was a much thinner one (around 0.06 mm) composed of white-lead mixed with a proteinaceous medium, likely an animal glue, as in preparation for gold leaf. This original gilding is covered by dirt, carbonaceous matter likely from candle-burning, and by two later gildings; the first again applied on a thin lead-white layer mixed with a proteinaceous binder. The second is not preserved, however from the presence of a thick red layer of minium (0.2 mm) it can be assumed this was another preparation for gold leaf, as commonly used in the Venetian Renaissance. A natural oil-resin, most probably dammar impregnating all layers (Fig. 9), was applied in order to fix and consolidate the gildings: its discolouration into a brownish matter is probably responsible for the overall brown aspect currently assumed by the columns.

Figure 7: Photomicrographs of the polished cross section of sample A3 (column A), a) in reflected light mag; b) stratigraphic scheme; c) SEM in backscattered electrons; d) same as c), but detail of the strata. The strata represented in b) and d) are: 1) painted preparation layer, 2) preparation for the gold leaf, 3) the original gilding, 4) dirt layer, 5) preparation for the gold leaf, 6) the second gilding, 7) dirt laye, 8) red layer of minium.

Figure 8: Detail of sample A3, a) SEM in backscattered electrons.; b) EDS mapping of the distribution of Pb; c) same of Ca; d) same of Au.

Conclusion and future research

Considering the overall results of the petrographic and isotopic analysis of the four marbles, it is 156possible to conclude that they were carved out of the white Dokymaean marble. Two of them (A and D) originated from the same quarrying locus, while the marble of column C was extracted from a different quarry. The provenance of column B could not be positively identified but likely comes from an unknown ancient Dokymaean quarry. The ancient quarrying area of Docimium was one of the largest of Roman and Byzantine times, active from Late Hellenistic to at least the X c. (Pensabene 2011:131) adding weight to our hypothesis. This quarry has been massively re-exploited in modern times and is now the largest white marble supplying area in Turkey. It may well be that one or more ancient quarries have been completely destroyed by modern exploitation therefore escaping study and sampling by archaeologists and archaeometrists. The high quality of the Dokymaean marble, namely its uniform fine grain size, allowed the two sculptors to obtain the beautiful details of the small figures narrating the life of Christ and Mary. The different quarries of origin of the marbles support Weigel’s hypothesis of Byzantine re-use of the four columns, pillaged from an ancient Roman monument. Dokymaean marble was often used for prestigious statuary, portraits (Attanasio et al.2019:178), sarcophagi and architectural elements, and widely exported from its quarries especially in all the Microasiatic provinces (Monna and Pensabene 1977:29–71).

As for the gildings present on the columns, the first and most ancient one was made of two layers: a thicker yellow stratum obtained by mixing a small amount of orpiment, minium, cinnabar and yellow ochre to lead-white (Gettens et al. 1966), and a much thinner one of the same white pigment mixed with a proteinaceous medium. This type of preparation is quite unusual, similar only in colour to those found on the contemporaneous sculpted external arches over the main portal of the Basilica (Lazzarini 1979: 60) (Lazzarini 1995: 232–234). In both structures, the second gilding was obtained in the same way by applying a gold leaf on a minium ground, indicating similar re-gilding of internal and external sculptures of St. Mark’s. Analysis of the brown organic medium impregnating all the superficial layers of the columns revealed a composition of discoloured dammar, a resin much used for consolidating and protecting ancient gilding. This resin began to be used in the West as a varnish for paintings in the first half of the XIX century (Mills, White, 1994: 107), therefore we suggest this conservative treatment should be dated after that period.

Figure 9: FTIR spectrum of the brown patina of sample A1, showing all the peaks of dammar.

The present aspect of the four columns of the St. Mark’s ciborium is far from being satisfactory. The brown patina which almost uniformly covers both areas where gilding remains as well as the exposed marble surfaces, reduces both the impact and legibility of the figures narrating the lives of Christ and Mary and the beauty of the sculpture which leads us to suggest a thorough but delicate cleaning.