The Marble Revetment of the Piers of the Dome of the Rock

by Terry Allen         published by Solipsist Press, Occidental, California, 2014

Copyright © 2014 by Terry Allen         An electronic publication     ISBN 0-944940-13-7

Table of Contents

Visual Evidence
Pre-1873 Views of the Interior
Ordnance Survey Photograph
Engravings after Catherwood
Arundale's Section
De Vogüé's Temple de Jérusalem
Orientalist Paintings
Two Color Lithographs After Drawings by Francois Edmond Paris
Watercolors by Carl Werner
A Watercolor by William Holman Hunt
A Watercolor by Carl Haag
A Watercolor by William Simpson
Later Views of the Interior
The Scroll Friezes
The Scroll Frieze of the Octagonal Arcade
The Scroll Friezes of the Inner Arcade
The Scroll Frieze Above the Inner Arcade
The Style of the Scroll Friezes
Carved Spolia
Panels Published by Creswell
Early Byzantine Spolia
Other Revetment
Inner Sides of the Inner Piers
Outer Sides of the Inner Piers
Left and Right Sides of the Inner Piers
The Iron Grille and the Pier Revetment
Carved Frames
Billet Moldings
Short Friezes
Piers of the Octagonal Arcade
The Original Design and Its Sources
Appendix A. The Directionality of the Scroll Frieze of the Octagonal Arcade
Appendix B. The Marble Revetment of the Great Mosque of Damascus
Lozenge Panels
Lost Carved Ornament


The piers of both interior arcades of the Dome of the Rock (the foundation inscription of which is dated 72/691–92) are revetted in marble, some of it carved in low relief.1 Some of the original decorative scheme of that revetment, now changed considerably, can be recovered. In this article I describe the evidence I have used to recover that scheme and how I have interpreted it.

Pier Naming and Numbering

I follow K.A.C. Creswell's usage in referring to the piers,2 with a small difference. The inner, circular arcade is composed of the northwest, northeast, southeast, and southwest piers. They have inner and outer sides, and I refer to the other sides of all the piers as right and left (R and L), as they would be seen from the Rock. In other words, I look from inside to outside. Creswell used this approach in the captions of sixteen photographs of the mosaics of the inside of the outer, octagonal arcade, but in three captions of photographs of the mosaics of the outside of the inner arcade piers he turned around, as it were, and labelled right and left as seen from the outside looking inward. I find it easier to keep things straight with a single (virtual) viewpoint.

Creswell numbered the piers of the outer arcade in the captions to his pl. 11–21, and inconspicuously on his plan (fig. 21). Pier 1 is that inside the south entrance, to the right as one enters, and the numbering continues counterclockwise, ending with Pier 8, inside the south entrance to the left.

Visual Evidence

I have not made a thorough search for early visual evidence for the interior of the Dome of the Rock. My principal visual sources are the photographs made by the photographic department of the American Colony in Jerusalem (glass plate negatives and prints, including stereo pairs) from the Matson Collection in the Library of Congress. These photographs have been scanned at multiple resolutions, and I have used the highest available, which is in TIFF format.3 I refer to them by the unique alphanumeric string that appears in their filenames (which is the same for scans at all resolutions), e.g., 03231u. The Library of Congress dates these photographs “approximately 1900 to 1920”, apparently on the basis of the history of the American Colony's photographic division. They all appear to show the same state of the building, which is that after the restoration campaign that began in 1873 and before any of the many repairs and restorations of the twentieth century.

I have also made use of the photographs and line drawings in Creswell's Early Muslim Architecture, 2nd ed., v. 1, pt. 1, and the broader selection of views in Saïd Nuseibeh and Oleg Grabar, The Dome of the Rock (N.Y., 1996, cited below as Nuseibeh). And I have used photographs from the Creswell Archive at the Ashmolean Museum, Oxford (, which are numbered beginning with “EA.CA”; I cite them by those numbers. Creswell's account of the Dome of the Rock, while not comprehensive, is still the most useful in understanding the physical building. I have used his bibliography to locate pre-1873 views of the interior, and have come across a few views not mentioned there.

For the chronology of late Ottoman restorations of the Dome of the Rock I rely mainly on Beatrice St. Laurent, “The Dome of the Rock: Restorations and Significance 1540–1918”.4 The restoration St. Laurent dated to 1291/1874–75, apparently on epigraphic evidence, seems to have been the most extensive of the later nineteenth century. It is also significant in that Western observers were present for various periods during the progress of the work. Several are listed in the Survey of Western Palestine, Jerusalem,5 according to which work began in 1873.

Pre-1873 Views of the Interior

Predating the American Colony photographs are at least one other photograph, various watercolors, and unpublished drawings from which derivative engravings were made.

Ordnance Survey Photograph

A photograph published in the Ordnance Survey of Jerusalem was of course taken before its publication in 1865, but likely not long before.6 It is a view to the west between the two arcades, inside the south entrance, with Pier 8 partly visible behind several chandeliers (cf. 06639u and 11793u, with later chandeliers). This location would have been appealing to an early photographer looking for a representative view of the interior, as opening the south doors would have provided more light than was available elsewhere, and early photography required strong light (possibly, too, these doors were most commonly left open, as they face the Aqṣā Mosque). This photograph shows nothing different from the American Colony photographs regarding the piers, and provides no otherwise missing detail of them.7

Engravings after Catherwood

Figure 1. Survey of Western Palestine, Jerusalem, p. 249.

Several published engravings, in quite different styles, are based on now-lost drawings by Frederick Catherwood, who was in Jerusalem in 1833 and drew with the aid of a camera lucida.8 The most careful engraving was published in the Survey of Western Palestine. It is a view in the opposite direction of the same part of the building shown in the Ordnance Survey photograph, with part of Pier 8 on the right, a bit of Pier 1 in the background, and the southeast pier of the inner arcade mostly occluded by two columns (cf. 03231u).

So far as I have checked, where decorative details are shown in this engraving they match those in the American Colony photographs. The scroll frieze of the southeast pier is shown correctly, as is the ablaq arch of the inner arcade and the adjacent marble revetment. Quite a bit of detail is omitted, the areas in question simply hatched. The decoration of the inner side of Pier 8 matches the American Colony photographs (see 05028u and 06639u). The veining of the marble in the leftmost column is apparently invented. Two striking differences from the American Colony photographs and the Ordnance Survey photograph are the lack of elaborate chandeliers and the appearance of what looks like a light screen on a thin wooden framework above the iron grille around the Rock, as if to keep out birds. I know of no other source that shows or mentions such a screen, but it would have been a good idea from all but the aesthetic standpoint.

Two engravings of low quality, one a reduced and less detailed vignette of the other, were published by James Fergusson after another Catherwood drawing, which gave the view seen in the Ordnance Survey photograph: to the west between the two arcades, inside the south entrance.9 They show different revetment on Pier 8 L than appears in the Ordnance Survey photograph; the same revetment is shown on Pier 7 L. In both cases rectangles (perhaps four) of darker stone appear beneath the poorly drawn arcade frieze near the top of the pier, occupying most of the width of their register, and a very large rectangle of darker stone occupies most of the space below, above a single register comprising the base. While it is possible that this revetment was changed between 1833 and the taking of the Ordnance Survey photograph, it is more likely that the engravings are inaccurate. All the piers had a register of decoration above the base after the restoration of 1873–75 and the Ordnance Survey photograph shows this register on Pier 8 too. The large rectangles of darker (presumably colored) stone might be exaggerations of the rectangular piece of colored spolia on Pier 8, duplicated on Pier 7, or of the rectangular pieces of colored marble on another pier (such as sides of Piers 2 and 3). Similarly, the smaller rectangles above might have been borrowed from Pier 2 or 6.

There are also incorrect details in the fuller version, which shows the dikkah, but misunderstood. The stairway to it is surely rendered incorrectly as it blocks the entrance to the cave below the Rock. The scroll frieze on the southeast pier, at right, is incorrect and the rest of the outer side of that pier is shown as an unbroken expanse of veined marble, as are the frames of the large rectangles of darker stone on Piers 7 and 8, which cannot have been the case. The designs of the mosaics of the soffits of the arches between Piers 7 and 8 do not match Creswell's pl. 23 and 24.

It is not possible to conclude in what details these derivative engravings accurately reflect Catherwood's drawings, or whether Catherwood was accurate in matters of detail. Details could have been added later rather than recorded on the spot (the value of using a camera lucida would have been to get the main lines of a view right rather than capturing everything visible); if not they could have been invented by the engravers.

Arundale's Section

Catherwood was accompanied and assisted by Joseph Bonomi and Francis Arundale, and a now-lost drawing of a section of the Dome of the Rock by Arundale was published as an engraving by Fergusson.10 It is of much higher quality than any of the engravings after Catherwood. It shows the building cut along the north-south axis, viewed to the west, with schematic and repetitious detail of the ornament. This ornament includes the revetment, in particular that of the two western inner piers, the arches, and the spandrel zone. But here, too, the detail cannot be trusted (details below). All the detail of all the ornament was probably filled out from samples drawn separately from the main lines of the section.

Fergusson also published an engraving titled “Order of the Dome of the Rock” after a drawing by Arundale; it is inaccurate in most details.

De Vogüé's Temple de Jérusalem

Figure 2. De Vogüé, Temple, pl. 19.

In 1864 a grand volume by Charles-Jean-Melchior de Vogüé, Le Temple de de Jérusalem: Monographie du Haram-ech-Chérif suivie d'un essai sur la Topographie de la Ville-Sainte, was published in Paris, London, and Liége. It covered work he had done in Jerusalem beginning in 1853. He devoted nineteen pages to the Dome of the Rock, illustrated by a plan, section, and ten color plates.

Figure 3. De Vogüé, Temple, pl. 20.

Figure 4. De Vogüé, Temple, pl. 22, left.

Of these the section and two plates of the interior decoration are relevant, and I shall consider them below.

Orientalist Paintings

The views discussed above were made by men whose interest was straightforwardly archaeological, although no doubt coupled with an appreciation of the exotic and colorful nature of the subject matter. Other views were painted by men whose interest was primarily the exotic and colorful subject matter, who are called today “Orientalist”. Despite their Romanticism, these painters often took care to observe and record their subjects accurately (at least in some details). I have found some such paintings either executed before 1873 or likely to have been executed later at least in part from sketches made before 1873. No doubt there are many others.

Two Color Lithographs After Drawings by Francois Edmond Paris

The French admiral Francois Edmond Paris commanded a naval squadron that visited Palestine, and he was in Jerusalem in 1861. His Souvenirs de Jérusalem was published the following year, with fourteen plates of which two are views of the interior of the Dome of the Rock: one to the west from the octagonal arcade, across the area inside the south entrance; the second from an elevated viewpoint corresponding to the eastern end of the dikkah, to the north-northwest, across the Rock.11 The first of this pair of plates shows usefully Pier 1 L; the second shows the inner and right sides of the northeast pier and less clearly the inner side of the northwest pier, as well as the interior of the dome up to the apex.

Watercolors by Carl Werner

According to the Bayerischen Staatsbibliothek's Deutsche Biographie project the German painter Carl Werner visited Jerusalem in 1862.12

The Victoria and Albert Museum possesses a watercolor by Werner titled “Interior of the Dome of the Rock, Jerusalem”, dated to 1863.13 It shows the right side of the southeast pier.

Another Werner watercolor, titled “The Holy Rock”, dated 1864, was auctioned in London in 1983, but I do not know where it is.14 It shows a man praying on a carpet actually on the Rock itself and does not show any useful detail of the marble revetment.

A Watercolor by William Holman Hunt

A watercolor by William Holman Hunt, dated 1869, was painted in Jerusalem. It is a view to the west from the outer ambulatory across the area inside the south entrance, showing Pier 1 L in the middle ground, Pier 8 L more distantly, and part of the outer side of the southeast pier; part of the top of the left side of the southwest pier seems to be shown also, over the entrance to the cave below the Rock. Hunt first visited Jerusalem in 1854–55 and visited the Dome of the Rock, then in 1869–72 and 1875–78. He was religious and in 1869 thought the Dome of the Rock stood on the site of Christ's tomb. He was interested in archaeological detail, and as the watercolor accords well with photographs, I think it is probably accurate.15

A Watercolor by Carl Haag

Carl (or Karl) Haag, a German-born painter working in England, visited Jerusalem at some date I have not discovered, but from the dates of his published engravings, before 1873. A painting of his from 1891, apparently titled The Holy Rock, Summit of Mount Moriah, Jerusalem, may be based on sketches from that visit (or a later one, or may incorporate visual material created by others, including from after 1873).16 It is a view to the southeast across the Rock, with parts of the inner sides of the southeast and southwest piers visible.

A Watercolor by William Simpson

William Simpson, best known as a correspondent for the Illustrated London News, visited Jerusalem, it seems in 1869. In 1872 he exhibited forty views of Jerusalem with a catalogue titled Underground Jerusalem. Numbers 32 and 33 were of the Rock, the former apparently a watercolor “from the north-east corner” and costing 120 guineas, the latter, a “drawing” showing “the north end of the rock” and priced at 35 guineas. He continued to produce views of Jerusalem over the years. A watercolor by Simpson sold at auction in 1994 and now in a private collection has been published,17 and I found an illustration of it online, described as “signed, inscribed and dated ‘Wm Simpson. 1887./The Sakhrah [etc.]’”.18 Its viewpoint does not fit either of the Underground Jerusalem descriptions, so it is either a different view or Simpson had his compass points confused in that publication.

Later Views of the Interior

There are not a few views of the interior of the Dome of the Rock drawn or painted after the restoration of 1873–75; none I have found is much help.

The Scroll Friezes

Scroll friezes carved in low relief occur in three places in the interior of the Dome of the Rock. In this article I use Robert Hamilton's terminology: the vine (or whatever it is) running along the frieze is a scroll, and each cycle of foliage branching off it is a coil.19

The Scroll Frieze of the Octagonal Arcade

Figure 5. Portion of Tie Beam and Part of Pier 4, 06635u, detail.

One frieze forms the lowest register of the entablature that runs along the tops of the piers of the octagonal arcade and the lower edges of the tie beams, on the interior side only. It is not continuous but broken into sections that generally alternate in the direction of their scrolling. I discuss these changes in Appendix A. De Vogüé's pl. 20 shows a portion of this scroll frieze colored gold with a blue background (Figure 3), and he described it so (op. cit., p. 83). This scroll frieze appears to me to be Umayyad in design, as has always been thought.

The Scroll Friezes of the Inner Arcade

Figure 6. Upper Part of Northeast Pier, Inner and Right Sides. 06636u, detail.

At the top of each pier of the inner arcade a scroll frieze, different from that of the octagonal arcade, carved in low relief, and raked, runs along each side, though not continuously: each side is a trapezoid composed of mostly rectangular stones a bit narrower than the coils of the scroll, with a double fillet running horizontally above and below and a single fillet at the corners. There is an integral number of coils on each side, six on the inner sides of the piers, seven on the outer sides, and five on the left and right sides, for a total of twenty-three. Because of the odd number of coils it would not have been possible to alternate them above and below the stem regularly. The vertical fillet obscures the the scroll turning the corner, as shown in Figure 6, where two above-the-stem coils are carved on either side of the vertical fillet and could not be joined correctly. As a result of the masking effect of the fillet, however, the irregularity is hard to see, and the overall composition is harmonious. The scrolls run left to right on all sides but two, the inside and left side of the northwest pier.20

De Vogüé's pl. 22 (Figure 4) shows a segment of this frieze as entirely golden, which is less plausible than that it had a colored background in his day.

The Scroll Frieze Above the Inner Arcade

Figure 7. Scroll Frieze over Northwest Pier and to East; 00526u, detail.

Finally, just below the drum, above the marble revetment of the spandrels of the arches, there is a taller and more elaborate scroll frieze than that of the inner arcade piers. Immediately above it is the lowest of the three cornices in the drum.21 This frieze is carved in low relief on rectangles of stone generally about two-thirds the width of one of the coils of the scroll and runs continuously, left to right.

The Style of the Scroll Friezes

The coils of the Umayyad scroll frieze on the tie beams and octagonal piers contain trilobed leaves and each coil has a central element. In the scroll frieze on the piers of the inner arcade and the scroll frieze just below the drum, which are larger and accordingly more elaborate, the central device is a flower (with other alternate elements in the frieze just below the drum), but the trilobed leaves have been replaced with highly unusual leaf forms something like crabbed human hands. These cannot possibly be Umayyad.

Creswell, unusually for him, presented a series of comparanda in an effort to show that the scroll frieze of the piers of the inner arcade and the scroll frieze just below the drum (he did not include the scroll frieze of the octagonal arcade) were Syrian in style, of a type developed in the fifth and sixth centuries.22 He did not complete the argument though, concluding that “it is but one step more” from his comparanda to the friezes of the Dome of the Rock. In fact the general arrangement of all the scroll friezes in the Dome of the Rock is the same, close to the slightly more complex stucco frame of the lowest wall panels in the reception room of the bath at nearby Khirbat al-Mafjar, particularly in that that the stem of the scroll is continuous, not a series of cornucopias, and runs as almost a straight line between the coils, which is not true in the material Creswell adduced.

That similarity shows that the scroll friezes of the inner arcade and below the drum are based on an Umayyad design, similar to the scroll frieze of the octagonal arcade. But, again, the leaves shaped like crabbed hands cannot possibly be Umayyad.

Instead, these forms must be Ottoman, as they have unmistakable parallels in certain pieces of blue-and-white Iznik ceramics that have been dated to the early sixteenth century. There may well be other parallels in contemporary Ottoman art, but I know of none from other periods.23

Here is a list of the Iznik ceramics I have found that feature the distinctive leaves:24

The occurrence of the same very unusual leaf forms on both ceramics and the Dome of the Rock scroll friezes indicates that the scroll friezes on the octagonal piers and the scroll frieze just below the lower cornice of the drum must have been designed during Sulaymān the Magnificent's long campaign of restoration, which began at least as early as 935/1528–2929 and continued at least until 972/1564–65.30 The design was presumably composed by a court designer, and the similarity of the basic scroll design to that of the scroll frieze of the octagonal arcade and the example at Khirbat al-Mafjar shows that these scrolls are replacements for Umayyad originals, lightly modified by Sulaymān's court designer. I am not the first to attribute these friezes to Sulaymān, but so far as I am aware no one has done so on stylistic grounds, and previous attributions have been of the entire internal marble decoration, without focus on the scroll friezes.31

As the scroll frieze replaces an Umayyad original, the marble revetment of the spandrel zone between it and the ablaq arches below, and the arches themselves, must have been revetted in marble originally; they probably were restored under Sulaymān as well. What is visible in the American Colony photographs may be an even later restoration. But it probably preserves the original scheme, at least in outline.

The revetment of the spandrel zone contains large square frames with small squares of colored stone within them and vertical strips suggesting squares, with small disks between them. A generally similar scheme of squares in spandrels existed until the fire of 1917 in the spandrel zones of the nave arcades of the church of Hagios Demetrios in Salonika, rebuilt in the first third of the seventh century.32 There, above an arcade of alternating light and darker stone revetment, the spandrels were filled with squares containing varied rather complex geometrical designs. There need be no direct connection between Hagios Demetrios—which seems to reflect the style of Constantinople—and Jerusalem; the style of Constantinople would have been known in Jerusalem through imperial building projects.33

Carved Spolia

There are two pairs of carved stone spolia elements in the pier revetment of the Dome of the Rock.

Panels Published by Creswell

Creswell published two “panels” of carved stone: “Pier No. 2 (next the staircase to the roof) is decorated on its two outer faces with magnificent panels of carved marble (Plate 10 a and b); the left one measures 1.13 x 0.85 and the other 0.95 x 1.01 m.”34 For convenience I refer to them as panels 10A and 10B. 10A is clearly composed of five vertical sections of stone.35

These carved panels have hardly been discussed in the literature. Myriam Rosen-Ayalon wrote of 10A (she also referred to 10B in a footnote) that “this marble panel should be classified as Umayyad [n. 30, citing only Creswell's plates]; many scholars accept it as certainly such [n. 31, referring only to A. Grabar, “Les mosaiques de Germiny-des-Pres”, Cahiers Archeologiques, v. 7, 1954, p. 182] or, if post-Umayyad, only a few decades later.”36 In the article cited André Grabar included panels 10A and 10B as part of a loose grab-bag of Islamic comparanda that he never subjected to detailed analysis. I think the two panels might well be pre-Islamic, like the other spolia discussed below, or if Islamic, pre-Umayyad.

I have no good interpretation of these reliefs, but in the hope of encouraging discussion of them I shall make a few observations.

The two panels have different proportions and, overall, look rather different. It is hard to imagine them both being parts of a single larger work, and as there is no other carved stone decoration like either in the Dome of the Rock I do not think they are remnants of its original decoration. They are much more likely spolia.

Both are doubled compositions. Considering one half of the doubled arrangement, the foliate elements of the two panels and their combination are really quite similar; this is clearer if one crops the composition and looks at the central forms, ignoring the development of the foliage that fills the differently shaped mandorla and rectangular frames.

The mandorla shape in Panel 10A occurs in one of the early wood panels from the Aqṣā mosque (published by George Marçais, Robert Hamilton, and Robert Hillenbrand, all using the same numbering): Aqṣā panel 19E; and perhaps, doubled vertically, in Aqṣā panel 18E.37 However the rest of the designs of those panels are not very similar to Panel 10A.

Aqṣā panel 14E (EA.CA.5027) is closely related to the other Dome of the Rock panel, 10B, both in overall design and details.38

George Marçais wrote of Aqṣā panel 14E:

Panel E 14 exhibits a rather disconcerting composition, a superimposition of bunches, presumably Mesopotamian. But it is not at Ctesiphon that we must seek its origin ; much rather in the sculpture of the Dome of the Rock (above, [Early Muslim Architecture, first ed.] i, Plate 4 d), which is only separated by some hundred metres from the Aqṣā Mosque (Fig. 142).39

There is no Plate 4 d in the second edition of Early Muslim Architecture, and I have not checked the first edition, but I think Marçais's reference to “sculptures” indicates that he had in mind the marble panels in pl. 10, a and b. Fig. 142 is a drawing comparing the Aqsa panel to one half of Panel 10B. So Marçais saw the similarity in composition, but he had nothing much to say about its significance, nor about anything else involving the carved panels.

Hillenbrand did not mention Marçais's comparison, and wrote of 14E only “the design varies from one border to the next, and in one case (14E; FIG. 49 [sic]) is markedly richer and quirkier than the norm.”40

The Aqṣā wooden panel is clearly based either on a common source or perhaps directly on Dome of the Rock panel 10B, but differs from it in its greater complexity and pictorial detail. As for its date, the group containing it has been considered either ʿAbbāsid (as Hamilton did in Structural History) or Umayyad (as Hamilton did in the section on the Aqṣā that he contributed to the 1989 revised edition of Creswell's Short Account of Early Muslim Architecture).41 I agree with Hamilton's revised chronology, and regard it as Umayyad, of the period of the caliph al-Walīd (with Hillenbrand), although in fact, if the Aqṣā of al-Walīd was the completion of a work begun by ʿAbd al-Malik the wooden panel could be contemporary with the Dome of the Rock.42

Stylistically there is nothing like these two panels in pre-Islamic Byzantine architectural decoration. They are also not Crusader in style. In fact, the closest comparanda are the mosaics of the Dome of the Rock, for example, the left and right sides of the piers of the octagonal arcade.43

My guess is that the carved stone panels came from some local church or churches. Their current location on the outside of one of the octagonal piers hardly shows them to good advantage, so they may have been moved there from a more visible location.

Early Byzantine Spolia

Figure 8. View to East-Southeast with Pier 1, 08633u.

Figure 9. Pier 1 R, detail, 08633u.

Figure 10. View to West-Southwest with Pier 8, 06639u.

Figure 11. Pier 8 L, detail, 06639u.

Two other panels of carved colored marble flank the south entry, facing each other on Pier 8 L and Pier 1 R. So far as I know they have never been discussed in the scholarly literature on Islamic art. Because of their locations they appear in early views (especially the panel on Pier 8 L), although it is hard to be sure just what is shown in the engravings after Catherwood. The panel on Pier 8 L is of dark green marble, framed in red and black.44 I do not know what color the other panel is. While they have surely been reset, and probably more than once, and while the red and black frame of the panel on Pier 8 L may not reproduce the original frame, if any, they seem attractively and appropriately located and hence likely in their original locations.45

The two panels are distinguished by the compound curves of their inner framing elements, which have been described as “wave-crested”.46 Related carved panels, with central recessed blank areas framed by curly elements, exist in Hagia Sophia (532–37), and the “wave-crested” profile was used extensively in the opus sectile revetment of the the Euphrasian Basilica of Poreč, Croatia (ca. 540) and the “champlevé” elements of the revetment of the Episcopal Basilica at Kourion, Cyprus (ca. 430, the revetment possibly but not probably from a later redecoration).47

The relief panels in the Dome of the Rock are pre-Islamic in date, from the fifth or the sixth century. They must be spolia, likely from a church built by some member of the imperial family, such as the Church of St. Stephen, rebuilt 444–460 by the empress Eudocia, (although left unfinished at her death)48 or Justinian's Nea Ekklesia of the Theotokos, completed in 543, which may or may not have suffered during the Persian occupation of Jerusalem and may or may not have been restored to some degree afterward.

Other Revetment

The piers of the inner arcade are covered in marble up to the springing of the arches and the octagonal arcade piers up to the arcade frieze below the scroll frieze. I leave aside the arcade frieze except to say that I think it is an original element but that it has so much variety in quality and detail that it must be much repaired.49

Ernest Tatham Richmond, who described the problem of water infiltration that must have led to repeated repairs to this marble revetment,50 reported that

The priests of the Armenian Patriarchate still speak of two great works of repair carried out to the Dome of the Rock during the nineteenth century, one about 1853 and the other about 1874. Both lasted several years. The repairs of 1853 were directed (I am quoting the information given me at the Patriarchate) by an Armenian architect named Garabet. Garabet Kalfa was an expert in dome construction. He came from Constantinople. With him came many Armenian craftsmen. It was probably Garabet who strengthened the Dome in the manner already described. He is said, too, to have repaired the interior decoration. Large numbers of Armenians were also used for the repairs carried out by Sultan ʿAbd al-ʿAzīz about 1874. … According to the Sheykh Khalīl al Dunuf, one of the oldest attendants at the shrine, the existing plain marble flooring and the marble casing of the bases to the columns of the intermediate colonnade date from the same period.51

Richmond gave further details:

The panelled plinth or dado covering the lower part of the octagon [the inside of the exterior wall] dates from the 1874 repair. At the same time a good deal of other work was done to the marble decoration. The casing in marble boxes of the bases of the columns forming the intermediate octagonal colonnade has already been referred to on p. 18. According to the Sheykh Khalīl al Dunuf the marble bases to the columns carrying the drum were also altered, having previously been made up of a number of small pieces of different coloured marbles. These were removed and fixed to the four piers of the drum, where they can now be seen forming a rather unnecessary part of the decoration.52

Richmond described two extensive campaigns, but no comprehensive redesign of the pier revetment. I see no evidence that any comprehensive redesign was ever applied to the piers, rather that the restorations Richmond mentions were the last (before the American Colony photographs) of a long series of successive restorations of the revetment that, whether piecemeal or general, were conservative, repairing rather than redecorating. The lead roof covering was probably maintained only sporadically, and water seeping into the piers (as Richmond described) would have produced a continuing problem, in changing areas, of marble slabs coming away from the underlying masonry.

The presence of the carved spolia discussed above, which probably came from local churches, suggests strongly that there was other marble spolia, as one would expect to have been available in those same churches. Throughout the pier revetment there is colored marble, including large rectangles framed as points of interest. I think it most likely that they are remnants of the original design, which included much colored marble that was spolia and could not be replaced in later centuries. The quarter-sawn grey-white marble that forms so much of the present revetment was part of the original design and continued to be available; as the colored marble slabs deteriorated with age they were reworked into smaller pieces and the grey-white marble gradually came to compose ever more of the pier revetment.

A topic purely for speculation, but not without interest, is when the spolia was gathered and for what building project: if it came from churches ruined by the Persians it would have been available decades before the Dome of the Rock can have been projected, and could have been stockpiled for some other purpose.

Inner Sides of the Inner Piers

The inner sides of the inner piers show starkly the loss of colored marble.

Figure 12. Southeast Pier, Inner Side, 05029u, detail.

Figure 13. Northeast Pier, Inner Side, 06636u, detail.

Figure 14. Northeast Pier, Inner Side, 05029u, detail.

Figure 15. Northwest Pier, Inner Side, 06638u, detail.

Figure 16. Southwest Pier, Inner Side, 05029u, detail.

All sides of the inner piers share the same division into registers. At the bottom is a plinth, generally of grey-white marble, corresponding roughly to the “panelled boxes of marble” concealing the original bases of the columns.53 I believe these were all redone in 1873–75 (they receive the most casual damage and had to be reworked whenever the pavement was raised or renewed), and I shall not bother describing them in detail. The next zone up, which may be called a base, is shorter than the plinth, though taller than the short frieze two registers above it. The bases are made of several varieties of colored marble, composed variously (as throughout this discussion, see Nuseibeh for color photographs). Above the base is a lower large field, then a short frieze, and an upper large field, capped by the scroll frieze. The lower large fields are composed of vertical panels of grey-white marble separated by colored strips, except for the northwest pier, the only one to have a large rectangular panel of colored marble. The short friezes are of three designs, discussed in the section titled “Short Friezes”. The upper large fields are all composed of four panels of grey-white marble. All the grey-white marble panels on these sides are either half- or quarter-sawn, and the panels are usually separated by raised fillets, some of which appear dark in the American Colony photographs (but may not be).

Some of the nonphotographic evidence introduced in the section titled “Visual Evidence” can be evaluated in connection with the inner sides of the inner arcade piers.

The section engraved after Arundale's drawing shows the two western piers decorated similarly to the northwest pier in the American Colony photographs, but more regularly, and the short frieze is indicated only by a horizontal line crossing the panels of the upper large field. Surprisingly, the top of the iron grille (including its spikes), which is depicted as of uniform height, is shown striking the sides of the piers a bit above the top of the lower large fields, whereas in the American Colony photographs this level is a good deal lower—and the grille is of two different heights on either side of the northwest pier. In the photographs the lower large field is a bit taller than the short frieze and the upper large field combined, but in the engraved section both lower and upper large fields are the same height. The engraving is simply not credible in these details.

The section published by de Vogüé depicts the building cut along the east-west axis and viewed to the north, showing the two northern piers. Like the engraving after Arundale it shows the iron grille as of uniform height on either side of the northwest pier, although the taller section is shown striking the pier at the height shown in the photographs. All the registers are shown, and are of about the correct height, but both piers are shown with a decoration loosely like that of the northwest pier as it appears in photographs but rather different in detail. If this was the state of the inner sides of the inner piers before the campaign of 1873–75 the northeast pier lost a considerable amount of detail then, but it seems more likely that de Vogüé simply was inaccurate, too.

The second lithograph after Paris appears to capture the difference in the height of the iron grille on either side of the northwest pier, although it is misunderstood as passing behind the piers; the visible part of the northeast pier's inner side matches the photographs generally, but the colors of the short frieze are reversed, which is very unlikely to have been done in 1873–75.

Haag's The Holy Rock manages to omit both the iron grille and the wooden balustrade within it, but the inner sides of the two southern piers are shown more or less as seen in the photographs, with the upper large fields too short.

Simpson's watercolor is a view to the southeast showing part of the southwest pier and all of the inner side of the northeast pier above the wooden balustrade (the southeast pier is obscured by the canopy suspended over the Rock). A gilded piece of what seems like woodwork with a trilobed arch projects above the wooden balustrade; it seems to be an interpretation of the stone-and-tile mihrab near the northwest pier but in a different relation to it than shown in the watercolor. The southwest pier's inner side seems sketchily observed, but it is shown as something like the northeast pier in the same watercolor. The inner side of the northeast pier has a decoration loosely like that of the northwest pier as it appears in photographs but rather different in detail, as with de Vogüé's section. It is a bit different from de Vogüés' rendering: there are five fields, two very short, within the lower large field's central panel. The central panel is narrower than shown in de Vogüé's section and much narrower than in photographs. Nor does this design match that of the same pier in the lithograph after Paris, which is, however, somewhat indistinct.

The nonphotographic evidence cannot be dismissed entirely, but it must be heavily discounted. It is also very possible that some of these depictions of the interior of the Dome of the Rock were based on others, if only for details. If they document a state of the revetment of the inner sides of the inner piers that was changed in 1873–75 that state would seem to be that the two northern piers bore the same design, which was possibly (Simpson, not de Vogüé) more like that of some of the sides of the piers, such as the right side of the northwest pier, where the framed field is divided into bands horizontally.

As far as the state shown in the photographs, not even the northwest pier has a plausibly original design (although cf. the rightmost framed panel on the qiblah wall of the Great Mosque of Damascus, discussed below). All of the inner sides show the result of making the best of a diminishing supply of colored marble, reused by being cut into strips and distributed more widely amidst ever more grey-white marble. The original design seems impossible to recover, except that the division into registers is very likely to have survived successive remodelling, if only because at any one time many of the pieces of marble would have fitted that division.

Outer Sides of the Inner Piers

The American Colony photographs show the entire outer side of only one of the piers of the inner arcade, the southwest pier. None of the nonphotographic views is useful except for the lithograph after Paris from the octagonal arcade and the watercolor by Hunt, both of which show the southeast pier. For the rest I have referred to Nuseibeh with some slivers of views from the American Colony photographs.

Figure 17. Southeast Pier, Outer Side, 00894u, detail.

Figure 18. Southeast Pier, Outer Side, 05041u, detail.

Figure 19. Northeast Pier, Outer Side, 03229u, detail.

Figure 20. Southwest Pier, Outer Side, 08633u, detail.

The base of the southeast pier's outer side is composed of rectangular tiles of colored marble, I believe in two alternating colors, framed.54 Filling the lower large field are three fields of quarter-sawn grey-white marble separated by vertical strips of colored marble, with narrower strips of the same material at the edges. The short frieze contains one of the designs to be discussed below. The upper large field is made up of five fields of quarter-sawn marble, with a square piece of colored marble, framed, in the center of the middle field.

The lithograph after Paris from the octagonal arcade shows this side behind a column that occludes almost all of it, but what is shown matches all other evidence only loosely. The Hunt watercolor shows the plinth edged in red marble, as it is not in the American Colony photographs, perhaps evidence that it was redone in 1873–75. It shows the base as the American Colony and Nuseibeh's photographs do, and also the lower large field, in which the vertical strips are red and green. The short frieze and the upper large field are as shown in photographs.

The same scheme appears on the other three outer sides, with variations. On the northeast pier55 the base and the vertical strips are all of blackish marble. On the northwest pier56 the framed piece in the upper large field is larger, and the upper large field has vertical strips of colored marble. In the center of the base of the southwest pier was one of the panels of “small pieces of different coloured marbles” that Richmond (quoted above) reported were taken from the marble bases of the columns carrying the drum in 1873–75.57 The framed piece in the upper large field is large.

Left and Right Sides of the Inner Piers

The sides of the inner piers are by far the most interesting, not least because in the American Colony photographs, as in Creswell's, the iron grille is still in place.

Figure 21. Southeast Pier, Right Side, 03227u, detail.

It is useful to describe the eight left and right sides of the inner piers briefly before discussing the decoration they have in common.

Southeast Pier, right side (Figure 21).

The right side of the southeast pier shows the prevailing decorative scheme of the left and right sides. Above the plinth and base, which appears to be a horizontal, framed rectangle, the lower large field is filled by a large panel of colored marble (in some cases divided into several pieces), with a single or double carved frame (here it is double). The short frieze continues the pattern of the inner and outer sides and the upper large field contains a roundel in relief within a square (except for two pier sides), which is framed (as here) with colored marble bordered by billet moldings (again, except for two sides on which the square is unframed).

Southeast Pier, left side.

The roundel is placed within a square but the square is unframed, and in the center of the singly-framed rectangle in the lower large field there is a gilded disk within a disk of white marble within a gilded square. See Nuseibeh p. 64 (in background). A section drawn and painted by William Harvey in 1909 and reproduced as the frontispiece to Early Muslim Architecture, 2nd ed. v. 1, pt. 1, shows this side.58

Northeast Pier, right side.

The lower large field's panel has a single frame. There is no roundel; the upper large field is divided into three fields of quarter-sawn grey-white marble. This side can be seen in the background of Nuseibeh, p. 111. The section by de Vogüé shows it only schematically, as does the second lithograph after Paris.

Northeast Pier, left side (Figure 22).

The base is composed of a horizontal rectangle of black marble, framed in grey marble, within a frame of grey-white marble; at each side are short vertical strips of grey and black marble. The lower large field's panel has a single frame, enclosing a half-sawn panel of strikingly veined grey-white marble with horizontal bands of black marble at top and bottom. The roundel is fully framed. Well illustrated in Nuseibeh, pp. 64 and 67. Simpson's watercolor shows about a third of this side, and generally matches the photographs, though the proportions of elements within the registers are off.

Northwest Pier, right side.

Like the left side of the northeast pier, which it faces, except that the base is composed of a small rectangle of grey marble, framed in white, then in black, with short strips of white and black marble; also, the enframed grey-white marble in the lower large field is not so strikingly veined. See Nuseibeh, pp. 65 and 115 (very obliquely).

Northwest Pier, left side (Figure 23 and Figure 24).

The left side of the northwest pier is clearly shown in Creswell, op. cit., pl. 32 b. The frame in the lower large field is single and there is no roundel in the upper large field. The section by de Vogüé shows it only schematically.

Southwest Pier, right side (Figure 25).

The lower large field has a single frame and the upper large field has a roundel within a square that is otherwise unframed. Also shown in Creswell, op. cit., pl. 32 b; see also EA.CA.53. The section by Harvey shows that the lower large field is filled in the same way as that of the left side of the southeast pier.

Southwest Pier, left side.

Visible in Creswell, op. cit., pl. 32 a; well illustrated by Nuseibeh, p. 61. The lower large field has a double frame and the upper large field has a framed roundel.

Figure 22. Northeast Pier, Left Side, 05029u, detail.

Figure 23. Northwest Pier, Left Side, 11784u, detail.

Figure 24. Northwest Pier, Left Side, 05026u, detail.

Figure 25. Southwest Pier, Right Side, 08633u, detail.

The Iron Grille and the Pier Revetment

Figure 26. Iron Grille between Northwest and Southwest Piers, 05445u, detail.

Figure 27. Northwest Pier, Left Side, 05026u, detail.

Figure 28. Northeast Pier, Right Side, 14752u, detail.

Figure 29. Southwest Pier, Right Side, 08633u, detail.

The lower large fields of the sides of the inner piers exhibit prominently large rectangles of colored marble, emphasized by their carved and gilded frames. The iron grille crashes into all of them, with no regard for the compositions of which they are parts. (It smashes into the columns as well, but this is less disturbing to me, somehow.)

The grille was apparently added during the First Crusade, and there is no evidence that its location was changed before it was removed in the twentieth century. It seems to me that had the revetment been redesigned after the grille was installed its designer would surely have taken the existence of the grille into account.

I conclude from this design collision that the layout of the revetment, at least of the lower large fields on the left and right sides, predates 1099, when the Crusaders took Jerusalem. The Crusaders simply disregarded it (I can imagine the man in charge just ordering a grille so many feet high), and no change was made in the layout in later restorations. It may have been difficult to work around the grille in restoring the left and right sides of the piers, but that is what appears to have happened.

If the lower large fields retain their pre-1099 design, then it is likely that such features as the division into registers and their heights, the frames of the lower large fields, and the roundels are similarly old (although the frames and roundels were restored later). I think that in fact the original design had the same division into registers, and such frames and roundels.

Carved Frames

Figure 30. Northwest Pier, Left Side, 05445u, detail.

Figure 31. Southeast Pier, Right Side, 03227u, detail.

On all eight left and right sides of the inner piers the lower large field includes a carved and gilded frame surrounding a rectangular field and set in a surround of grey-white marble. These frames are of two designs.

In the single frames high-handled vessels of kantharos form alternate with smaller large-bellied urns that sprout bunches of foliage, both enclosed in long curved leaves, imbricated on the inside; the corners are occupied by roundels (Figure 30). The side sections grow from bottom to top and the top and bottom sections grow from the roundels toward the center, where they meet without any central device. This is the pattern of the frames of all the left and right sides except the right side of the southeast pier and the left side of the southwest pier, which face each other across the southern approach (although the pattern of the left side of the southeast pier is unclear to me).

On those two sides the frame is double, both inner and outer frame being composed of a rather rigid foliate interlace alternating pointed ovals and large circles, with small circles interposed, all elements filled with simple foliage. The design is laid out so that there are large circles in all the corners. The side sections grow from bottom to top. Nuseibeh p. 61 (the left side of the southwest pier) appears to show that the top and bottom sections grow from the corners toward the center, where they meet in a large circle, except that in this case that large circle is offset to the left in the outer frame.

The southern approach would have been that taken by visitors coming directly from the Aqṣā mosque, which is marked by other deliberate variations in design according to Sheila Blair.59 It is worth pointing out that the correspondence of these double frames would not have been so apparent after the installation of the iron grille.

These frame designs are not Umayyad.60 Nor are they even convincingly Islamic. But they could be European, and Sulaymān had Italian artists in his employ. As sixteenth-century Italian designs they are not outstanding, but they could well be of that period, and I attribute them to Sulaymān's restoration—again, a restoration of a design that was already in place.61 They all seem to be the same size, so where they are still filled with full rectangular pieces of marble those pieces could be original (compare the rectangle of marble on Pier 4 L, Nuseibeh, p. 67, which seems to match in size).

No Umayyad stone parallels for the use of such carved frames occur to me, but there could have been Byzantine examples in Jerusalem, as in the revetment of the piers of Hagia Sophia the sixth register is composed of rectangles of red and green marble with billet borders and a continuous carved frame around the whole set.62 A double carved frame appears in the mihrab of the Great Mosque of Cordoba, surrounding the large inscription band, perhaps an inheritance from Umayyad Palestine.

Billet Moldings

The billet moldings around the roundels occur in Hagia Sophia but they do not seem to occur in Ummayad architecture.

Billet moldings in exterior masonry appear in Mamluk architecture from as early as 801-13/1399-1411 (the Jāmiʿ al-Uṭrūsh in Aleppo)63 and in Mamluk Jerusalem from before 820/1417-18 (the minaret of the Khānaqāh al-Ṣalāhīyah),64 and into the Ottoman period.65


Roundels are missing from only two left and right sides: the left side of the northwest pier and the right side of the northeast pier. These two sides do not face each other but can be seen from the same viewpoint, looking north. Similarly, looking to the south, on the left side of the southeast pier and the right side of the southwest pier the squares within which the roundels are placed lack frames; these are the only such sides. These variations may have been some original attempt to create axiality. But perhaps it is more likely that all left and right sides originally had similarly framed roundels, and that some roundels and frames were damaged or lost and not replaced. It is possible that one or two of the surviving roundels have been moved, and the frames reworked, to achieve the present effect.

In their composition (which I will not attempt to describe; see Nuseibeh p. 70 for a good detail) they match well enough at least one masonry roundel in the Jerusalem city wall dateable to Sulaymān's time that they can be assimilated to his restoration of the Dome of the Rock,66 although the composition could well be earlier.

These roundels can be paralleled in Umayyad architecture, in the revetment of the qiblah wall of the Great Mosque of Madīnah, completed in 90/709.67 A description of this decoration was preserved by Abū ʿUmar Aḥmad ibn ʿAbd Rabbihi (d. 328/939–40), perhaps from a personal visit in 317/928–29:68

The southern wall of the Mosque has a marble wainscotting [dado] on the inside from its base up to a man's height. Joined to this is a convoluted ornamental cordon of marble of the thickness of a finger. Above this is another wainscotting narrower [shorter] than the first, painted with a bright red unguent called khalūq; then another like the first one, and containing fourteen gilt and decorated windows running in a line from east to west, of the size of the windows in the Great Mosque of Cordoba. Above this is another wainscotting also of marble, containing an azure space (?) on which are inscribed five lines in gold, in thick letters as big as a finger, out of the chapters of the Qur'ān called Qiṣār-al-Mufaṣṣal. Above this is another wainscotting like the first, i.e. the lowest, in which there are shields of gold, with decorations [in relief] (tirasah min dhahab manq[ū]shah).69 Between each pair is a column from which branch out branches of gold. Above this is another narrow [short] wainscotting of marble, with ornaments. In breadth [height] it is of about the length of a forearm. It has branches and leaves of gold in bold relief: in the middle is a square (or rectangular) mirror said once to have been the property of ʿĀesha.

Jean Sauvaget drew a reconstruction of the wall based on this description, which while reasonable contains much that is not provided by this passage.70

The register with the shields was the height of a man, so they should have been reasonably large. The use of the word tirasah to describe them confirms that conclusion and indicates that they were probably executed in relief. They could have been like the roundels of the left and right sides of the inner piers of the Dome of the Rock, and their existence at Madīnah makes it very likely that roundels were part of the original design of the Dome of the Rock's revetment.71

Short Friezes

The short friezes of the two north piers are similar and differ from those of the two south piers, which share another design. I omit a survey of the nonphotographic evidence for the short friezes, as it should be clear by now that much of this evidence is schematic in detail.

On the sides of the north piers the short friezes are composed of three widely spaced squares of colored marble, the outer two set at the corners in such a way that they are centered in rectangles a little higher than they are wide, with the rectangles divided from the central zone by vertical strips of colored marble (see Figure 13). On the outer sides the center square (an irregular hexagonal shape on the northeast pier's outer side) is similarly flanked by short strips of colored marble, which are omitted on the other sides. On the inner side of the northwest pier the center square (or rectangle) is the full height of the register.

Figure 32. Southwest Pier, Outer Side, 08633u, detail.

Figure 33. Southeast Pier, Outer Side, 00894u, detail.

The design of the short friezes of the southern piers may be unique. Small squares of colored marble are spaced nearly equally across the frieze, bordered on top and bottom by horizontal strips of colored marble that stop short of the corners, giving the effect of a full-height rectangle of grey-white marble around the outer squares. The squares are not all squares and not all quite the same size, showing that they are reused bits of marble, as the horizontal strips must be, too.

This design, which suggests framing but avoids framing, would have been avant-garde in Vienna of 1890, but it seems to date back to before the 1873–75 restoration, as it appears in the Hunt watercolor of 1869, which shows the exterior of the southeast pier.

The 1863 watercolor by Carl Werner titled “Interior of the Dome of the Rock, Jerusalem” shows the right side (and, almost indistinctly, the outer side)of the same pier, without the design of small squares. Instead there is a wide and short frame of green marble, set toward the top of the frieze and framing grey-white marble (which is not the usual enframed material). This side is also shown with a framed roundel (which it does not have in photographs) and a double frame in the lower large field (it is single in the photographs). I suspect Werner's watercolor does not show a previous state of this pier but rather a composite of the decoration of various piers. The horizontal frame may have existed on one of the north piers in 1863. In any case, it is the key to understanding the design of the short friezes of the south piers: assuming that it actually filled the full height of the short frieze of whatever pier it occupied, it could have been transformed into the design of small squares by eliminating the short vertical sides, leaving the top and bottom sides of the frames as the horizontal strips seen in the photographs, filling the ends of the frieze with grey-white marble, and inserting the small squares.

There is no way to know how far back the idea of a wide and short frame goes in the revetment of the Dome of the Rock, but it also appears in some of the bases and on the inner side of the northwest pier.

Piers of the Octagonal Arcade

The piers of the octagonal arcade retain less marble that might date to a time before the nineteenth century restorations than do the inner piers. The same is true of their designs. I shall describe them as a group rather than in detail.

Figure 34. Pier 1, 03228u, detail.

Figure 35. Pier 2, 03227u, detail.

Figure 36. Pier 3, 03229u, detail.

Figure 37. Pier 8, 08633u, detail.

Broadly, there are two schemes of division into registers in the revetment of these piers, which are shorter than the inner piers. Adapting the terminology I used for the inner piers, one scheme is composed of a plinth, base, large field (taller than the large lower fields of the inner piers), and a short frieze of varied height. This scheme was used for Piers 1, 3, 7, and 8. Pier 2 is a variant, without base and with a taller short frieze. The other scheme is that of the inner piers, with some change in proportions, used for Piers 4, 5, and 6.

These variations suggest that whatever the original scheme of decoration was, it was changed differently in different campaigns of restoration. As Clermont-Ganneau saw Pier 7 stripped of revetment in 1874, which means it was subsequently rerevetted, I am inclined to associate the first scheme generally with that campaign, but to interpret this material further is probably to overinterpret it. It is worth noting that all of the designs composed of small pieces of marble of various colors occur on piers employing this first scheme.

(The registers of Piers 1, 2, 3, and 8 are shown in the figures above. For Pier 7 see Creswell, op. cit., pl. 5 a. Nuseibeh illustrations: Pier 1, pp. 57, 112, and 116; Pier 3, p. 111; Pier 4, pp. 67 and 111; Pier 5, p. 115; Pier 6, pp. 62, 110, and 114; Pier 7, pp. 60–61; Pier 8, pp. 58 and 60.)

Other than compositions of small pieces of colored marble, the decoration of the octagonal piers is composed of vertical strips of colored marble, rectangles of colored marble (Piers 2, 3, 4, and 6, in addition to the pre-Islamic Byzantine carved spolia on Piers 1 and 8 and the probably pre-Islamic carved spolia on Pier 2), frames of colored marble around quarter-sawn grey-white marble (Piers 6 and 7), and, of course, the grey-white marble itself.

The Original Design and Its Sources

The piers of the octagonal arcade seem much more affected by their many restorations than the inner piers, but it appears that their original scheme was a reduction of that of the inner piers.

The inner pier revetment was divided into registers, each of which probably bore a different kind of decoration, as seen most clearly in the large fields of their sides: framed panels of colored marble spolia in the lower fields and gilded relief roundels in the upper fields. While no carved frames exist on other sides of either set of piers, the idea of centrally placed spolia slabs survives.

The roundels are not uniquely appropriate to their location on the left and right sides of the inner piers, as the upper large fields are rectangles, not squares. The outer sides of the inner piers might have had roundels as well, although in both arcades it is the inner side that is more heavily decorated. Perhaps something less striking was used in these locations, such as framed panels of colored marble. On the inner sides of the inner piers something more striking might have been displayed, such as opus sectile spolia or newly executed panels in the low-relief “champlevé” technique.72

The interior decoration of the Dome of the Rock preserves the distribution of materials characteristic of the early Byzantine period: marble below, gold-ground mosaic above. It also preserves the taste for exotic materials: mother of pearl in the mosaics and marble, both colored and strikingly veined, in the revetment. And the use of spolia is no innovation: Hagia Sophia's revetment already contains spolia.

Figure 38. Southwest Pier, Outer Side, 08633u, detail.

Figure 39. Hagia Sophia, Istanbul, South Pier on West Side.

In what can be recovered of the arrangement of the Dome of the Rock's revetment it follows early Byzantine practice. The alternation of short and tall registers is seen most notably at Hagia Sophia (where the height of the inner piers to the springing of the arches is roughly twice that of the Dome of the Rock's piers and marble continues above, where mosaic is used in the Dome of the Rock). Even closer may be the organization of registers in the revetment of the piers around the central space of San Vitale in Ravenna (completed 546–48).73 And in Hagia Sophia, in addition to registers of full-height colored marble there are also panels of colored marble framed in carved marble and set in fields of less colorful marble, in the upper part of the first story in the apse and in the second story under the barrel vault before the apse,74 and particularly (though more simply framed) in the inner narthex.75 These or something similar may be the prototypes of the large rectangles of colored marble framed in carved marble on the left and right sides of the inner piers of the Dome of the Rock. Ibn ʿAbd Rabbihi's description of the roundels in the Great Mosque of Madīnah suggests that the original roundels of the Dome of the Rock would have been carved in relief also. There do not seem to be any early Byzantine roundels carved in relief of this scale, but in their disposition in the overall composition the Dome of the Rock roundels might have been conceived as the equivalents of the porphyry disks so prominently displayed at Porec and in Ravenna (St. Vitale, the Orthodox Baptistry).

In short, the marble revetment of the piers of the Dome of the Rock was modelled closely on early Byzantine imperial architecture, which was represented in Jerusalem by such buildings as Justinian's Nea Ekklesia of the Theotokos, whence much of the spolia used in the Dome of the Rock may have come. The only clear differences were the greater use of spolia, a necessity in the late seventh century, and perhaps a less restrained approach to decoration.

Appendix A. The Directionality of the Scroll Frieze of the Octagonal Arcade

The scroll frieze of the octagonal arcade is composed of the eight wooden tie beams and the wooden or stone sections of the entablature atop the eight piers. Each of those sections has four faces: two on the inner side of the pier, one on the left side, and one on the right side.76

In the center coil of the frieze on each tie beam is a shell that, remarkably, is not a point of origin for two scrolls running to either side, but simply replaces the filling of its coil in a scroll that runs the length of the beam in the same direction.

Noticing this unexpected design prompted me to check the directionality of the scroll friezes on all the tie beams (Creswell published only the four on the cardinal points) and on the sections atop the piers. I began with the American Colony photographs and continued with Creswell's published plates and photographs from the Creswell Archive at the Ashmolean Museum, and made some use of Nuseibeh.

A Warning to Researchers

I tried to use Nuseibeh's plates of the mosaics of the inside of the octagonal arcade, pp. 84–105, to help identify the location of some piers shown in Creswell Archive photographs and found, insofar as I checked, that the compass points given are 180° off, so that “Inner Octagonal Arcade: North” actually shows the south side (between Piers 1 and 8). I confirmed this discrepancy by examining Creswell's plate captions and the inscription as given by Max van Berchem in Matériaux, v. 2, pp. 230–31. No location given by anyone who may have used Nuseibeh as a source can be trusted unless this unfortunate error is recognized.

Below I give the results of my investigation with the photographs I used. Here “Li” means the left half of the inner side of the pier and ”Ri” the right half; “R” is the segment on the right side of the pier connecting Ri with a tie beam and “L” the left segment connecting Li with another tie beam.

The scrolls of segments Ri and Li always run in opposite directions, so far as I can see, so it is not possible to join them smoothly, but some joints are particularly poorly made, by which I mean that coils are truncated. It is more difficult to see the joints between the R and L segments and the tie beams, but I have noticed the same problem there, too; perhaps it is partly the result of bad restoration. At the corners between L and Li, and R and Ri, a leaf form masks the joint. The three figures below show this leaf; Figure 40 shows a poorly made joint between Ri and Li; the same joint in Figure 41 is a bit better, though the coils are still out of phase at the joint. EA.CA.186 shows clearly the fairly good joint at this point on Pier 6 and the symmetry of reflection between Ri and Li.

Figure 40. Pier 5, Inner Side, Detail of Scroll Frieze, 06635u, detail.

Figure 41. Pier 4, Inner Side, Detail of Scroll Frieze, 06635u, detail.

Figure 42. Pier 8, Right Side, Detail of Scroll Frieze, 05894u, detail.

The pattern on the piers is regular: on the right side (R) segment the scroll runs left to right and the direction reverses at every joint through Ri, Li, and L. From the leaves at the joints between R and Ri, and L and Li, the scrolls run toward the center joint and toward the joints with the tie beams. On the tie beams the scrolls run left to right except on the east and west beams (between Piers 2 and 3 and between Piers 6 and 7), where they run right to left.

Given the pattern of directional reversal in the segments atop the piers, it is odd that the scrolls of the tie beams run one-way only. It would be more natural to use the shells in the middle of each tie beam as points of origin for scrolls running in opposite directions. Jerusalem woodworking practice had been to use a central circular element such as a wreath as the starting point for scrolls growing in opposite directions (as in the tie beam soffits illustrated by Hamilton, Structural History, pl. 47; see also pp. 89–90). The high quality of the carving suggests no loss of local craft tradition on the part of the woodworkers, so the responsibility for the one-way tie beam design should lie with the designer. For a possible explanation of how the reversal of direction on the east and west tie beams could have occurred see Footnote 20.

Appendix B. The Marble Revetment of the Great Mosque of Damascus

With the foregoing discussion of the revetment of the Dome of the Rock in mind it is worth looking at the Great Mosque of Damascus (built by the Umayyad caliph al-Walīd, 88–96/707–715), to see what can be recovered of the design of its original revetment.

Figure 43. Qiblah Wall before 1893, Félix Bonfils, after von Oppenheim.


Creswell pointed out the existence of two small flat pilasters in the eastern vestibule, 4.88 m. above the (modern, not original) pavement, which he thought were in situ because they were associated with quarter-sawn marble. He also published a photograph of the western vestibule taken in 1920, which shows a set of such pilasters and other revetment.78

Figure 44. Flat Pilaster Capital, Hagia Sophia.

Flat pilaster capitals carved in low relief in the east porch of the Dome of the Rock resemble these and also resemble pilaster capitals at Hagia Sophia; there is also a series of small flat carved pilasters in the apse of San Vitale in Ravenna.79 These examples indicate that the pilasters in the Great Mosque of Damascus are Umayyad and not somehow later additions.

As the Damascus pilasters matched in height and vertical position two marble grilles 1.77 high in the western vestibule (which I do not recognize in the 1920 photograph), Creswell imagined a combination of the two, illustrating it with a restoration drawing of the wall of the western riwāq that incorporates two grilles there, which were 4.90 m. above the (modern) pavement.80

Creswell described his reconstruction:

Thus there must have been, first a surface of marble panelling 4.88 m. in height, of which part (and perhaps all) consisted of quartered rectangles, divided by small baguettes. Then came a band 1.77 m. wide which extended upwards from the 4.88 m. level to 6.65 m., also of quartered marble but with pilasters and marble grilles at intervals. … Directly above it, that is to say directly above the tops of the marble grilles, began the coating of mosaic (fusaifisā') as may still be seen under the west riwāq (Plate 54 b). This extended up to the roof level which is about 15 m. from the ground (or 14.65 m. from the raised pavement of the riwāq), so it must have been over 7 m. broad [tall]. This scheme doubtless existed under the northern and eastern riwāq[s] also.

The grilles were of course window grilles, although light shows through only one of them in Creswell's plates and the other three appear blocked.

Creswell continued:

As regards the decoration of the sanctuary, we are told that marble panelling ran all round to several times the height of a man; at this point ran the famous golden vine (karma) in which we must recognize a frieze decorated with a long undulating vine-stem gilt, or an acanthus scroll carrying grapes, after the pattern of some of those in the Dome of the Rock. This is confirmed by a photograph of Bonfils (Plate 62A (d)), taken before the fire [of 1893] and published by Baron von Oppenheim, which shows that a band with pilasters similar to those in the eastern vestibule ran along the southern end of the transept. But it shows something more, for immediately above is a band consisting not of a true vine stalk but of an acanthus scroll carrying bunches of grapes here and there, as in some of the fantastic plant ornament in the Dome of the Rock; it can only be the golden ‘karma’ mentioned by Ibn ʿAsākir. Its width, scaled off Bonfils photograph, is 60 cm. (Note 2: The pilasters measure 35 mm. in the photograph and the karma 8½. Taking the former as 1.77 m. [note the assumption] the width [height] of the latter must have been 60 cm.81) As the windows of the sanctuary are 10.33 m. above the floor, they must have cut into the mosaic instead of coming between the pilasters.

Above this frieze, up to the ceiling, the walls were entirely covered with polychrome mosaics … .82

Finbarr Barry Flood cited this passage, with footnote, noted the assumption of the height of the pilasters, and remarked, “At approximately one quarter the width [height] of the pilasters, however, the true figure for the width [height] of the karma should be about 45 centimeters.”83 This reasoning seems to be based on reading what Creswell wrote. In fact, on the published plate and measured with its plain borders the karmah is about 8½ mm. tall at the right side of the second pilaster from the right (because it is an oblique view things of the same size appear progressively larger toward the left) but that pilaster is 25 mm. tall (that on the left is 28 mm.), which makes the karmah 60 cm. high. So Creswell scaled correctly but failed to catch a typographical error.

Photographs of the eastern half of the qiblah wall published by Badr el-Hage show no pilaster register, which may be why Creswell limited his comments regarding the pilaster register to the area of the “transept”.84 Nor do they show a continuation of the karmah, at least at the same level; the band that Flood thought was such a continuation85 is far too low, as can be seen in el-Hage's p. 96, a photograph by Trancrède Dumas et fils, in which both the western portion of that band and the karmah over the mihrab are visible. Flood referred to a “faint narrow band … at the appropriate height on one of the piers to the left of the main entrance as it appears on [Richard Phené] Spiers' water-colour of the mosque's interior” (Flood's color pl. 2, also available online at the Victoria and Albert Museum web site). I am not sure whether he meant the freestanding pier that would be to one's left as one entered or the large pilaster applied to the north wall, supporting the western nave arcade, but both seem to be at the right level. However, by comparing all the four piers visible in the watercolor it can be seen that the “band” is a raked cornice, gilded on the other piers and probably decorated with acanthus, like the corresponding cornice seen in the view of the center of the qiblah wall by Sulaymān al-Hakīm (to be discussed immediately). The cornices are also shown clearly in Flood's fig. 22 and 35, photographs taken soon after the fire. As for the “fragmentary nature of the karma as it survived in the last century” (p. 61), a comparison of the painting by Frederic Leighton (Flood's color pl. 1) with photographs shows it to be unreliable in details. A photograph by Bonfils, which seems to be the one Flood published part of as his fig. 30, happens to show part of Flood's band on the eastern wall in good light (at left), and its decoration looks far more like acanthus than a vine scroll.86

Figure 45. Bonfils no. 1950, “Damas—Colonnade dans l'intérieur de la grande mosquée”.

Figure 46. Bonfils no. 1950, “Damas—Colonnade dans l'intérieur de la grande mosquée”, detail.

Creswell's assumption that the qiblah wall pilasters were the same height as those he measured in the eastern vestibule can be checked roughly, using a much better prefire photograph of the qiblah wall taken by the Damscene photographer Sulaymān al-Hakīm.87 This photograph, unlike the Bonfils one, shows the wall down to the ground and establishes clearly that the karmah was at the same level, and as tall as, the cornices of the large pilasters attached to the qiblah wall that support the nave arcades. The sections that Creswell published as his fig. 86 and 87 are after Phené Spiers, who wrote that “the four great piers of the transept and the south wall were encased with marbles enriched with arabesque inlays and borders of mosaic up to a height of 20 feet.”88 The sections show the height of the piers up to the bottoms of the cornices as about 22 feet, or 6.7 m. (both are very small and of unknown accuracy). It seems reasonable to use 6.65 m., the height of Creswell's reconstructed revetment.89 The photograph by Sulaymān al-Hakīm is perspective-corrected, so the higher up the wall it is the taller an element appears. Assuming a total of 6.65 m. from the top of the carpets on the pavement to the top of the pilasters, I obtain a result of 1.7 m. for the height of the pilasters without correcting for distortion.90 That result, about four percent off, confirms the likelihood that Creswell's assumption was correct, although the height of the pilasters may have varied from place to place in the mosque.

Perhaps, as I argued for the inner piers of the Dome of the Rock, the marble revetment of the Great Mosque of Damascus retains its original division into registers.

Creswell's pl. 47 a shows the eastern vestibule. There there is a short register above the pilaster register composed of narrow rectangles of various kinds of stone. I calculate its height as 53 or 54 cm., which matches the height of the karmah well enough to be encouraging.91 The short register corresponds to the karmah in height and location. And another check can be made using pl. 47 a. Using the same procedure as I used on the photograph of the qiblah wall, I find the height of the pilaster to be 1.7 to 1.8 m.92

The revetment of the eastern vestibule seen in pl. 47 a below the pilaster register is rather varied and does not seem to be the original composition. It is unclear what the register below should be, except that over the doorway the rectangle containing the two lozenges and their frame is about 1.8 m. high, probably intended to be the same height as the pilaster register, given that joints in the revetment throw off these calculations some. The lozenges are about 1.3 m. tall.93 Whether the framed lozenges are in their original location or not, this is a height given by the architecture.

The photograph Creswell published of the western vestibule in 1920 (pl. 50 b) shows that on all three sides the vestibule retained its pilaster register and a short register above that corresponding to the karmah register of the qiblah wall, composed of what I assume to be veined grey-white marble and vertical rectangles of darker stone. Scaling from this plate is quite inexact because of its small size and relatively low resolution, but comparing the height of the rightmost pilaster (on the north wall) to that of the register above it, that register is about one-third as tall, which is about right. Just to the right of the rightmost column the space between the bottom of the pilaster register and the top of the doorway is about as tall as the pilaster register, as in the eastern vestibule. But the half-lozenge visible there is not framed; rather, its top tip touches the bottom of the pilaster register. I make it very roughly 1.2 m. tall; more likely it is 1.3 m. tall, like the lozenges in the eastern vestibule.94 A register this tall runs along the west wall at this level for its entire width. It contains lozenges and roundels (perhaps disks). On the south wall over the doorway there is a lozenge and a half, perhaps taller.

On the north wall to the right (east) of the doorway some revetment is missing, and one rectangle is left, perhaps out of its original position. Once could reconstruct here a register about the same height as the pilaster register (as to the right of the dark rectangular slab) or about a third taller (as to the left of the slab).

Returning to the qiblah wall, between the piers and below the pilaster register there is a register shorter than it. I make it about 1.6 m. tall.95 Both the shorter register and the revetment between the pilasters in the pilaster register are in good repair and quite plain. They are composed of grey-white marble, apparently quarter-sawn, and there are two or three inset panels. Comparison of the al-Hakīm and Bonfils photographs with the Creswell Archive photographs of the eastern vestibule (such as EA.CA.394) suggests that even given the difference in dates of the photographs the qiblah wall had been restored far more recently (and perhaps more often) than the vestibules, and it is possible that in restoration the shorter register acquired the height shown in the photographs. It is even more likely that it had lost some decoration, and the pilaster register, too, may have lost something between the pilasters (such as the shields of the corresponding register in Madīnah).

In any event, below the shorter register on both sides of the mihrab are thin horizontal bands above and below panels surrounded in grey-white marble with vertical strips of colored marble between them, handled differently on the two sides. Below the lowest horizontal band is a low dado of vertical rectangles.

Figure 47. Damascus, Madrash al-Zāhirīyah.

Figure 48. Damascus, Madrash al-Zāhirīyah.

The framed panels, some of which seem to have been partly painted over, appear to date from the restoration by the Mamlūk governor Tankiz (727–30/1326–30).96 This restoration may well have retained elements of the earlier restoration by the Mamlūk sultan Baybars (668/1269–70),97 as the low dado is paralleled in the mausoleum of Baybars within the Madrasah al-Zāhirīyah, begun after his death in 676/1277 and completed around 1281.98 Perhaps the rightmost panel, which is a rectangle of colored marble framed in grey-white marble and colored marble (cf. the inner side of the northwest inner pier of the Dome of the Rock, Figure 15) belongs to the same campaign as the dado. And perhaps this Mamlūk work preserves something of the design of the original revetment, as the style of the workshop Baybars established to make restorations in Damascus and Jerusalem was deliberately Umayyadizing.

Among the framed panels are two that may be later tilework: the panel two to the left of the mihrab and the one part of which is visible to the right of the minbar (more in the Bonfils photograph than in al-Hakīm's); they appear similar to the panels flanking the west gate in the western vestibule.99

I think the best that can be said is that there may have been several register arrangements, differing for zones over doorways and elsewhere, and that in prominent places there was probably a register of highly decorative framed rectangles, whether low, as on the qiblah wall, or, less likely, immediately below the pilaster register, as in the eastern vestibule, or both.

The qiblah wall of al-Walīd's Great Mosque of Madīnah (88–90/707–09), as reconstructed from textual sources by Jean Sauvaget (including the account of Ibn ʿAbd Rabbihi, which I quoted above, also had a pilaster register above which was a band like the karmah in Damascus. Below the pilaster register was an inscription band, and below that three registers, the middle one shorter than the other two, which were the height of a man (perhaps 1.5 m. or so).100 This alternation in height of registers is similar to the arrangements of revetment in Hagia Sophia and the Dome of the Rock, although the scale differs. Flood, who has written extensively about the karmah in Damascus and has considered the example of Madīnah as well, placed the epigraphic frieze mentioned in descriptions of the Damascus prayer hall below the pilaster register, but it is unclear on what basis, other than to parallel the qiblah wall of the mosque of Madīnah. According to Ibn Ḥauqal, for example, the inscription was under the ceiling.101 However, the example of Madīnah perhaps shows that the register below the pilaster register in the Damascus qiblah wall is not part of the original arrangement.

Lozenge Panels

Figure 49. Lozenge Panels and Inscription, after van Berchem.

A set of rectangular panels carved with lozenges, some filled entirely with foliage, were preserved in the mosque before the fire on two of the freestanding piers supporting the dome. Creswell described their situation (in part translating from van Berchem, see Footnote 103):

A carved panel (Plate 62A (c)), now preserved in the Damascus Museum, was removed thither from the débris of the Great Mosque after the fire of 1893. Van Berchem's photograph (Plate 62A (a)), taken in the following year, shows that there were two such panels (that is to say eight in all), set vertically on either side of four practically identical inscriptions of Malik Shāh, which were fixed to the north and south faces of the two southern dome-bearing piers. The pair flanking the inscription on the north face of the south-western pier [has] been published by van Berchem. They are very much alike but not identical. Ours [the panel shown in pl. 62A c] again shows variations in detail, but the composition of all three is the same.

As these panels are applied to composite piers, of which slightly more than half is due to Malik Shāh, it follows that they cannot occupy their original position, which still remains a problem.102

Van Berchem published a photograph of a pair of such lozenge panels, with inscription. That photograph (above, Figure 49) shows part of a carved frame around the ensemble.103 The pattern of this frame is not the same as that of the two lozenge panels, but the foliate forms are similar, and I conclude that it is original. The pattern has symmetry of reflection with respect to the central vertical axis of the ensemble, so at the least the frame is not simply reused arbitrarily. I cannot tell how the pattern turns the corner, if it does, but it is entirely possible that the ensemble was reused in its original form, except of course for whatever occupied the place of the Saljuq inscription.104

The panel Creswell published has been dated by Michael Meinecke to the period of the construction of the mosque under al-Walīd.105 Meinecke gave its dimensions as 162 by 45 cm., which is the same proportion as the photographs of it show. The two lozenge panels in van Berchem's photograph are somewhat narrower, but taking the height of the panel Meinecke gave dimensions for and scaling from van Berchem's photograph, the frame shown in it is about 9 cm. high, including the joint between it and the lozenge panel, giving a total height for the full ensemble of about 1.8 m. The width of the ensemble would be about 2.4 m. Both calculations are thrown off a bit because the view of the ensemble is slightly foreshortened.

Creswell reported that “three [inscriptions] still remained in place [in] 1894 …, but are no longer there. The two which belonged to the south-west pier, however, are preserved in the Museum.”106 According to Meinecke, the lozenge panel he catalogued was the only one preserved, of marble, with traces of gilding; he added that “old photos show that new reliefs were added with a pattern similar to the older ones, identical in composition but of simpler ornamentation”. He thought its original location was in lower zone of decoration (below the mosaic), which was composed of marble revetment.

Of the “old photos” Meinecke referred to only Creswell's pl. 62A, a, is cited in his notes. It does show lozenge panels of a different design, and two such were published by Henri Stern.107 Stern obtained photographs of the two from Jean Sauvaget, whom he quoted as stating that they were in the “dépôt de la Mosquée”.108 Stern described the vegetal ornament and cited Roman and later precedents for lozenges in rectangles, including their use in stucco at Qaṣr al-Ḥayr al-Gharbī and in paint at the palace at Ruṣāfah excavated by Katherina Otto-Dorn.109

Figure 50. Qaṣr al-Ḥayr al-Gharbī Facade as Reconstructed.

A large quantity of marble panelling, presumably in relatively good condition, should have been made available by al-Walīd's demolition of the Church of St. John the Baptist, which Creswell showed stood in the west side of the ancient temple enclosure now occupied by the mosque.110 That church was built or converted from a temple by the Byzantine emperor Theodosius (379–95) according to sources cited by Creswell,111 and may well have been redecorated during the succeeding centuries. Of course an immense amount of newly quarter-sawn marble must have been fabricated for the mosque revetment, but spolia is reported to have been used also.112

Among the spolia there may have been carved panels with lozenges, such as those found in or thought to be from the Episcopal Basilica in Kourion.113 While the Kourion panels are executed in the champléve technique, similar panels must have been the prototypes for the Damascus lozenge panels. The overall composition of which the Kourion panels were a part is lost. The related opus sectile lozenge panels in the Euphrasian Basilica, Poreč, occur in the apse, oriented vertically and in a set of rectangular panels of various designs, without a frame such as the Damascus panels were set in.114

The ensemble is too wide to have fit on the north or south face of the original piers that were enlarged to support the Saljuq dome, although it could have fit on one of the sides (see Creswell's fig. 81). At about 1.8 m. high it is taller than the qiblah wall register containing the pilasters.

There is no way to know, but the lozenge ensemble could have been part of one of the registers below the pilaster register. The only possibility I see is the register of the enframed panels, which, depending on how one demarcates it, is 1.9 m. or more. Such a location would have put the ensemble at eye level. It might have been on one side of the mihrab, with a mate on the other side.

What the original central element of the lozenge ensemble was cannot be deduced from the Kourion material, and Poreč is not much help. Were the central space square I would restore a roundel or disk (cf. the western vestibule) unhesitatingly, but it is not. An ellipse is not out of the question (there are ellipses in the apparently spolia panels in Hagia Sophia and in the revetment of the upper part of the inside of the exterior wall of the Dome of the Rock, and although the latter are surely modern they might reflect an original design). Perhaps the lozenges flanked an earlier inscription, and were reused by Malik Shāh in the same sense.

Lost Carved Ornament

Five fragments of carved marble ornament were published by Henri Stern from photographs and drawings given him by Jean Sauvaget.115 Two were narrow strips in a secondary location on the west face of the north pier of the eastern vestibule. The other three were loose, “au dêpot de la mosquée” in 1945, and might be remnants of decoration that survived the fire of 1893 but was not restored (or they could have wandered in from elsewhere in the neighborhood).

The size of the narrow strips can be estimated by comparing them with the lozenge partially visible to their left. If it is the same size as the lozenges in the eastern vestibule, roughly 1.3 m. tall, they are very roughly 20 to 30 cm. wide and nearly a meter long. Their lengths are not entirely meaningful, as they seem to be parts of borders or frames, of different but closely related designs, in which foliage grows in either direction from a circle filled with a foliate element or a rosette. Each has thin matching borders on both sides, of different designs. The strip with a foliate element in its circle has a clear vertical orientation; the one with a rosette could have been intended to be seen either horizontally or vertically. Their widths are apparently slightly different, and clearly wider than the frame of the lozenge ensemble, even discounting the thin borders. What they may have framed or bordered is not suggested to me by other surviving elements of the revetment.

Guessing on the basis of the likely scale of the matting against which they are displayed, the other three fragments might be in the range of fifteen to thirty cm. wide. They all differ in style and technique of execution, and do not match either of the narrow strips. Two seem to be borders; again, of what they were borders I have no idea. The third has a slightly curved side, so it must have been set next to an arch, of which there are of course many in the mosque, from the original mihrabs to the arcades. The only obvious Late Antique parallel that occurs to me is Hagia Sophia, where the spandrel zones of the arcades are filled with vegetation.

1. I use “marble” indifferently to describe the grey-white veined marble that constitutes most of the revetment and the various colored stones included within that background, whether they are geologically marbles or not. Some pieces of colored marble are heavily veined, and there are also pieces of heavily veined marble that are not conspicuous for their color; I call them all colored marble here. I am inclined to think that the stone used for low-relief carved decoration is also marble, but I do not know whether it really is or not.

In “Notes on a Particular Technique of Architectural Decoration”, Israel Exploration Journal, v. 24, 1974, pp. 232–36, Myriam Rosen-Ayalon wrote that the low-relief friezes of the Dome of the Rock are executed in the technique called “champlevé” in the scholarly literature, in which the background is “coarsely chipped off” and filled “with color [meaning paint?] or colored paste”. The technique is a substitute for opus sectile.

I do not know whether these friezes are colored with anything but paint, and Rosen-Ayalon cited no evidence or observation for her assertion. Several fragments of the arcade frieze of the inner side of the exterior wall, in the Islamic Museum, Jerusalem, have been published on the Internet, and while they appear to be late replacements of the original decoration, their backgrounds are not roughened to receive a colored filling (;ISL;pa;Mus01;3;en as of 1 April 2014).

2. In Early Muslim Architecture, 2nd ed., 2 v., Oxford, 1969, v. 1, pt. 1.

3. An overview of the Matson Collection is given at , with a search function. I used the URL on 30 May 2012 to obtain thirteen pages of results, reviewed thumbnails, and selected black-and-white views of the interior. Some scans are mirror-reversed; I have corrected such errors in the images used in this article.

4. In Ottoman Jerusalem, The Living City: 1517–1917, ed. Sylvia Auld et al., 2 v., London, 2000, v. 1, pp. 415–24. This article seems to me fuller in the information relevant here than St. Laurent and András Riedlmayer, “Restorations of Jerusalem and the Dome of the Rock and Their Political Significance, 1537–1928”, Muqarnas, v. 10, 1993, pp. 76–84.

5. Ed. Charles Warren and Claude Regnier Conder, London, 1884, p. 246.

6. Capt. Charles William Wilson, Ordnance Survey of Jerusalem, 2 v., London, 1865, v. 2, pl. 4. In August 2013 I found a useable scan of it online at and an inferior one at the New York Public Library web site (which sells a presumably better version).

7. A photograph by Félix Bonfils, no. 1216, was taken from farther to the east, showing Pier 1 in the foreground. It shows the ceiling of the inner ambulatory repaired, as in 11793u, in places where it was in tatters in the Ordnance Survey photograph, so I take it to postdate the 1873–75 campaign. As of 3 June 2014 scans can be found at the Library of Congress, . Information at dates it between 1867 (when Bonfils established his Beirut studio) and 1899.

8. Michael Darby, The Islamic Perspective: An Aspect of British Architecture and Design in the 19th century, London, 1983, pp. 35–37, with a long quote from Catherwood that derives from W. H. Bartlett, Walks About the City and Environs of Jerusalem, 1st ed., London, 1844, pp. 161–78, taken from a letter from Catherwood in which he mentioned the camera lucida, p. 162. On the loss of Catherwood's drawings see Victor Wolfgang von Hagen, Frederick Catherwood Archt., New York, 1950, p. 37 and n. 33 and 34 on p. 150.

9. In An Essay on the Ancient Topography of Jerusalem, London, 1847, frontispiece; the vignette of the same view is in The Holy Sepulchre and the Temple at Jerusalem, London, 1865, fig. 9.

10. Essay, facing p. 104, reproduced in Darby, op. cit., fig. 21.

11. I found a list of the plates in a bookseller's online description as of 1 April 2014, and a thumbnail image of the first plate in another bookseller's online description. Images of the second plate are for sale online as of 1 April 2014; see .

12., with text apparently from the Allgemeinen Deutschen Biographie; as of 18 April 2014.

13. Shown with no information but a caption at as of 1 April 2014. It is no. SD.1192; the date follows the signature at lower right. I found a larger image of the work at .

14. Described, with thumbnail, at as of 1 April 2014. The information about the auction and the date are from,_1864 as of 1 April 2014.

15. Judith Bronkhurst, William Holman Hunt: A Catalogue Raisonné, 2 v., New Haven, 2006, v. 1, pp. 13, 17–18, 20; v. 2, pp. 75, 140–41.

16. Images for sale at as of 1 April 2014.

17. Nicholas Tromans, ed., The Lure of the East: British Orientalist Painting, New Haven, 2008, fig. 165, unfortunately cropped at the left edge.

18. Among the pages at as of 5 May 2014; see also , as of 5 May 2014, without illustration. A smaller view with the same title sold in 1996: , as of 5 May 2014.

19. Khirbat Al Mafjar: An Arabian Mansion in the Jordan Valley, Oxford, 1959, p. 262.

20. This disparity must not have been intended. It could have occurred if the stone panels to be carved were cut to shape and the outline of the design—the stem and coils—were drawn on the panels with the aid of geometric construction, with a detailed drawing of one or two individual coils provided separately. The man who drew the outline made the error, and the carvers had only to turn the detailed drawing upside down to suit it to the outline.

21. More of the frieze is visible in 00526u than is shown in Figure 7. See also 00527u and 06635u and Creswell, op. cit., fig. 361–63.

22. Op. cit., pp. 122–23.

23. Possibly Walter Denny had such a piece in mind when he wrote of “a ceramic bowl with mittenlike curled-in oak leaves” in Masterpieces from the Department of Islamic Art in the Metropolitan Museum of Art, ed. Maryam Ekhtiar et al., N.Y., 2011, p. 286; unfortunately he did not provide its accession number and I have not found it.

24. The descriptions and dates are from Nurhan Atasoy and Julian Raby, Iznik: the Pottery of Ottoman Turkey, London, 1994.

25. Op. cit., fig. 292.

26. Op. cit., fig. 295.

27. Op. cit., fig. 300, top.

28. Op. cit., fig. 307; Esin Atıl, The Age of Sultan Süleyman the Magnificent, N.Y., 1987, p. 256.

29. Inscription on the colored glass windows, Finbarr B. Flood, “The Ottoman Windows in the Dome of the Rock and the Aqsa Mosque”, Ottoman Jerusalem, v. 1, pp. 431–63, p. 433.

30. Inscriptions on two of the doors, Max van Berchem, Matériaux pour un Corpus Inscriptionum Arabicarum, pt. 2, Syrie du Sud. Jérusalem “Ḥaram,” 3 v. pag. cont., Cairo, 1925–27 (Mémoires de l'Institut Français d'Archéologie Orientale. v. 43–45), v. 2, pp. 339–40.

31. De Vogüé attributed all the marble decoration to Sulaymān, op. cit., pp. 83 and 95, where he gave his idea of the chronology of the interior decoration: originally it was all mosaic, the Crusaders replaced some of it with figural paintings, Ṣalāḥ al-Dīn expunged those paintings and installed some decoration, and at the time of Sulaymān that decoration was thought out of harmony with the rest and was replaced with what we see now (marble); but he cited no stylistic reason.

Gülru Necipoğlu discussed Sulaymān's restoration project in “The Dome of the Rock as Palimpsest”, Muqarnas, v. 25, 2008, pp. 17–105 (citing inter alia a work by Michael Meinecke that is unavailable to me). She wrote (p. 62):

I have found confirmation of that date [959/1551–52] in the unpublished chronicle of Mustafa ʿAli, which lists the renovation of the Dome of the Rock among Sultan Süleyman's major building projects. The passage suggests that the project involved more than external tile revetments, and perhaps included the renewal of marble revetments as well: “In Jerusalem the Noble, the interior and exterior of the exalted Rock of God (ṣaḫratu'llāh-i müşerrefe) was [re]built with the installation of tiles (kāşī) and its construction reached completion in 959.”

The passage quoted does not mention marble either inside or out, and Necipoğlu did not say why she suggested “renewal of marble revetments”.

Creswell, op. cit., p. 99, wrote of the inner arcade that “the arches proper, as well as their soffits, are faced with alternate slabs of black and white marble. This facing, hitherto believed to date from Sultan Sulaymān (1520–66), is older,” but did not specify who believed it to date to Sulaymān's work.

Oleg Grabar, The Dome of the Rock, Cambridge, Mass., 2006, p. 200, wrote that under Sulaymān, “inside the building, much was repaired and redone, in particular the sequence of marble panels in the lower part of the wall”, without citing evidence.

32. Richard Krautheimer, Early Christian and Byzantine Architecture, 4th ed. 1986, fig. 82, p. 127. Georgios A. Soteriou and Maria G. Soteriou, He basilike tou Hagiou Demetriou Thessalonikes, 2 v., Athens, 1952, contains additional views of the north side of the nave, pl. 1 b, 2 a, that add little to Krautheimer's figure for present purposes. A detail may be found in the article by Susan Boyd, available online, cited in Footnote 46. Further photographs appear in Charalambos Bakirtzis, ed., Mosaics of Thessaloniki 4th–14th century, Athens, 2012, pp. 134, and 139–41 (including the south side of the nave).

Slobodan Ćuřcić, Architecture in the Balkans, New Haven, 2010, p. 107, provided a photograph (fig. 104) and remarked that “the splendid opus sectile decoration of the nave arcade spandrels must be seen as original fifth-century work, and not a result of the seventh-century restoration”, but did not detail his reasons, instead referring to an extensive literature on Hagios Demetrios in which “issues related to its archaeology are much contested”, with a bibliographic reference (n. 87 on p. 844) to Charalambos Bakirtzis, The Basilica of St Demetrius, Thessaloniki, 1988, repr. 1998. Bakirtzis, however, attributes the spandrel decoration to the seventh century, constrasting it with earlier revetment elsewhere in the church (pp. 34–35).

33. Jerusalem would have been a parallel case to Alexandria: Judith MacKenzie, The Architecture of Alexandria and Egypt 300 BC–AD 700, New Haven, 2007, has shown the close connection between the architecture of Constantinople and Alexandria, although I am certain she erred in concluding that Alexandria led Constantinople.

34. Op. cit., p. 99. The caption to Creswell's pl. 10 says they are on Pier 1, but this must be a mistake because the staircase, described by Creswell, p. 93, and shown on his plan, fig. 21, and other plans, is directly east of Pier 2. For more of the plain marble surrounding these panels see EA.CA.722 and EA.CA.723.

35. Scaling from the published plates, the vertical sections of Panel 10A are about 23 cm. wide, and the field of Panel 10A, inside the border, is about 90 by 64 cm.

36. “The First Mosaic Discovered in Ramla”, Israel Exploration Journal, v. 26, 1976, pp. 104–19, p. 113.

37. Marçais, “The Panels of Carved Wood in the Aqṣā Mosque at Jerusalem”, in Early Muslim Architecture, 2nd ed., v 2., pp. 127–37; Hamilton, The Structural History of the Aqsa Mosque: A Record of Archaeological Gleanings from the Repairs of 1938–1942, Jerusalem, 1949; Hillenbrand, “Umayyad Woodwork in the Aqṣā Mosque”, in Julian Raby et al., Bayt al-Maqdis: Jerusalem and Early Islam, (Oxford Studies in Islamic Art, v. 9), pt. 2, pp. 271–310, p. 289 for illustrations of these two panels.

38. Hillenbrand, op. cit., p. 288.

39. Op. cit., p. 135.

40. Op. cit., p. 301, n. 70. The figure reference is apparently faulty; I think Hillenbrand must have meant Fig. 50, as the border in Fig. 49 is ordinary.

41. James W. Allan, ed., A Short Account of Early Muslim Architecture, Revised and supplemented by James W. Allan, Aldershot, 1989, pp. 79–82. Hamilton fudged this argument in “Once Again the Aqsa”, Bayt al-Maqdis: Jerusalem and Early Islam, (Oxford Studies in Islamic Art, v. 9), pt. 1, pp. 141–44; the argument in Short Account is much clearer.

42. See Creswell, Early Muslim Architecture, 2nd ed., v. 1, pt. 2, p. 374, n. 1.

43. Such as Early Muslim Architecture, 2nd ed. v. 1, pt. 1, pl. 17 b. For the motif of the heart-shaped surround of palmettes in Panel 10A, which may be mistaken as Sasanian, consider the floor mosaic from the “late Byzantine” (pre-Islamic) Burnt Palace at Madaba, which is typical of the whole group of Late Antique or early Byzantine church mosaics in the area (Michele Piccirillo, The Mosaics of Jordan, Amman, 1993, p. 79, no. 54). It also occurs in the Dome of the Rock mosaics, in the soffit of one of the arches of the inner arcades, Early Muslim Architecture, 2nd ed., v. 1, pt. 1, pl. 24 c, though in more developed form.

44. Nuseibeh, pp. 58, 60.

45. Creswell, op. cit., p. 86, cites Charles Clermont-Ganneau, Archaeological Researches in Palestine During the Years 1873–1874, 2 v., London, 1899, v. 1, p. 214, to the effect that he saw Pier 8 “completely stripped of its marble casing”, but Clermont-Ganneau identified the pier in a footnote as follows: “I think, but owing to the disorder of my notes I cannot be certain, that this was the pier at the south-west angle (No. 10 on the plan)”, which is on p. 154; the pier marked on this plan by the number 10 is Creswell's Pier 7, not 8.

46. Ann Terry, “The Opus Sectile in the Eufrasius Cathedral at Poreč”, Dumbarton Oaks Papers, v. 40, pp. 147–64, p. 149, n. 9, citing Susan Boyd, “A Little-Known Technique of Architectural Sculpture: Champlevé Reliefs from Cyprus”, Jahrbuch der Österreichischen Byzantinistik, v. 32, 1982, pp. 312–25, which I have not consulted. However, Boyd's article “The decorative program of the champlevé revetment from the episcopal basilica at Kourion in Cyprus”, Actes du XIe congrès international d'archéologie chrétienne (Publications de l'École française de Rome, v. 123), pp. 1821–40, is online as of 3 May 2014 at ; it draws parallels among the Byzantine monuments I cite here, without noting the two panels at the south entry of the Dome of the Rock, and adds some other material.

Boyd's definitive discussion of the Kourion revetment is in A.H.S. Megaw, ed., Kourion: Excavations in the Episcopal Precinct, Washington, D.C., 2007, pp. 235–320; for “wave-crested” designs pp. 276–87, pl. 6.14 d and e, 6.15–6.17; conclusions and dating pp. 291–94; examples of the “champlevé” technique elsewhere, with catalogue, pp. 294–301.

I thank Dr. Terry for comments on the Dome of the Rock panels.

47. For Hagia Sophia see Cyril Mango and Ahmet Ertuğ, Hagia Sophia, A Vision for Empires, pp. 15, 65, 70, 71, and 101; related panels, p. 41. For Poreč see Terry, op. cit.; Ann Terry and Henry Maguire, Dynamic Splendor: The Wall Mosaics in the Cathedral of Eufrasius at Poreč, 2 v., University Park, Pa., 2007, v. 2, pl. 1; and Peter Brown, Through the Eye of a Needle: Wealth, the Fall of Rome, and the Making of Christianity in the West, 350–550 AD, Princeton, 2012, fig. 10. The profile occurs in the fifth-century Orthodox Baptistry in Ravenna, but not in relief; see Friedrich Wilhelm Deichmann, Ravenna, multiple v., Wiesbaden, 1969–89, several reprints, v. 3, Bauten und Mosaiken von Ravenna, pl. 95.

48. Denys Pringle, The Churches of the Crusader Kingdom of Jerusalem, 4 v., Cambridge, 1993–2009, v. 3, p. 372.

49. Some fragments have been published online as of 4 May 2014 at;pa;ISL .

50. The Dome of the Rock in Jerusalem. A Description of its Structure & Decoration, Oxford, 1924, pp. 82–83.

51. Op. cit., pp. 19–20. In 1874 Clermont-Ganneau saw one pier “completely stripped of its marble casing” (see Footnote 45).

52. Op. cit., p. 84.

53. Creswell, op. cit., p. 85; Richmond, quoted above.

54. Nuseibeh, p. 111, an oblique view.

55. Nuseibeh, p. 67, a good view.

56. Nuseibeh, p. 65, an oblique view.

57. In Nuseibeh, p. 110, it has been changed.

58. It is now in the Victoria and Albert Museum, no. E.1267–1963, and can be found online at as of 9 May 2014.

59. “What is the Date of the Dome of the Rock?”, Bayt al-Maqdis, v. 1, pp. 59– 87, pp. 76–77.

60. The roundels in the corners and the construction of the design of the double frames are not right for the period, nor have I found a kantharos in Umayyad art.

61. It is also possible that either the single frames or the double frames or both are are later restorations of Sulaymān's work employing a new design or designs, which I doubt. But if so, they would be Neoclassic, and we may have the name of the designer. St Laurent, “The Dome of the Rock: Restorations and Significance”, p. 423, cites documents dating from 1232 to 1234/1815–17 relating to the restoration of Sultan Maḥmūd II: “The master builder (qalfa) came from Istanbul to supervise the project, and a master architect Abu Salah Efendi is mentioned. … According to the second part of the document written in 1233/1817, large quantities of marble were brought from Damascus. The document lists a large number of craftsmen, twelve in number, working with marble, including Ilyas Shanawi—the master and supervisor—Butros Lutfi, Yusuf ibn Abu Touma, Antonios Touma, and Nicola Danani. Two of them were masters (wali al-nʿim) and ten were craftsmen (muʿallim). From their names it is clear that most of these artisans were Christians.” Nicola Danani was surely an Italian.

62. Mango and Ertuğ, op. cit., pp. 53–60.

63. Michael Meinecke, Die Mamlukische Architektur in Ägpten und Syrien (648/1250 bis 923/1517) (Abhandlungen des Deutschen Archäologischen Instituts Kairo, Islamische Reihe, v. 5), 2 v., Glückstadt, 1992, v. 2, pp. 295–96, and v. 1, pl. 91; see also the index s.v. “Zahnschnitt”.

64. Michael Hamilton Burgoyne, Mamluk Jerusalem, an Architectural Study, n.p. 1987, pp. 517–18.

65. For example, the Dār Bāyrām Jāwīsh, ca. 959/1551–52, no. 13 in Yusuf Natsheh, “Catalogue of Buildings”, in Auld, et al., ed., Ottoman Jerusalem, v. 2, pp. 657–1085, p. 733, pl. 13.5.

66. Sylvia Auld, “Stars, Roses, and Interlace: Architectural Decoration in Ottoman Jerusalem”, in Auld, et al., ed., op. cit., pp. 355–88, pl. 24.31, 24.32; locations and dates in fig. 24.26.

67. Creswell, op. cit., pp. 142–43.

68. Ibn ʿAbd Rabbihi, Al-ʿIqd al-farīd, ed Aḥmad Amīn, et al., Cairo, 1940–53, pp. 260–62; trans Muḥammad Shafīʿ, “A Description of the Two Sanctuaries of Islám by Ibn ʿAbd Rabbihi (d. 940)”, A Volume of Oriental Studies Presented to Edward G. Browne, ed T. W. Arnold and Reynold A. Nicholson, Cambridge, 1922, pp. 416–38; on the author's possible visit p. 422, this excerpt pp. 432–34.

69. The Arabic phrase is from the text ed. Ahmad Amīn, et al., p. 261.

70. La mosquée omeyyade de Médine, Paris, 1947, p.80, fig. 3, reproduced in Creswell, op. cit., fig. 75.

71. Note the absence in the interior of the Dome of the Rock of the low-relief engaged colonnettes, which, however, appear in the Great Mosque of Damascus (see Appendix B). There is such work at the Dome of the Rock in the east porch (Creswell, op. cit., pl. 2 c-e, EA.CA.724, EA.CA.726), but the only good place for such elements inside would have been on the inside of the outer wall.

72. Opus sectile is prominent at Porec (see Footnote 46); for opus sectile panels in Hagia Sophia see Mango and Ertuğ, op. cit., pp. 33, 41, 68–69, 71, and 72. Square plaques in the “champlevé” technique found at Kourion (Boyd, in Kourion, pp. 270–74, fig. 6.20, pl. 6.13, 6.14 a and b), perhaps suggest designs that might have been used, but they are far too small for the surfaces available in the Dome of the Rock.

73. Krautheimer, op. cit., fig. 187 and 191; Deichmann, op. cit., pl. 285–303.

74. Mango and Ertuğ, op. cit., p. 41.

75. Op. cit., p. 11.

76. For the material of the sections atop the piers see Creswell, op. cit., p. 86, relying on observations by Clermont-Ganneau. In Clermont-Ganneau's “Letters: VII-X”, Palestine Exploration Fund Quarterly Statement, v. 6, 1874, pp. 135–178, pp. 137–39 (available online as of 20 April 2014 at, he reported the results of his examination of the exterior side of the entablature between the capital of one of the columns and the springing of the masonry arch above it, where he was able to see the wooden core of the tie beam. This account was reprinted in his contribution to the Survey of Western Palestine, pp. 309–10, and in modified form in his Archaeological Researches, pp. 208–11, from which Creswell took his fig. 29 a. Creswell supplemented this account with a section of the octagonal arcade drawn by Richmond (and published by him in his Dome of the Rock in Jerusalem, fig.10) that shows a wooden tie beam in section and the inside face—not section—of a pier. Creswell wrote “this entablature is carried round the inner faces of the angle piers” (p. 86), without specifying the material of the portion of the entablature atop the piers. I think it would be odd for this entire zone to be made of wood, and I imagine it is actually stone, but it could be stone faced with wood.

Clermont-Ganneau also wrote that “this part [the exterior side?] of the beam, now masked by a marble casing, was originally intended to be seen, because we found the ornamentation of the beam continuing under the marble” (“Letters”, p. 138). What did he mean by “the ornamentation of the beam”? Creswell took it to apply to “the metal plate covering the outer side of the tie beam” (loc. cit., where the reference in n. 9 is faulty). If that is correct, then this metal must have been present and must have been removed for Clermont-Ganneau to see the wood behind it, although he does not note that fact.

77. An elevation detail by William Harvey. If I understand the elevation correctly, the L and Ri segments of the pier are shown; on both the scroll frieze is R → L, whereas the tie beam to the right of the pier is L → R. The pier can be identified as Pier 2 by comparing its foliate and inscription mosaics with Creswell, op. cit., pl. 12 a and b.

78. Op. cit., pp. 174–75, pl. 47 a–c (EA.CA.107, EA.CA.390 and EA.CA.426; EA.CA.388 and EA.CA.395; EA.CA.389, EA.CA.394, and EA.CA.424) and 50 b.

79. Op. cit., pl. 2 c–e; see also EA.CA.61, EA.CA.724, and EA.CA.726. For San Vitale, Deichmann, op. cit., pl. 299.

80. Op. cit., western vestibule grilles, pl. 59 a and b; western riwāq grilles, pl. 59 c and d; restoration drawing, fig. 92.

81. Creswell's pl. 62A d, the photograph by Bonfils, is represented in the Creswell Archive by EA.CA.715.

82. Op. cit., pp. 174–76.

83. The Great Mosque of Damascus, Leiden, 2001, p. 59, n. 10.

84. Badr el-Hage, Des photographes à Damas 1840–1918, Paris, 2000, pp. 96 and 97.

85. Op. cit., p. 60 and fig. 30 and 31.

86. I obtained a scan of this photograph from the Library of Congress, , with identifying information at , on 2 June 2014.

87. El-Hage, op. cit., p. 92, where the dating “ca 1900” perhaps refers to the print. See p. 47 for brief information on Sulaymān al-Hakīm. The central part of this photograph was published by Phené Spiers, “The Great Mosque of Damascus”, Palestine Exploration Fund, Quarterly Statement, 1897, pp. 282–301, fig. 5, which can be found online.

88. Op. cit., p. 288, fig. 3 and 6.

89. That would be 7.15 m. to the tops of the cornices including 60 cm. for them. Max van Berchem, who was sometimes offhand about measurements, left notes giving the height of the piers and the large pilasters that correspond to them as the suspiciously round 7.50 m. to the springing of the nave arches, “Notes archéologiques sur la mosquée des Omeyyades”, Bulletin d'Études Orientales, v. 7–8, 1937–45, pp. 39–56, p. 41.

90. I measured along the axis of the first pilaster from the right on el-Hage's published plate of the al-Hakīm photograph. The height of the wall to the top of the pilaster register is 83 mm. and the height of the pilaster is a bit over 21 mm. Dividing 6.65 m. by 83 and multiplying by 21 gives a result of 1.68 m., which rounds to 1.7 m. Trying the same measurements on a scan of the plate, at a smaller size, I obtain 1.73 m., which rounds to 1.7 m. I cite both processes to point out the often-ignored necessity of observing the law of significant digits and to illustrate that for me, at least, scaling from photographs is much less accurate than measuring the real world. The interested reader may wish to try his own hand at it.

91. I first measured the published plate. The pilaster is 46 mm. high and the register above it 14 mm. If the pilaster is 1.77 m. high the register above it is 54 cm. high. I then measured a print of the scan of EA.CA.395 that is available online, which is a close-up of the same pilaster. The pilaster is 83 mm. high and the register above 25 mm., giving 53 cm. for its real height.

92. The total height of the wall to the top of the pilaster is 175 mm. and the height of the pilaster is 46 mm., so if the elevation to the top of the pilaster is 6.65 m., the pilaster is 1.75 m, to be rounded.

93. Again, the total height of the wall to the top of the pilaster is 175 mm. and the height of the enframing rectangle is 47 mm. measured below the pilaster, so it is 1.79 m., to be rounded. The left-hand lozenge is 35 mm. tall, so it is 1.33 m. high, to be rounded.

94. On the published plate the pilaster register above the long, vertical side of the half-lozenge is 16 mm. tall and the half-lozenge is 11 mm. tall. If the pilaster register is 1.77 m. tall, the half-lozenge is 1.22 m. tall, to be rounded.

95. I measured from el-Hage's plate of the al-Hakīm photograph to the right of the leftmost pilaster. The pilaster register is 25 mm. tall and the shorter register is 22 mm. tall; if the pilaster register is 1.77 m. tall the shorter register is 1.56 m. tall, to be rounded.

96. Meinecke, Die Mamlukische Architektur v. 1, p. 96 and pl. 60; v. 2, p. 145.

97. Meinecke, op. cit., v. 2, p. 33.

98. Meinecke, op. cit., v. 2, pp. 52, 57–58; Michael Meinecke, “ Das Mausoleum des Qalā'ūn in Kairo: Untersuchungen zur Genese der mamlukischen Architecturdekoration”, Deutsches Archeologisches Institut, Abteilung Kairo, Mitteilungen, v. 27, 1971, pp. 47–80, pp. 64–67, 74–75.

99. These are possibly Mamlūk; see Michael Meinecke, “Syrian Blue-and-white Tile of the 9th/15th Century”, Damaszener Mitteilungen, v. 3, 1988, pp. 203–14, p. 210, n. 26.

100. Sauvaget, La mosquée omeyyade de Médine, Paris, 1947, p.80, fig. 3; Creswell, op. cit., pp. 142–49, who remarks the similarity between Damascus and Madīnah.

101. Flood, op. cit., pp. 196–97; for Ibn Ḥauqal see the translation by Guy Le Strange, Palestine Under the Moslems, London, 1890, p. 236.

102. Op. cit., p. 177. To clarify, van Berchem's photograph, Creswell's pl. 62A a, shows two pairs of lozenge panels, each pair arranged like the one van Berchem published a photograph of, reproduced here as Figure 49. Van Berchem enumerated the four inscriptions and gave their locations in the article Creswell used.

103. Inscriptions arabes de Syrie (“Extrait des Mémoires de l'Institut Égyptien” [v. 3, see Creswell, op. cit., p. 167, n. 4, but repaginated]), Cairo, 1897, pp. 12–19, pl. 4, fig. 7.

104. If the inscription panels are still to be found they should be inspected with a view to determining whether they might be spolia.

105. In Kay Kohlmeyer et al., Land des Baal, Mainz, 1984, no. 262, pp. 284–85; a reduced version of this entry appears in the English version of this exhibition catalogue, Harvey Weiss, ed., Ebla to Damascus, pp. 525–26. The museum number for the panel is A.11.

106. Op. cit., p. 167 n. 3; the detail about the three inscriptions remaining after the fire is taken from van Berchem, op. cit., p. 91. Meinecke confirmed that two inscriptions were preserved in the museum.

107. “Notes sur les mosaïques du Dôme du Rocher et de la mosquée de Damas a propos d'un livre de Mme. Marguerite Gautier van Berchem” Cahiers Archéologiques, v. 22, 1972, pp. 201–32, fig. 36 and 37. I thank Lorenz Korn for a copy of this article.

108. Op. cit., p. 228. Where are Sauvaget's papers?

109. Shown schematically in a drawing, Otto-Dorn, “Grabungen im Umayyadischen Ruṣāfah”, Ars Orientalis, v. 2, 1957, pp. 119–33, pl. 2, fig. 7, between the thin pilasters of an arcade, described on p. 125.

110. Op. cit., pp. 152, 180–96.

111. Op. cit., pp. 164–65.

112. Flood, op. cit., p. 2, 141–42, referred to reuse of materials from the church and illustrated a stone in the south wall on which a partially effaced relief figure can be seen, fig. 9 and 69, attributing it to the church. For spolia from outside Damascus reportedly used in the mosque see Flood, op. cit., pp. 201–02.

113. Boyd in Kourion, pp. 276–88, pl. 6.14–6.18. The article by Boyd available online, cited in Footnote 46, illustrates one of the Kourion panels next to a related opus sectile panel from Poreč, fig. 1 and 2.

114. Terry and Maguire, op. cit., v. 2, pl. 1.

115. Stern, op. cit., pp. 225–27, fig. 27 (photograph), 28 (drawings), and 39 (photograph); see also Lucien Golvin, Essai sur l'architecture religieuse musulmane, v. 2, L'art religieux des Umayyades de Syrie, Paris, 1971, p. 171–73 and fig. 40, which is two drawings each of the two pieces seen in Stern's fig. 27, apparently sketches after Sauvaget's drawings and his own somewhat more complete drawings after the photograph. Stern's fig. 39 was reproduced by Flood, op. cit., pl. 29.