234

/Abstract / Just behind the colonnade of Hagia So-phia’s south gallery on the wall of the southeast pier and now concealed beneath a thin layer of plas-ter, a panel of mosaic shows a monumental cross with flaring ends embellished with pearls. Schol-ars have largely overlooked this panel. Its back-ground – almost certainly of gold—is made of tesserae laid in parallel horizontal rows, each one slightly separate from the adjacent rows. This kind of pattern is a characteristic of Middle Byzantine Hagia Sophia: it is a prominent feature in parts of the poorly pre-served geometric borders of Emperor Alexander’s panel in the north gallery. Thus, the panel can be considered to be an example of this distinctive ninth- and tenth-century mosaic technique.

/ Keywords / Hagia Sophia, Constantinople, Middle Byzantine Period, Wall Mosaic, Foliated Cross

Alessandro TaddeiSapienza Università di Romaalessandrotaddei@hotmail.it

235

A recently published article by Natalia Teteri-atnikov sheds new light on some critical aspects regarding the famous wall mosaic of Emperor Alexander in the north gallery of Hagia Sophia / Fig. 1/. A new reading of the historical background to the reigns of Basil i’s sons, Leo vi (886 – 912) and Alexander (912 – 913), leads her to advance, among other conclusions, a new chronological hypothesis for the panel1. A significant point to bear in mind here is that Alexander’s panel is placed on the east face of the nw main pier of the nave, at the level of the gallery. The upper portion of each of the four main piers of Hagia Sophia shows a complex superimpo-sition of architectural and decorative phases, mainly due to the fact that their original architectural layout had been modified on many occasions since the first rebuilding of the church in 558 – 562. Most of these alterations were made during the restorations and/or reconstructions subsequent to the major earthquakes of 740, 869 and 989 / Fig. 2 / 2.

The Emperor Alexander mosaic – which was discovered under the nineteenth-century Ottoman plaster in 1958 –1960 by Robert Van Nice – was ac-curately reported by Paul Underwood and Ernest Hawkins, and represents the most substantial among the few surviving fragments of the ninth- and tenth-century mosaic decoration of the upper gal-leries of the church 3. Except for such fragments, we

know most of these middle Byzantine mosaics only through the drawings and sketches produced mainly from the seventeenth century onwards. Thanks to – among various other illustrations – Cornelius Loos’ engravings (1710 –1711) and Wilhelm Salzenberg’s watercolours (1854), we are able at least to form a fairly clear idea of part of the lost wall mosaic revet-ments of the galleries 4.

A Neglected Mosaic in the South Gallery of Hagia Sophia, Constantinople*Alessandro Taddei

* I am particularly grateful to Professor Andrea Paribeni (Università degli Studi di Urbino “Carlo Bo”) who discussed patiently with me during the writing of the text, and helped me with his thorough knowledge of Hagia Sophia. It goes without saying that whatever faults or inac-curacies are mine and mine alone.

1 Natalia B. Teteriatnikov, “Why is he hiding? The mosaic of Emperor Alexander in Hagia Sophia, Constantinople”, Arte Medievale, 4th series/ii (2012), pp. 61– 76.

2 Recent scholarly approaches to the building’s history and the related architectural problems have often come to different and, indeed, contrasting conclusions. The investigations conducted by Rowland Mainstone and Robert Van Nice in the upper structures of the building were published in: Rowland Mainstone, “The Reconstruction of the Tympana of St. Sophia at Istanbul”, DOP, xxiii–xxiv (1969 –1970), pp. 355 – 368, sp. p. 367; Rowland Mainstone, Hagia Sophia. Architec-ture, Structure and Liturgy of Justinian’s great church, New York 1988, pp. 97– 98, figs 117–124; Urs Peschlow, “Die Hagia Sophia und das Erdbeben des Jahres 740”, in Die Hagia Sophia in Istanbul. Akten des Berner Kolloquiums vom 21. Oktober 1994, Volker Hoffmann ed., Bern 1998, pp. 89 –107. Both Mainstone’s and Peschlow’s conclusion were largely questioned by Eugenio Russo, Le decorazioni di Isidoro il Giovane per S. Sofia di Costantinopoli, Rome 2011.

3 Paul A. Underwood, Ernest J. W. Hawkins, “The Mosaics of Hagia Sophia at Istanbul. The Portrait of the Emperor Alexander”, DOP, xv (1961), pp. 187– 217.

4 Alfred Westholm, Cornelius Loos, teckningar från en expedition till Främre Orienten 1710–1711, Stockholm 1985; Wilhelm Salzenberg, Altchristliche Baudenkmale von Constantinopel vom v. bis xii. Jahrhundert, Berlin 1854, pp. 95 –100, pl. 23 – 30.

Loos’ engravings provide us with an accurate, albeit black-and-white, representation of the whole ensemble of the galleries’ mosaics around 1700, before they were largely destroyed. The Swedish engineer saw almost all the vault surface of the gal-leries of Hagia Sophia still covered with glass cubes. These drawings place some crucial information at our disposal: first of all, the wall mosaics executed in the galleries following the middle Byzantine restorations were both decorative (non-figural) and figurative. Second : they were almost all still visible in the eighteenth century.

We owe our almost complete knowledge of the entire decoration of the area to the remarkable quantity of drawings and short descriptions made by the brothers Gaspare and Giuseppe Fossati in 1847–1849. During the extensive restorations the Fossatis conducted in the then Ayasofya mosque, many of the Byzantine figurative wall mosaics in the galleries were discovered or reported, following the removal of the late Ottoman plaster. Soon after they were, understandably, again covered over. This was the case, for example, of Emperor Alexander’s panel. Many others – most, in fact – perished due to their poor state of preservation, which made any attempt to keep them in place under the new coat of plaster impossible 5.

The first middle Byzantine restoration work in Hagia Sophia had probably begun under the reign of Basil i (867– 876), after the 869 earthquake had damaged the upper structure of Justinian’s building. Since discussion of the architectural phases of the galleries of Hagia Sophia is beyond the scope of the present paper, it is useful now to summarize only the most relevant points. First, the 869 rebuilding works appear not to have been very extensive: only a few months after the earthquake, the south gallery of Hagia Sophia was ready to house the sessions of the eighth Ecumenical Council (October 869 to February 870). Second: in the Peri Ktismaton catalogue of the Vita Basilii – very likely authored or drawn up by order of his grandson Constantine vii 6 – no mention is made of radical interventions performed on the tympana and the gallery structures. What rather shocked the con-temporaries was the severe crack that the earth-quake caused on the western main arch of the naos, for we are told that Basil i promptly reinforced

1 / Panel with Emperor Alexander, mosaic, Hagia Sophia, Constantinople

2 / Layout of the northwest pier at the level of the gallery, Hagia Sophia, Constantinople

237

it and decorated it with new mosaics depicting the Virgin Mary holding the Son and the apostles Peter and Paul 7.

As a whole, the new wall-mosaic decoration of the upper sections of the church would prove to be a much longer-lasting enterprise than the structural interventions. A complex of figurative wall mosaics then made its first appearance inside Hagia Sophia, whereas Justinian’s original wall mosaic decoration was – for many reasons – definitely non-figurative 8. This means that the late ninth-century Byzantine craftsmen took on a new challenge, arranging a figurative cycle over ample surfaces which had hitherto been covered with extensive revetments of gold tesserae or simple geometrical patterns, gar-lands or frames9. How painstakingly this difficult task was accomplished appears clearly from the new figurative decoration of the apse and the bema 10.

Apart from the apse semi-dome and the barrel vault of the bema, the artists must have started their work from the two great tympana of the nave, subse-quently proceeding to the mosaic decoration of the vaults, soffits, arches and pendentives in the bays of the north, south and west galleries / Fig. 3 /. As for the tympana, only the portraits of the Church Fathers in the recessed niches at the base of the two walls have

5 Cyril Mango, Materials for the Study of the Mosaics of St. Sophia at Istan-bul, Washington d.c. 1962, pp. 7– 21; Natalia B. Teteriatnikov, Mosaics of Hagia Sophia, Istanbul: The Fossati Restoration and the Work of the Byzantine Institute, Washington d.c. 1998; Sabine Schlüter, “L’opera di restauro di Gaspare Fossati per Santa Sofia”, in Santa Sofia ad Istanbul. Sei secoli di immagini e il lavoro di restauro di Gaspare Fossati (1847– 49), exhibition catalogue (Mantova, Casa del Mantegna 14 novembre – 31 dicembre 1999), Volker Hoffmann ed., Bern 1999, pp. 139 –147; Sabine Schlüter, “Gaspare Fossatis Restaurierung der Hagia Sophia in Istanbul, 1847– 49”, in Die “Denkmalpflege” vor der Denkmalpflege, Akten des Berner Kongresses 30. Juni – 3. Juli 1999, Volker Hoffmann ed., Bern 2005, pp. 361– 376.

6 See Mario Gallina, “Arte della propaganda e propaganda dell’arte a Bisanzio: la sezione Περὶ κτισμάτων della Vita Basilii”, in Studi in onore di Salvatore Tramontana, Enrico Cuozzo ed., Castel di Serra 2003, pp. 227– 244.

7 Chronographiae quae Theophanis Continuati nomine fertur liber quo vita Basilii imperatoris amplectitur, Ihor Ševčenko ed., Berlin/Boston 2011, p. 265 – 266 (§79); See Mango, Materials (n. 5), pp. 76 – 77; See Idem,

“Byzantine Writers on the Fabric of Hagia Sophia”, in Hagia Sophia from the Age of Justinian to the Present, Robert Mark, Ahmet Ş. Çakmak eds, Cambridge 1992, pp. 41 – 56, p. 53. 

8 I tried to find a possible religious, doctrinal and cultural background for this radical choice in Justinian’s original decoration of the church: Alessandro Taddei, La decorazione musiva aniconica delle chiese della Santa Sofia e della Santa Irene a Costantinopoli (unpublished PhD Thesis, Sapienza Università di Roma, 2012), pp. 4 –12; See also Alessandro Taddei, “L’impresa musiva giustinianea nella Santa Sofia di Costan-tinopoli. Una lettura orientale”, in Un medioevo in lungo e in largo. Da Bisanzio all’Occidente (vi–xvi secolo). Studi per Valentino Pace, Vittoria Camelliti, Alessia Trivellone eds, Pisa 2014, pp. 10 –19.

9 According to Loos’ drawings, the mosaic decoration of the west gallery probably remained non-figurative even after the Middle

Byzantine restorations, the only remains of which appear to be those in the soffits of the triple arch looking toward the nave. See Mango, Materials (n. 5), pp. 40 – 43, figs 40 – 41, 46. For the mosaics in the triple arcade, see also Alessandro Taddei, “La decorazione musiva aniconica della Santa Sofia di Costantinopoli da Giustiniano all’età mediobizantina”, in Vie Per Bisanzio. vii Congresso Nazionale dell’Associazione Italiana di Studi Bizantini, Venezia, 25–28 novem-bre 2009, Antonio Rigo, Andrea Babuin, Michele Trizio eds, Bari 2013, vol. i, pp. 373 – 382, sp. p. 377, fig. 3a – c.

10 Cyril Mango, Ernest J.W. Hawkins, “The Apse Mosaics of St. Sophia at Istanbul. Report on Work Carried Out in 1964”, DOP, xix (1965), pp. 115 –151; Recently: Natalia Teteriatnikov, “Hagia Sophia, Constantinople. Religious Images and Their Functional Context after Iconoclasm”, Zograf, xxx (2004 – 2005), pp. 9 –19; Alessandra Guiglia Guidobaldi, La decorazione musiva della prima età macedone. Questioni aperte, in Bisanzio nell’età dei Macedoni. Forme della produzione letteraria e artistica. viii Giornata di Studi Bizantini (Milano, 15 –16 marzo 2005), Fabrizio Conca, Gianfranco Fiaccadori eds, Milan 2007, pp. 119 –149.

238

survived, scattered remnants of the decoration that once covered their entire surface, probably complet-ed in the decade 880 – 890 11. 

A cycle consisting of four iconographic themes was then designed to adorn the four hemispherical vaults of the two central bays in the south and north galleries. Thanks to the documentation available (Loos / Salzenberg / Fossati), we can be quite sure that the eastern vault of the north gallery was decorated with the scene of the Baptism of Christ in the river Jordan. According to Loos’ view of the north gal-lery, at the centre of the vault there was a medallion encircling a cross 12. On the other hand, nothing can be said about the few mosaic fragment reported by the Fossati in the western vault of the same gallery (adjacent to Alexander’s panel): Loos’ engraving is of no help; the presence of a mosaic depicting the Trans-figuration on Mount Tabor has nonetheless been suggested – a hypothesis to be discussed shortly 13.

In the south gallery reconstruction of the lost cycle can be made with more certainty: in the east vault a bust of Christ Pantokrator dominated. Having sur-vived until the middle of the nineteenth century, the Pantokrator mosaic is well documented not only by Loos but also through several of the sketches made by the Fossatis. Christ stood out within a medallion

supported by figures of cherubs and seraphs, these latter located at the level of the vault pendentives. A small fragment of this mosaic is preserved today, having reappeared in 1960 after the removal of the old plaster in the area of the northeast pendentive of the vault / Fig. 4 /. It shows what Cyril Mango called a gold “parasol ” pattern, enclosed in a dark-blue clypeus standing out against the gold background. Above and below the clypeus appear pairs of sym-metrical green acanthus and vine leaves 14.

A representation of Pentecost found its place in the west vault. Though almost complete at the time of Loos (1710), it was preserved only in fragments when the Fossatis recorded it 15. For this reason, there are a number of substantial discrepancies among the different sketches and reproductions now available. The general layout of the composition consisted of a central medallion with a throne. The medallion ema-nated rays, which reached the figures of the twelve apostles seated on a continuous bench depicted at the base of the vault. Depicted on the pendentives, groups of figures representing the “nations” ( glossai or phylai ) attended the scene 16.

The mosaics in both the vaults of the south gallery “were intended, contrary to normal practice, to be viewed from the east”17. This should be accounted

3 / Plan at the level of the galleries, Hagia Sophia, Constantinople

4 / Northeast pendentive of the so--called Pantokrator Vault, Hagia Sophia, Constantinople

239

for, as Cyril Mango pointed out, with the fact that – when certain particular ceremonies were held – the imperial cortege used to access the south gallery from the southeast ramp, coming directly from the palace. This circumstance could imply that the cen-tral bay of the south gallery is to be considered as the main focus of the whole gallery complex of the middle Byzantine Hagia Sophia18. This kind of ar-rangement is to be taken into account as a premise for the following remarks.

A mosaic beneath the plaster (south gallery, southeast pier)

As we have seen, Emperor Alexander’s panel occupies the eastern face of the northwest pier. This wall constitutes the short side of a narrow “cor ridor ” / Fig. 6 / or elongated bay separating the colonnade of the gallery and the triple arcade serving for reinforcement behind it. The south gallery shows ex-actly the same architectural layout / Figs 3 – 4 /. Barrel vaults cover these narrow corridors in both the north and the south gallery. The narrow plan and high ceilings of these corridors / bays combine to admit only dim lighting: an “unfitting place”, as Natalia Teteriatnikov pointed out 19. Nonetheless, some light

11 As is well-known, the generally accepted dating is circa 880 after the presence of the portrait of the patriarch of Constantinople Saint Ignatius the Younger, who actually died on 23rd October 877. The epi-graph accompanying the portrait – iγnatioc o neoc – as C. Mango and E. Hawkins noted, share similar graphic features with two monumental inscriptions that once ran in two lines above and below the upper rows of windows of both the tympana. Some traces of these latter still sur-vived when the Fossatis carefully reproduced them in their drawings. The whole text of the inscription was later traced by Silvio Giuseppe Mercati in a manuscript he published in 1922 : Silvio Giuseppe Mer-cati, “Sulle iscrizioni di Santa Sofia”, Bessarione. Rivista di studi orientali, xxvi (1922), pp. 199 – 222. See Mango, Materials (n. 5), pp. 63 – 66, fig. 87; Teteriatnikov, “Religious Images” (n. 10), p. 14. The text consists of a four-stanza epigram in which an anonymous commissioner celebrates the repairs made on the building as a measure against decay (and not the earthquakes!). The phraseology is quite comparable with that used in the Vita Basilii: Cyril Mango, Ernest J. W. Hawkins, “The Mosaics of St. Sophia at Istanbul: The Church Fathers in the North Tympanum”, DOP, xxvi (1972), pp. 3 – 41, sp. p. 40. Thus far, the identification of the anonymous commissioner as Basil i seems quite acceptable.

12 The mosaic was described by Mango on the base of Loos’ drawing: “The Saviour himself is shown in the centre, submerged in the waters of the Jordan which rise to from a kind of hillock. The Baptist stands to the left, his hand over Christ’s head. Two angels approach from the right with outstretched hands; a third angel as well as another figure stand on the left (…)”: Mango, Materials (n. 5), p. 47.

13 Ibidem, p. 47: “All one can distinguish is the lower extremity of one figure”; See also Catherine Jolivet-Lévy, “Présence et figures du souverain à Sainte-Sophie de Constantinople et à l’église de la Sainte-Croix d’Aghtamar”, in Byzantine Court Culture from 829 to 1204, Henri Maguire ed., Washington d.c. 1997, pp. 231– 46, sp. p. 235.

14 Mango, Materials (n. 5), pp. 29 – 35, figs 22– 27. See also the Fossa-tis’ sketches published in the catalogue Santa Sofia ad Istanbul (n. 5), pp. 173 –175, 265 (cat. no. 15). See also Taddei, “Decorazione musiva aniconica” (n. 9), p. 376 – 377, fig. 2b – c.

15 Mango, Materials (n. 5), p. 35.16 Ibidem, pp. 35 – 38.17 Ibidem, p. 37.18 Ibidem, p. 38.19 Teteriatnikov, “Why is he hiding?” (n. 1), p. 70.

240

comes indirectly from the windows of the galleries, thus creating a sort of permanent twilight. Those wishing to attend the main liturgy held in the nave from the gallery colonnades had to position them-selves right inside these “corridors ”.

Due to the feeble light the barrel vaults of the corridors are decorated with a simple mosaic “carpet” of geometrical elements / Fig. 4 /. Very roughly exe-cuted, the mosaic employs cubes of white-greyish Proconnesian marble and a great amount of blue-greyish rock tesserae 20. The rough mosaic sur-face, above all, attests to work accomplished hastily and with little care. It shows nonetheless the same technical features as the geometrical mosaic borders framing Emperor Alexander’s panel, those classified by Underwood and Hawkins under phase 3 of the mosaic palimpsest of the northwest pier / Fig. 2 / 21.

We have no precise idea as to the kind of mosaics that were to be found in the wall surfaces sym-metrical to the surface occupied by Alexander’s panel, i. e. the inner faces of the northeast, southeast and southwest piers. There, the mosaic revetment was covered by the Fossatis with a uniform coat of plaster, adorned with a geometrical pattern clearly inspired by the original motif of the barrel vaults of the corridors. No substantial information about the mosaic revetment is provided – as far as we know

– by Loos, Salzenberg or even the Fossatis’ dossier of sketches and drawings. As a matter of fact, most of the nineteenth-century plaster was gradually re-moved from the gallery walls and vaults from the 1930 s onwards, as the work by the Byzantine Insti-tute of America went ahead on the basis of the Fossati drawings 22. Nevertheless, scant attention was paid to the east end of the corridor of the south gallery, i. e. the western face of the southeast pier.

The Southeast pier – immediately to the left of the surviving mosaic fragment of the Pantokrator vault / Fig. 6 / – preserves some surprising elements: the west face of the pier, though still completely whitewashed, shows evident traces of a mosaic lying beneath the thin coat of plaster. A careful look at the plaster surface when the sunrays provide oblique lighting could prove revelatory. Here, very likely, the Fossatis hastily whitewashed the wall in order to con-ceal a mosaic. Thereafter, they painted the coat with the usual pattern, imitating the geometrical elements of the barrel vault above (Alexander’s portrait had

undergone the same treatment)23. As moisture or salt efflorescence gradually deteriorated the painted coat, the nineteenth-century colours practically dis-appeared while the underlying mosaic texture reap-peared, though its colours are faded under a film of filth, decayed paint, and soot.

One can now barely discern the subject of the mosaic: a monumental cross with flaring ends embel-lished with pearls / Fig. 7 /. The outline of the cross is seemingly obtained with dark-red (?) cubes on a gold background. It is difficult to ascertain whether the body of the cross was jewelled or not. From the base of the cross sprout a couple of thin symmetrical acanthus leafs, slightly curving upward and then spreading horizontally over the entire width of the panel surface. Robert Van Nice actual ly reported the existence of this cross in the plate no. 29 of his architectural survey of Hagia Sophia 24.

Technical and stylistic remarks

The reconstruction of the wall mosaic decoration of the southeast pier raises a number of issues:

The (almost certainly) gold background of the panel / Fig. 7 / is made of tesserae characteristically laid out in parallel horizontal rows, each row being slightly distanced one from the next. This kind of layout is well attested in Middle Byzantine Hagia Sophia. It is a prominent feature in the mosaic panel of the lunette mosaic above the Imperial Door and is to be found in some areas of the poorly preserved geometrical borders of the Alexander panel as well. Thus, it can be maintained as a distinctive sign of the ninth- and tenth-century mosaic technique.

Just above the flaring end of the upper arm of the cross the mosaic revetment appears to break off abruptly / Fig. 8a – b/. Nowadays the area from here up to the barrel vault is covered by the usual coat of painted plaster laid by the Fossatis. This upper edge of the mosaic-covered area, perfectly horizontal, does not even show any trace of a decorative border. So far, it is impossible to ascertain which kind of wall decoration was originally intended to cover the

20 Underwood / Hawkins, Emperor Alexander (n. 3), p. 204.21 Ibidem, pp. 203 – 210.22 Teteriatnikov, Mosaics of Hagia Sophia (n. 5), pp. 31– 68.23 Underwood / Hawkins, Emperor Alexander (n. 3), p. 190.24 Robert L. Van Nice, Saint Sophia in Istanbul: an architectural survey, 2nd

installment, Washington d.c., 1986, pl. 29.

5 / Barrel vault of the “corridor” of the south gallery, Hagia Sophia, Constantinople

6 / West face of the southwest pier of the south gallery, Hagia Sophia, Constantinople

7 / Southeast pier, wall surface with graphic high-lighting of the outlines of the original mosaic panel with the cross, Hagia Sophia, Constantinople

242

surface above the panel with the cross. In the case of the Emperor Alexander mosaic / Fig. 1/ we may observe the same horizontal cutting of the mosaic re-vetment immediately above the emperor’s halo (the crop of the panel slightly damaged the clypeus with the name Alexandros) 25. Underwood and Hawkins reported that only few tesserae were visible above the upper margin of the Alexander panel: these are of large dimensions and of blue-greyish rock, and the authors consequently inferred that very likely there had once been a mosaic with pattern similar to that of the barrel vault above the portrait of Alexander 26.

Incidentally, the wall onto which the imperial portrait was laid appears to be narrowed by the pres-ence of what Rowland Mainstone called a “truncated pilaster” / Fig. 2/, an enigmatic architectural feature, probably evidence of a previous arrangement of the colonnade structure, now acting as a buttress. It is worth noting that the upper margin of the mosaic of Alexander is not linear and roughly coincides with the abrupt termination of the pilaster itself. We conjecture that the scattered tesserae noticed by Underwood and Hawkins were indeed part of a hasty restoration of the mosaic texture with the same decorative pattern and the same materials as those of the vault. The southeast pier, too, has a “truncat-ed pilaster” which creates the same narrowing of

the wall but the area covered by mosaic appears to break off well below the level of the upper end of this pilaster / Fig. 8a – b /. In the other two symmetric walls (on the northeast and southwest pier), every trace of mosaic having disappeared, no conclusive evaluation appears possible.

No medium other than mosaic can be imagined as suitable revetment for these lunette-shaped areas above the panel of Alexander and that of our cross. We would, rather, suggest that an original simple texture made up of golden cubes extended above the two panels. Thus, undecorated as it was, and confined within the most shadowy part of the “cor-ridors”, it rapidly fell victim – probably well before the late Ottoman whitewashing – to deliberate de-struction in order to collect precious metallic cubes. This may explain why, noticeably, the portrait of the emperor and the cross were left undisturbed or suffered limited damage. The area was thereafter covered with a new texture imitating the same dec-orative pattern as that of the vault.

But, how did the original decoration of the upper section of the wall actually appear? The other mon-umental mosaic crosses in middle Byzantine Hagia Sophia deserve attention: they were often employed simply to decorate lunettes or other circumscribed architectural surfaces. Jewelled crosses in gold on a

8 / Southeast pier, Transverse section (looking East), Hagia Sophia, Constantinople

8 a – b /  Southeast pier, Transverse section (looking East), Hagia Sophia, Constantinople

243

blue background replaced, for instance, the busts of saints in the clypei of the tympana under the vault of the room above the southwest ramp (the small sekreton). A case of surviving iconoclastic decoration, these crosses were in fact introduced by patriarch Niketas between 768 and 769 to clear the then for-bidden images from the Patriarchate’s rooms on the occasion of renovations 27. This decorative type, none-theless, continued to be used well after iconoclasm, for it can be made out starting from the eighth-cen-tury mosaics in the tympana of the small sekreton up to the later mosaics of the southwest vestibule, where a monumental example of cross in a globe (still to be cleaned from plaster) appears on the wall of the central bay 28.

A further, large, jewelled cross with flaring ends enclosed in a blue clypeus made its appearance in the ninth- and tenth-century wall mosaic decoration of the south wall of the western gallery. The Fossatis documented the mosaic in some sketches before its destruction, and we are thus informed that it was located in the lunette just above the door opening toward the so-called “Room above the southwest vestibule ”. The clypeus (framed by a garland?) had a diameter of circa 2. 25 m and stood out on a gold background. The horizontal arm of the cross mea-sured 92 cm in length. Diagonal rays of light ema-

25 The mosaic decoration ends more or less at the level of the upper end of what was called by Mainstone the “truncated pilaster”. This is an unfinished architectural element of the joint between the upper colonnades and the main piers of the nave, attributed either to the 532– 537 phase or to the 558–562 rebuilding: see Mainstone, Hagia Sophia (n. 2), p. 191, fig. 221. Russo, Decorazioni di Isidoro il Giovane (n. 2), pp. 21, 71.

26 Underwood / Hawkins, Emperor Alexander (n. 3), p. 190.27 Mango, Materials (n. 5), p. 94; Robin Cormack, Ernest J.W. Hawkins,

“The mosaics of St. Sophia at Istanbul the rooms above the south-west vestibule and ramp”, DOP, xxxi (1977), pp. 175 – 251, sp. p. 205; See Leslie Brubaker, John Haldon, Byzantium in the Iconoclast Era c. 680 – 850. A History, Cambridge 2011, pp. 201– 202, 214 –215; Leslie Brubaker, Inventing Byzantine Iconoclasm, London 2012, p. 35, fig. 4 (with bibliography).

28 This latter is discussed in an article by Philipp Niewöhner and Natalia Teteriatnikov on the architecture and mosaic decoration of the ves-tibule: Philipp Niewöhner, Natalia Teteriatnikov, “Architecture and Ornamental Mosaics in the South Vestibule of St. Sophia at Istanbul: The Secret Door of the Patriarchate and the Imperial Entrance to the Great Church”, DOP, lxviii (2015), pp. 117–156.

29 Mango, Materials (n. 5), pp. 40 – 42, figs 42– 44.

nated from the crossing of the arms, from where the blue background shaded into a brighter area formed by two concentric circles of grey and light green29. Though no leaves sprout at the base of the cross, a large horizontal border of geometrical elements (lozenges and St Andrew’s crosses) ran below it – a feature which, in any case, resembles the solution adopted for the lower margin of our southeast pier mosaic cross.

A similar cross enclosed in a blue globe was recorded in the opposite (north) wall of the west gallery as well: here the mosaic cross showed rather

larger dimensions (the horizontal arm of the cross measured 1.26 m) but a symmetrical arrangement of the two lunettes is readily perceived30. A noteworthy detail is that, unlike the south lunette, here the mo-saic was better preserved and the Fossatis had the possibility to describe the upper semi-circular cir-cumference of the lunette itself. There ran a double band; the outer one in gold and, beside it, an in-ner border consisting of lozenges and St Andrew’s crosses pattern. This could be the solution adopted for the framing of the mosaic cross panel on the se pier. But let us now turn our attention to the scattered traces of it still preserved in situ / Fig. 9 /.

In the lower part of the pier, just above the vine-leaf stucco moulded frieze, the darkness usually prevailing in the “corridor ” is dispelled by direct sunlight at a certain time of day. There, the lower margin of the mosaic revetment is character-ized by a band decorated with a row of alternating golden lozenges outlined in red and filled with light-green quatrefoil elements on a dark grey-blue background. A slightly different version of the same lozenge border is still distinguishable along the lower edge of the Alexander panel / Fig. 1/ 31. A wide decorative band clearly ran along the vertical right edge of the cross panel, presumably up to the top of it / Figs 8 – 9 /. No room for such a border appears on the left edge, for the body of the “truncated pilaster” creates the above-mentioned asym metry of the wall. Before cleaning, it is impossible to ascertain which kind of decorative motif actually filled the band on the right edge. The roughness and the relatively large size of the tesserae suggest a certain similarity with the technical features of the mosaic in the near-by soffit of the triple arcade which separates the

“corridor ” from the east bay of the gallery and the Pantokrator vault. Moreover, the panel frame and the nearby soffit mosaic seem to be part of the same phase of decoration realized together with the Pan-tokrator vault.

The lack of symmetry in the panel with the mo-saic cross was probably resolved by covering the body of the “truncated pilaster ” with a decorative band which must have been similar or identical to the one still distinguishable on the right edge of the panel / Fig. 8a – b/. The same band could have run along the vertical edges of the recessed niche created by the pilaster as well as along the profile

9 / Southeast pier, particular of the lower edge of the mosaic panel with the cross, Hagia Sophia, Constantinople

245

of the barrel vault above. A layout similar to that adopted for the lunette walls of the west gallery may, after all, have provided a suitable frame for the area of the wall covered by the same gold background of the mosaic cross.

Had there actually been a “set ” of four mosaic panels on the piers?

Questioning the traditional 912 – 913 chronology currently assigned to the Alexander panel, Teteriat-nikov described from new points of view the ceremo-nial costume and garments worn by Alexander 32. She convincingly reaffirms the similar features between the representation of the sovereign in the mosaic and the portrait of the same, together with his brother Leo, as young co-emperors in the manuscript of the Homilies of Gregory of Nazianzus (bnf, Paris. gr. 510). Teteriatnikov’s conclusion is that the mosaic panel was probably executed before 912, i. e. before Alex-ander’s accession to the throne as sole emperor, thus implying a hypothetical intervention by Alexander’s brother, Leo vi, as a patron of this part of the mosaic decoration in the galleries 33.

On the basis of the above remarks we will advance some preliminary hypotheses. First of all, according to the stylistic and technical elements observed on both the panels, we can argue that the cross mosaic could be contemporary with the mosaic of Alexander. Given their position we may suggest a tantalizing hy-pothesis, i. e. that the two panels belonged to a “set ” of four panels, arranged to decorate the inner faces of the four piers (the other two, located on the north-east and southwest piers now being completely lost).

What might have been actually depicted on the other two panels?

We must take into account that Van Nice reported another cross in mosaic, apparently having smaller dimensions, on the inner face of the northeast pier (the east wall of the “corridor ” of the north gallery)

– “symmetrical” with that located on the southeast pier. This latter area, where the mosaic revetment is poorly preserved under the plaster, deserves, for the future, an in-depth survey with specific equip-ment. Therefore we can imagine two mosaic panels with foliated crosses on the two eastern piers and

can suggest two imperial portraits (Alexander and one unknown) on the western ones. Thus we obtain a counterpoint (cross/east ― emperor/west) – for each gallery.

The hypothesis inevitably lead to the following corollary: if the portrait of Alexander was realized in the years immediately before 912, the other portrait one may expect to find on the southwest pier, in juxtaposition with Alexander’s, should reasonably be that of the then (886 – 912) ruling emperor, Leo vi. Although the possibility of a further dynastic set of panels in the galleries of Hagia Sophia would represent a fascinating case, unfortunately we have no real evidence of a panel depicting Leo vi on the southwest pier. It will remain a mere theory, but – in our opinion – not, a priori, to be rejected.

In accordance with Teteriatnikov’s suggestive hypothesis, we will in any case try to place our four panels within the broader context of the dynas-tic patronage. Paris. gr. 510 once again represents a reference model. If we focus our attention on the original arrangement of this manuscript pro-duced in Constantinople ca. 879 – 88234, we see that it opened with Christ enthroned (f. Av). A cross richly embellished with jewels and gold pendoulia (f. Cr; / Fig. 10/ ) appears before the representation of Emperor Basil i between Elijah and Gabriel (f. Cv). This latter is followed by the second imperial por-trait (f. Br), depicting Empress Eudokia with her two sons Leo and Alexander, both bearing the title of despotis they were attributed with respectively in 870 and 879 – the same borne by Alexander in his mosaic. The second cross (f. Bv) closes the series. Such a se-quence, as Ioli Kalavrezou stated in the case of Paris. gr. 510, “was similar to an ivory diptych” with the crosses decorating the reverses of the leaves 35. A sim-ilar sequence was adopted again, due – evidently – to its importance and meaning, in the f. 2r and 3v of

30 Mango, Materials (n. 5), p. 42, fig. 45.31 The lozenge-and-quatrefoil pattern is, by the way, widely employed

in the sixth-century decorative mosaics of St Sophia, and it is con-stantly imitated, still being fashionable, by the Middle Byzantine restorers, so that any comparison could unfortunately entail the risk of unwarranted generalization.

32 Teteriatnikov, “Why is he hiding?” (n. 1), pp. 62 – 69.33 Ibidem, pp. 72 – 73.34 Leslie Brubaker, Vision and Meaning in Ninth-Century Byzantium.

Image as Exegesis in the Homilies of Gregory of Nazianzus, Cambridge 1999, pp. 5 – 6.

35 Ioli Kalavrezou-Maxeiner, “The portraits of Basil i in Paris (gr. 510)”, Jahrbuch der österreichischen Byzantinistik, xxvii (1978), pp. 19–24, sp. p. 24.

the Bible of the patrikios Leo (bav, ms. Vat. Reg. gr. 1), composed most likely after 940 36. Here / Figs 11 –12/, the crosses are not stepped, while inserted into an architectural framework formed by a decorated ar-cade resting on two red-marble columns with gold-en Corinthian capitals 37. The role of the crosses in Vat. Reg. gr. 1 is again that of providing the sym-bolic frame to the scenes where the patrons’ por-traits are represented – that of Leo protospatharios with the Virgin Mary and those of abbot Makar and the late Constantine, Leo’s brother, kneeling before Saint Nicholas. 

Applying this diptych-idea to the decoration of the walls of a building, we may note – as Leslie Brubaker did in her analysis of the Paris. gr. 510 – how the gallery of portraits in the manuscripts found precise correspondence in the mosaics of the Kainourgion, the private rooms that Basil i built in the Great Palace complex. Although now lost, the Kainourgion was extensively described in the Peri Ktismaton and we thus know that it contained a celebrative cycle of wall-mosaics depicting the imperial family. There, mosaics showing the enthroned figures of Basil i and his wife Eudokia Ingerina decorated the walls togeth-er with images of their offspring, sons and daughters,

“holding books containing the holy Commandments that they had been brought up to obey”38. The vault was dominated by a cross in mosaic, thus providing, in Brubaker’s words “a symbol of victory and repre-sentation of God’s approval (…) linked, again, with portraits of Basil and his family”39.

The very same “dynastic aspirations”40 possibly lay behind the decoration of the galleries at Hagia Sophia. Let us think of our four symmetrical panels on the piers of the galleries as if they were detached pages of a manuscript in mosaic, obviously intended as a means of imperial propaganda41. In particular,

36 Irmgard Hutter, “The decoration”, in La Bible du Patrice Léon. Co-dex Reginensis Graecus 1. Commentaire codicologique, paléographique, philologique et artistique, Paul Canart ed., Vatican City 2011, pp. 195 – 272, sp. pp. 270 – 272.

37 Suzy Dufrenne, “Les miniatures”, in La Bible du Patrice Léon (n. 36), pp. 81–194, sp. pp. 103–108.

38 Chronographia Theophanis Continuati (n. 7), p. 293 (§ 89).39 Brubaker, Vision and Meaning (n. 34), pp. 156, 162–163, 170–171, 187;

See also, Leslie Brubaker, “Looking at the Byzantine Family”, in Approaches to the Byzantine Family, Leslie Brubaker, Shaun Tougher eds, Farnham/Burlington 2013, pp. 177– 206, sp. pp. 191, 195.

40  Ibidem, p. 170.41 See Catherine Jolivet-Lévy, “L’image du pouvoir dans l’art byzantin

à l’époque de la dynastie Macédonienne, 867–1056”, Byzantion, lvii (1987), pp. 441– 470, sp. p. 444.

10 / Homilies of Gregory of Nazianzus, bnf Paris. gr. 510, f. Cr, Bibliothèque nationale de France, Paris, 879 – 882

11 / Bible of the patrikios Leo, bav, Reg. gr. 1, f. 3v, Bibliotheca Apostolica Vaticana, Vatican City, after 940

12 / Bible of the patrikios Leo, bav, Reg. gr. 1, f. 2r, Bibliotheca Apostolica Vaticana, Vatican City, after 940

248

42 This concept has been developed by Antonio Iacobini concerning the production of manuscripts: see Antonio Iacobini, “Il segno del possesso: committenti, destinatari, donatori nei manoscritti bizantini dell’età macedone”, in Bisanzio nell’età dei Macedoni (n. 10), pp. 151–194, sp. p. 158. 

43 Jolivet-Lévy, “Présence et figures du souverain” (n. 13), p. 237.44 Ibidem, p. 237.45 For Hagia Eirene at Constantinople and the Koimesis at Nicaea see,

recently, Brubaker / Haldon, Byzantium in the Iconoclast Era (n. 27), pp. 203, 212–214; See Brubaker, Inventing Byzantine Iconoclasm (n. 27), pp. 35 –36, 39 – 41 (with updated bibliography).

46 Brubaker, Vision and Meaning (n. 34), p. 157.47 Anatole Frolow, La relique de la Vraie Croix. Recherches sur le dévelop-

pement d’un culte, Paris 1961, p. 222. More recently, Holger A. Klein, Byzanz, der Westen und das ‘wahre’ Kreuz. Die Geschichte einer Reliquie und ihrer künstlerischen Fassung in Byzanz und Abendland, Wiesbaden 2004, pp. 58 – 68 (with bibliography).

48 Leonis vi Sapientis imperatoris byzantini homiliae, Theodora Antonopoulou ed., Tournhout 2008, p. 240 : “θεμέλιος τῶν ἐκκλησιῶν […] βασιλέων διάδημα […] παντὸς τοῦ κόσμου φυλακτήριον” (16, vv. 195 –196, 198).

the two (?) imperial portraits can also be regarded as some sort of “donor recognition plaques”, a visual reminder of the patronage of the mosaic decora-tion in the vaults 42. In this sense, the proximity – by chance? – between the depiction of the Baptism in the north gallery (central bay, east vault) and the portrait of Alexander on the northwest pilaster has rightly been interpreted by Catherine Jolivet-Lévy as indi-cating a connection. Through the portrait, Alexander, who seemingly favoured the iconography of the Baptism (at least as suggested by a nomisma he struck, representing the emperor blessed by St John Prodro-mos)43, might be seen to be gazing on the represen-tation of Baptism in the east vault. At the same time, he also looked towards the opposite northeast pier, where the mosaic panel with the cross reported by Van Nice was set. Thus far, the southwest pier, just under the Pentecost vault and directly opposite the depiction of the Pantokrator, could be the most fit-ting place for a portrait of the ruling emperor Leo vi, when considering the relevance attributed to the south gallery as a whole.

Once again, iconographies may become the means to express a clear political and religious mes-sage in a particular architectural context like the galleries of Hagia Sophia: human sovereignty as a result of the divine will44.

Let us now turn back to the foliated cross on the se pier / Fig. 7 /. As far as we can judge for the time being, its linearity shows a resemblance to mon-umental examples of the recent iconoclastic past. Both the nearby Hagia Eirene with the gigantic mo-saic cross standing out from the gold background of the apse semi-dome and the similar crosses that should have existed in the apses of the Koimesis church in Nicea and of Hagia Sophia at Thessaloniki are obviously to be taken into account, and the stylistic heritage of the iconoclastic period is not to be ruled out 45.

On the other hand, the ubiquitous simple cros-ses with pearled ends that belong to the decoration conceived by Justinian for the ground-floor vaults and soffits of Hagia Sophia, with their sober and elegant outlines – red or dark blue on gold – played a significant stylistic role as well, thus restraining

the ninth- and tenth-century craftsmen from the risk of the overemphasized embellishment of Paris. gr. 510. Even the capitals ic xc nika hanging from the cross arms in Paris. gr. 510, a reminiscence of Constantine’s vision, seem to have been omitted in our mosaic, although it would have to be cleaned to verify whether this was in fact the case.

In any case, the monumental cross with foliated base in the south gallery of Hagia Sophia might well reflect what Brubaker observed about the ful l-page crosses in Paris. gr. 510 : i. e. that “the cross had shed its brief and temporary role as a barometer of iconoclast and iconophile tendencies”46. We can, nevertheless, add that our cross in its context also constituted a precise reminiscence, once again of the eighth council of 869 – 870, which was held in the south gallery in the presence of the relics of the True Cross, a circumstance the importance of which certainly cannot be underestimated47.

This, indeed, may be one of the sources of in-spiration for the realization of our mosaic cross. It was set out on one of the main piers of the Megale Ekklesia and stood there as if intended as a power-ful reminder of what Leo vi wrote in his sixteenth Homily: “the Cross is the solid foundation of the churches, the crown of the rulers, the phylactery for all the world”48.

These preliminary – and by no means adequate – notes can only constitute the first step toward a more complete survey of this significant mosaic. More precise conclusions will be possible once the plaster is removed and closer analysis officially permitted.

249

Summary / Opomenutá mo-zaika v jižní galerii chrámu Hagia Sofia v Konstantinopoli

Mozaiková výzdoba nacházející se v horních galeriích chrámu Hagia Sofia v Konstantinopoli stále nabízí pozoruhodná odhalení a podněty k novým bádáním, a to i přes podrobnou znalost místa ply-noucí z mnohaletého výzkumu a restaurátorských prací vedených Americkým byzantským institutem (Byzantine Institute of America).

Tento článek se zabývá především obložením čtyř hlavních pilířů v naosu, jejichž mozaiková výz-doba na úrovni galerií zůstala až do 50. let 20. století z velké části ukryta pod tenkou vrstvou omítky. Ta sem byla nanesena během radikálního restau-rování tehdejší mešity Ayasofya, uskutečněného v letech 1847–1849 ticinskými architekty Gasparem a Giuseppem Fossatiovými. Podobně jako ve zbytku budovy vyzdobili bratři Fossatiové omítku dekora-tivními prvky, inspirovanými motivy z Justiniánova období, které se zachovaly na zdech a klenbách ve spodním patře budovy. Toto obložení bylo od konce 30. let minulého století postupně téměř kompletně odstraněno. Na několika místech však ještě zůstává a zakrývá tak původní mozaikovou výzdobu.

Výchozím bodem tohoto výzkumu jsou čtyři vel-ké pilíře podpírající centrální kupoli, jež na úrovni galerií rámují dvě velká sloupořadí (severní a jižní), navazující na ústřední prostor naosu. Obzvláště na západní straně jihovýchodního pilastru je omítka poměrně silně poškozena a pod ní tak můžeme tu-šit monumentální mozaikový panel, na kterém je zobrazen drahokamy vykládaný kříž s akantovými listy, vyrůstajícími z jeho spodní části.

Existence mozaiky byla zaznamenána během pečlivého mapování galerií provedeného Robertem

Van Nice v roce 1960. Van Nice kromě toho upo-zorňuje také na dnes již zmizelé „dvojče” této mo-zaiky, umístěné symetricky na severovýchodním pilíři. O mozaikovém kříži na jihovýchodním pi-lastru chybí zatím podrobná studie, která bude po-chopitelně možná teprve po kompletním odstranění omítkové vrstvy. I přesto na tomto místě autor nabí-zí několik předběžných úvah. Na základě prvního pozorování způsobu pokládání jednotlivých dílků mozaiky a jejího „topografického umístění” se zdá, že by mozaika mohla být současná s portrétem cí-saře Alexandra (912 – 913). Ten je dosud zachován na severozápadním pilastru a byl již podrobně studo-ván Paulem Underwoodem, Ernstem Hawkinsem a v současné době Natalií Teteriatnikovou.

Autor článku tedy navrhuje představit si kompo-zici čtyř symetrických panelů; dvou s monumentál-ními kříži (na jihovýchodním a severovýchodním pilíři), jednoho s Alexandrovým portrétem (na se-verozápadním pilíři), a jednoho s úplně neznámým a ztraceným vyobrazením (na jihozápadním pilíři). Na základě úvah Teteriatnikové, která přesvědčivě navrhuje datovat Alexandrův panel již do let těsně předcházejících prohlášení tohoto panovníka za jediného císaře, tj. do posledních let jeho společné vlády s bratrem Lvem VI. (886 – 912), autor předklá-dá hypotézu, že na jihozápadním pilíři mohl existo-vat další císařský portrét – portrét Lva VI. Pokud by tomu tak bylo, portréty spolucísařů by mohly sloužit jako jakési „značky” donátorů rozsáhlé mozaikové výzdoby galerií, která pravděpodobně začala za vlády Basila I. (867– 886), a to po opravách vrchní části budovy po zemětřesení v roce 869.