The Uncovering of the Mosaics of Hagia Sophia in Constantinople
I sent you a post card the other day from CP,  dearest Mildred; I hope you get it. Now, I can wait no longer to tell you what I found there.
First of all, St. Sophia  is no longer a mosque, thank God : Those horrible great round green shields with Koranic inscriptions on them that defaced the pendetives of the main cupola have been taken down. The carpets which, laid in the sense of the Mihrab (and not in the axis of the building) distracted the eye and gave the place a circus-tent look, have been removed, revealing the original pavement of cipollino and verde antico. This is also an immense improvement. And little by little it may be hoped that all the Moslem junk, Mihrab, this, that, and the other, will go, including the pastiche pavement laid askew across the entrance to the E. end (so as to be in proper relation to the Mihrab). But of course they have to go slow, as there are many ill-wishers. Another danger, at times serious, is that the church shd. be turned into a museum to the extent of being filled with show-cases and God knows what, reconstruction of dynosaurs and stuffed crocodiles, for choice . The Turk is apt to shake his head and say that the building, now that it's not a mosque, is dead, and that it must be used for some living purpose. Halil bey,  formerly head of the Museum and now the sort of local Koechlin,  takes this line. Fortunately, Aziz bey,  the present head of the Museum, doesn't. So there's hope.
And now for the mosaics!  From what has been discovered by now, it is certain that the original plan, under Justinian, didn’t provide for any figure mosaics at all. There were only crosses and formal (most of them textile) patterns. As you’ll have realised from the photos, of the interior, the space available for mosaics isn’t large in relation to the building, for all the interior, up to the spring of the semi-domes, is wainscoted in marbles. Thus, apart from the vaults, the gallery vaults and walls and those of the narthex, the mosaic surfaces are limited to the cupola and its pendentives, the semi-domes, the soffits of the huge main arches, and the spaces, between the windows (3 tiers of them), under the N and S main arches supporting the cupola (see plans in Whittemore’s  reports). Here, there is plenty of figure mosaic, all of it post-iconoclastic-struggle, of course. Much of this mosaic was sketched by Salzenberg  a century ago, when the Fossati  carried out their restoration and uncovered as much as they could (only to cover it up again), but not all. Whittemore has already found a lot unknown to Salzenberg.
Wllittemore’s two published reports, which you’ve seen, deal with the vestibule and the narthex. In the vestibule, there is recul  enough to see the composition of the Virgin and the two Emperors, over the door leading into the narthex, and the light is good. The prostrate Emperor (Leo VI, late IXe cent.) before Christ in the lunette over the central one of the 9 doors leading from the narthex into the interior, is not so fortunate.
The lunette is too high, too narrow the narthex, relatively, for this picture. Only by retreating through the door leading into the outer narthex (exonarthex), can one see it, pretty well. I felt that the artist had been cramped by the position and size of the lunette to be filled by the composition desired by Leo the Sage (who like many sages, was a very foolish person). The artist made up for all this by pulling off a most astonishing symphony, or better synchromy, in pale buffs and golds and dark silvery greens – but there it is: the original Justinian cross, when all is said and done, was better adapted to the space.
The narthex, facing W. whence much rain comes, has suffered, in the vaults especially, and the mosaic there is pretty patchy (also in the ground of the Leo VI composition), a lot of the Fossati paint remaining to fill up holes.
Inside the church, aside from some very beautiful textile-pattern mosaics which have been freed, W. has concentrated on the S. gallery and the E. semi-dome. Here he has made magnificent discoveries. You’ve perhaps seen photos, of some of them, but I’ll go ahead just the same: Imperial portraits. Deesis group and Virgin and Child.
The Imp. portraits are on the E. wall of the S. gallery. Two groups of them: 1) Constantine X Monomachos and his Empress Zoë, with Christ between them, all life size or a trifle over; 2) John II (καλοριάννυς = Handsome Jack) with his spouse Irene (a Hungarian princess, and a regular butter-blond, with a fat Gretchen-like  mug, and two long braids of yellow hair like Melba  doing Gretchen ) the Virgin between them and, L. of Irene, their son Alexios.
Group 1) has been marred by having all three faces redone: I can’t imagine why. It must have been, done not very long after the original work. Group 2) is very well preserved – down to the waist, about; for below the waist all the cubes of these groups (including the Deesis) had evidently been poked out by Turkish guardians to give as souvenirs to visitors.
The Deesis, also in the S. Gallery, has rather less left than the others, but it is a marvel of beauty – all 3 heads: the Christ, the Virgin and St. John Baptist.
This is by far the most beautiful mosaic I’ve ever seen, and it shows where Daphni comes from, and the Sicilian mosaics. Its perfection and accomplishment, in drawing, colour and cube-setting, are amazing. Looking at it, I felt that if Rubens could have seen it, he’d have sat up and sneezed – It has all his lightness of touch and of colour, richness of impasto. Until one has seen that, one doesn’t know what XI cent. mosaic is. Even the Pródromos  (St. John B.) at Daphni pales a bit before this Pródromos – tho’ it’s honourably near, and very grand. (I don’t know if I told you in my card, but I went to CP via Saloniki,  Hosios Loukas  and Daphni,  and end July I saw Palermo  and Cefalú,  so I had these things in my eye).
The Virgin in the E. semi-dome is on a colossal scale – I suppose some 6 metres tall – but I’m not good at sight measurements. She is a grand vision, seated on her throne, with the Child in her lap. When one is on the face of the mosaic – it was still being worked on when I was there one sees that the cubes forming her robes are all of varying shades of green (a surprising colour for her to wear) descending to almost black. But when one retreats as far as the scaffolding allows one, (one can’t see her at all from the floor at present) her robe becomes blue: A full marine blue. The Child’s is gold, with brown shadings. A (to me) unaccountable detail: the cushion the Virgin is seated on is green, with yellow lights. When one is on the mosaic, the green in the cushion seems to be of the same green as much of that in the robe. Well, when one goes back, and the robe becomes blue, the cushion stays green! Is it kept green by the yellow lights? Is the robe turned blue by the apparently black (blue-black perhaps) shadings?
The semi-dome that carried this Virgin collapsed towards the middle of the XIVe Cent., was promptly rebuilt and the present mosaic was executed about 1350,  with the aid, as Whittemore says he has contemporary literary evidence to show, of subscriptions collected in Russia. It is clear that the present mosaic reproduced as closely as possible that which was lost when the semi-dome fell – the earlier one is known to have been very celebrated, and many copies of it on a small scale no doubt existed. The artists seem to have tried to work in a style several centuries earlier than that of the XIVe, which is well-known from the CP mosaics of that date in the Chora (Kahrié Djami)  and the Pammacháristos (Fetieh Djami)  and the Apostles at Saloniki. And with much success, though on close inspection the system of drapery gives the thing away. Iconographically, they were faultless. Morey,  of Princeton, was at CP just before me. W. took him up and showed him the E. semi-dome Virgin – without recalling the collapse of that semi-dome in the XIVe – Morey, on seeing it, said “Yes, certainly Xe cent.; perhaps even IXe.” Which is a neat illustration of the limitations of the iconographic approach to Byzantine art.
By the way, this story – and everything else in this letter – is highly confidential. W. begged me to regard all I saw as confidential. I can’t believe he meant it literally, but you know him and can judge. I pass the seal of secrecy on to you.
Well, the foregoing is what has been uncovered so far, in figure mosaics.
W.  is now working on a huge archangel, on the S. face of the arch in front of the E. semi-dome. On the same scale as the Virgin, he was one of her two guards. Whether his colleague, on the N. face, is preserved or not W. doesn’t yet know.  But the one on the S. face is very well preserved indeed: enough tests have been made to establish that. And he may be of the early Macedonian period: X or even IX – after 842,  when images were finally restored. You may imagine with what thirst I await the revelation.
Then, there’s a Pentecost  – the 12 Apostles with the Holy Ghost descending on them, in the S. gallery vault, near the Deesis – and God knows what all. The whole thing is going to be an incredible revelation of Byz. art of the capital itself, at the highest power, during the Macedonian period, with Comnenian and Paleologue stuff as well. Its importance for the history of Byz. art is, I should almost venture to say, greater for the post-iconoclastic periods than that of all the mosaics known hitherto. Indeed, I don’t think this statement is an exaggeration. Of course, there are schools which, as far as one can tell at present, don’t seem to be represented at S. Sophia, where only art of the most accomplished, metropolitan, courtly, polished, was admitted. There’s probably nothing there in the vein of S. Sophia at Saloniki  – and I adore those rather wild, fauve, Matisse-like apostles and angels, so remote from the courtly, but so enchanting. Nor like Hosios Loukas in Phocis. But S. Sophia CP represents the central Byz. stream, there’s no doubt about that, and all the rest is more or less excentric. The discovery of the CP mosaics is by far the most important event in the whole annals of Byz. studies.
Whittemore is doing his work well. Happily, the fabric of S. Sophia – barring earthquakes and explosions – gives rise to no anxiety. If W. had run into the necessity of rebuilding piers and arches and vaults, as Marangoni  has had to do in Venice, he wouldn’t have been able to apply his technique, and as he isn’t an architect he presumably would have had to pass. But given the state of the S. Sophia structure, his technique is excellent – couldn’t be better, and he is applying it with increasing skill and resourcefulness as he gains experience.
He has had most trouble with the mosaics on which the Fossati worked most – i.e. those in the narthex, and they were the first he tried his hand on. W. at any rate started on the principle that the Fossati campaign was part of the history of S. Sophia, and should be preserved where it does not obscure original Byz. work. Thus, in the Leo VI lunette, where the Fossati filled up a hole with plaster and then painted on it in imitation of mosaic, or reset cubes (the whole thing was afterwards covered up with plaster and paint, of course) W. has left the Fossati daubs, and there are not a few of them in the field of the Leo VI group, producing a disconcertingly patchy impression. I think he’d have done better to remove all the Fossati stuff, and to have reduced the plaster to a neutral tone by the excellent expedient, which he applies elsewhere, of rubbing with cob-web: as he says, the colour of Time.
I conveyed this criticism gently to W., and he said that if he’d known, when he started, all he knows now, he’d very likely have done this. There’s no harm done, and he can still do it – it wouldn’t be a big job – in the Leo VI lunette, The formal designs in the Narthex vault, which have been treated in the same way, present a bigger problem (in time), and one that is much less urgent than the prosecution of the work on the figure mosaics inside the Church. Here, on those he has uncovered so far, there’s no Fossati daubing at all. The Fossati’s never found these mosaics. And there may be none on any of the inside figure compositions. W. proceeds, first to consolidate, and then to clean. In order to consolidate, however, he has to find out what there is left, and thus has to remove the oil paint with which the Fossati covered all the surfaces that are not marble-sheathed, first of all. He does this, and all the subsequent cleaning, by mechanical means, without acid or liquid, even water, thus avoiding any running of the Fossati paint, in solution, into the interstices between the cubes. He just works away with a variety of tools, scraping like a dentist cleaning teeth.
Of course the oil paint can be chipped off without the infinite care that has to be lavished when the light plaster couche under it is removed. But even the removal of the oil paint is a very long business. When he has got it off, he sees the cubes (tho’ not their colour) and can feel, with his hand, what state the intonaco that bears them is in. Very often it has blistered off the brick wall, is just hanging like blistered wallpaper, and has to be consolidated before anything further is done.
He proceeds, first of all, to anchor the mosaic island, and so to prevent it from shifting on the wall, it or any part of it. He does this by surrounding the mosaic with a zone of new plaster, full of copper clamps which he runs into the brick (or stone) of the wall, the heads being embedded in the plaster. When this has been done, he consolidates the mosaic itself by means of smaller copper clamps, of hairpin shape with the two ends bent at right angles. All over the mosaic, in places where a cube or two are missing, he inserts these clamps, running them through the intonaco (which carries the mosaic) and a couple of inches into the wall behind. All that appears on the surface are the bent-over ends – usually imbedded in new plaster filling old holes, or if the chink is very small, the heads are left visible, and as they are about the size of a small cube they can only be distinguished when one is right on the face of the mosaic.
The mosaic has thus been pinned securely onto its wall, and at the original level (which is necessarily disturbed when new cement is squirted in to bind the blistering intonaco to the wall, as at Ravenna). And this level, with its intentional irregularities, is a component part of the mosaic.
Then starts the endless business of cleaning off the coating of plaster, done as with the paint by exclusively mechanical means, and dry: no acids, no liquid. Some 25 different tools, of metal, wood, glass, wire-brushes and air-syringes to blow the dust out of crannies, so that the intonaco’s own colour may show, also. Where cubes are wobbly, they have to be fixed with new intonaco, which is then reduced to neutral tone (like the holes that have to be plugged) with cob-web. W.’s men, English and Americans, the best of them lent from the Brit. Mus. and the Office of Works, proceed with the greatest care, and W.’s eye is on them all the time. The results, especially where there is no Fossati problem, appear to me to be admirable. He never replaces a cube.
Happily, the Turks regarded these mosaics as heretical and to be covered up, certainly, but with no little awe. When covering them, they took care not to injure them. The Moslem of course regards Christ as a prophet, and the Virgin, Bibi Miriam, as a very great celestial personage – also the angels. When it comes to Christian Saints, he might be less respectful, but even here he’d tread warily, lest he might incur the wrath of some malevolent and puissant being. So there was, in all probability, no Turkish destruction of these mosaics. The weather, esp. in the W. and N. galleries, caused enormous losses – huge surfaces of mosaics just having fallen out – and of course the Turk did nothing to hinder the process until the Fossatis were called in a century ago. But he didn’t destroy wilfully except what the cube-hunting visitors, and guardians supplying them, did in the galleries, and even there they seem to have spared the faces, and equally important, he didn’t try to restore, as the Italians have done.
It seems too good to be true that there is such a mass of the noblest mosaics ever created, waiting there to be revealed. What irony that this should be at the moment when El Greco’s masterpieces at Toledo are reported missing! All this to be had, if W. finds the money to proceed. And I needn’t say that in the whole field of art, there’s nothing that seems to me to touch this work, for importance, and for the unutterable joy these things give when they are uncovered.
W. was very kind to me, allowed me to spend all the time I wanted, during the 4 days I spent at CP., on the scaffoldings, to see his men work, talk with them, examine everything. He gave me no photos – and I, knowing my W., asked him for none, except of things he has already published. I think, and certainly hope, we’ve found a modus vivendi. Of course I’ll do all I can to support and help him – the work is of such compelling splendour that there can be no question of sparing any trouble to that end – and I only wish I could be sure of doing something effective. All I ask of him is to give me photos, of everything he publishes as soon as he has published it. He has promised he’ll do this.
The object of this letter, besides relieving my feelings at not having had you there, is to put it to you that when you next come to Europe . . . I’ll say no more.
 Hagia Sophia, the former Early Byzantine basilica, later a mosque, and now a museum in Istanbul, Turkey. Built in 537, until 1453 it served as the seat of the Patriarch of Constantinople and then as a mosque from 1453 until 1931. It was secularized and opened as a museum in 1935.
 The initial plan for the museum was to have display vitrines with Byzantine and Ottoman objects. However, when it opened, there were no exhibits. See Wendy Shaw, “Museums and Narratives of Display, from the Late Ottoman Empire to the Turkish Republic,” in Muquarnas, An Annual on the Visual Culture of the Islamic World: History and Ideology: Architectural Heritage of the “Lands of Rum,” 24, ed. Julia Bailey, Sibel Bozdoğan, and Gülru Necipoğan (Leiden: Koninklijke Brill, 2007): 270.
 Halil Edhem Bey (Halil Edhem Eldem) (1861–1938), a Turkish archaeologist, who in 1892 became the second director and later, in 1910, Director General of the Archaeological Museum of Istanbul until his retirement in 1931.
 Raymond Koechlin (1860–1931), a personal friend of the Blisses and the Tylers, the Alsatian-born Raymond Koechlin collected Chinese, Japanese, Islamic, and medieval art and acquired works by the Impressionists and Post-Impressionists. He displayed his collection crowded together in a small apartment at 24, boulevard Saint-Germain. This grand amateur wrote a number of books about the arts that he collected, but he is best known for his studies of French Gothic sculpture (especially ivories), Islamic ceramics, and Japanese art. At the end of his life, Koechlin wrote a privately printed memoir about French collectors of Asian art, titled Souvenirs d’un vieil amateur d’art de l’Extrême-Orient (1930), which he sent to the Blisses. He served as a curator at the Musée du Louvre and made significant donations to the museum after his death in 1931.
 Aziz Ogan (Aziz Ogan Bey) (1888–1956), director of the Istanbul Archaeological Museum between 1931 and 1954.
 On June 7, 1931, Thomas Whittemore (1871–1950), director of the Byzantine Institute, received permission from Mustafa Kemal Atatürk (ca. 1881–1938) to uncover the Byzantine mosaics of Hagia Sophia. Whittemore and his team uncovered fragments of sixteen figural mosaics in the vault southwest of the gallery. When in 1934 Hagia Sophia was secularized, Whittemore was able to uncover additional mosaics, work which continued over the next eighteen years. See Wendy Shaw, “Museums and Narratives of Display, from the Late Ottoman Empire to the Turkish Republic,” in Muquarnas, An Annual on the Visual Culture of the Islamic World: History and Ideology: Architectural Heritage of the “Lands of Rum,” 24, ed. Julia Bailey, Sibel Bozdoğan, and Gülru Necipoğan (Leiden: Koninklijke Brill, 2007): 270 and n. 47.
 Thomas Whittemore (1871–1950), an American art historian, was born in Cambridge, Massachusetts. He graduated from Tufts College in Massachusetts with a BA degree in English in 1894 and took graduate courses at Harvard University through 1898. Beginning in 1902, at Tufts College, he gave lectures on ancient and medieval art and, by 1906, on art history. Whittemore met the art connoisseur Matthew Prichard (1865–1936) when Prichard was at the Museum of Fine Art, Boston. Prichard introduced Whittemore to the artist Henri Matisse (1869–1954), who became his life-long friend, and to Byzantine art, which became his passion. In 1911, Whittemore became the American representative for the British Egyptian Exploration Fund and participated in the excavations at Abydos and Balabish. He taught art history at New York University between 1927 and 1929 with the rank of assistant professor. In 1930, he established the Byzantine Institute of America in Boston with a research center in Paris and a field office in Istanbul. In 1931, he began the work in Istanbul of uncovering the Byzantine mosaics of the church of Hagia Sophia which had been covered over in 1849. Whittemore became keeper of Harvard's Byzantine coin collection in 1933 and was named a research fellow in Byzantine art in 1938. He retired from Harvard in 1942. He died in 1950 in Washington, D.C. A character modeled on Thomas Whittemore, Professor Darchivio, appears in Edith Wharton's novel Glimpses of the Moon (1922).
 Wilhelm Salzenberg (1803–1887), a German architect who surveyed Hagia Sophia in 1851–1852.
 Gaspare (1809–1883) and Giuseppe (1822–1891) Fossati, brothers and Swiss architects who undertook renovations in Hagia Sophia in 1847–1849.
 Gretchen, a fictional character from Faust by Johann Wolfgang von Goethe (1749–1832).
 Dame Nellie Melba (1861–1931), an Australian operatic soprano.
 A role frequently-sung by Nellie Melba was that of Marguerite (Gretchen) in Charles Gounod’s opera Faust, which she had studied under the supervision of the composer.
 Greek for “the forerunner.”
 The fourteenth-century Byzantine Church of the Holy Apostles in the Greek city of Thessaloniki.
 The mid tenth-century Byzantine Church of the Theotokos (Panagia) in the monastery of Hosios Loukas in Boeotia Greece.
 The eleventh-century Byzantine monastery of Daphni in Athens, Greece.
 The twelfth-century Norman-Byzantine Church of Santa Maria dell'Ammiraglio (the Martorana).
 The twelfth-century Norman-Byzantine Cathedral of Cefalù.
 The apse mosaic is now traditionally dated ca. 867, having been inaugurated on March 29, 867 by the Patriarch Photius (ca. 810–893).
 The eleventh-century Church of the Holy Saviour in Chora (Kariye Camii), Istanbul.
 The eleventh-twelfth-century Byzantine Pammakaristos Church (Church of the Thotokos Pammakaristos) (Fethiye Camii), Istanbul.
 Charles Rufus Morey (1877–1955), an American art historian and professor and chairman of the Department of Art and Archaeology at Princeton University between 1924 and 1945. He was best known for his expertise in medieval art and his Index of Christian Art.
 Thomas Whittemore.
 The partially-preserved Archangel Michael mosaic, north side of the bema of the apse.
 The Second Iconoclasm period, between 814 and 842.
 This mosaic probably was destroyed in the Istanbul earthquake of June 1894.
 The eighth-century Byzantine church of Hagia Sophia, Thessaloniki.
 Luigi Marangoni (1872–1950), an Italian architect and the proto or custodian of the Basilica of San Marco in Venice. See Otto Demus, The Church of San Marco in Venice: History, Architecture, Sculpture (Washington, D.C.: Dumbarton Oaks Research Library and Collection, 1960), 198.