Marble from Lod-Diospolis: Lessons in Recycling

Preface

The chief mosque of Ar Ramlah is in the market, and it is even more beautiful and graceful than that of Damascus. It is called Al Abyad (the White Mosque). In all Islam there is found no finer Mihrab than the one here, and its pulpit is the most splendid to be seen after that of Jerusalem; also it possesses a beautiful minaret, built by the Khalif Hisham ibn ‘Abd al Malik. I have heard my uncle relate that when this Khalif was about to build the minaret, it was reported to him that the Christians possessed columns of marble, at this time lying buried beneath the sand, which they had prepared for the Church of Bali’ah. Thereupon the Khalif Hisham informed the Christians that either they must show him where these columns lay, or that he would demolish their church at Lydda, and employ its columns for the building of his mosque. So the Christians pointed out where they had buried their columns. They are very thick, and tall, and beautiful. The covered portion (or main building) of the mosque is flagged with marble, and the court with other stone, all carefully laid together. The gates of the main-building are made of cypress-wood and cedar, carved in the inner parts, and very beautiful in appearance. (Al-Muqaddasi 164; Le Strange 1890 [2010]: 350; see also Marmardji 1951: 83)

The purpose of this article ¹My thanks go to Alon Shavit for inviting me to write this article—a true challenge with regard to the fate of marble in the Land of Israel in general, and in Lod and Ramla in particular. The research took me back a few years, to 1985, when Prof. Nurith Kenaan-Kedar drew my attention to the interrelationship between Lod and Ramla via the architecture and art they yielded, and invited me to engage in a joint study on the reuse of column capitals from the Great Mosque of Ramla. This research did not come to fruition for various reasons and was eventually cut short due to Kenaan-Kedar’s untimely death. This study did however take shape as part of a research project (in preparation) on reuse phenomena (spolia) in late antiquity. All these are dedicated to her memory. is to present the architectural-decorative finds made of marble or other imported stone,²In the past I have often referred to the fact that certain construction and decorative materials, such as marble, colored granite, alabaster and more, are not found naturally in the Land of Israel. Their presence in Land of Israel sites is the result of importation and all that it involved. A research organization called Association for the study of Marble & Other Stones in Antiquity (ASMOSIA), was established in 1988 for the purpose of focusing on and researching these phenomena in the ancient world. This article is a continuation of my research devoted to the subject in recent years (summaries in Fischer 1998). which characterized the city of Lod-Diospolis in the Roman and Byzantine periods, and to examine the implications of these finds for developments during the Islamic periods in Lod itself and in nearby Ramla. On the face of it, the finds appear quite meager. But is this indeed the case? The premise of the article is that architectural decorative elements from the Roman and Byzantine periods were reused during the intensive reconstruction the Muslim rulers carried out in the Land of Israel at the beginning of the eighth century; this is the first part of the article. The second part is devoted to a decorative element made of marble that reflects, albeit to a limited extent, the wealth of classical art in Lod-Diospolis.

The city of Lod became better known by its new name, Diospolis, at the end of the second century. It was one of the main cities in the Land of Israel, which later became the Roman province of Palaestina (and from the fifthcentury, Palaestina Prima). Due to its geographical location between the Mediterranean coast and the Judean foothills, and from there to the Judean Mountains and especially to Jerusalem, it developed into a central crossroads that was significantly nurtured by the Roman government, giving it a high urban status (Schwartz 1991; Fischer, Isaac, and Roll 1996). Despite the extensive archaeological activity carried out in Lod in recent decades, resulting in impressive discoveries (mainly of buildings with mosaic floors; Gorzalczany 2015a; and see Gorzalczany 2015b in this volume), the architectural-decorative finds, which include elements made of imported stone, are extremely scarce. Therefore, the discovery of any of the few elements of this type as well as their treatment may shed light on the city’s place in the architectural-decorative context of the Roman and Byzantine periods. This being said, the issue of the reuse of these architectural decorative elements in the Islamic and Crusader periods is not negligible at the least. On the contrary, it contributes greatly to the completion of the cultural picture of late antiquity and the Middle Ages.

Byzantine Period Finds in Secondary Use

1. The columns and capitals from Al-Umari Mosque

Within the distinctive Islamic architectural design of the Al-Umari Mosque (Figs. 1, 2), based mostly on local limestone, two pairs of columns crowned with two pairs of Corinthian capitals stand out. In both cases, the capitals were coated with a layer of paint, giving them a golden impression.

The first, eastern, pair (Fig. 3) in the Al-Umari Mosque consists of two monolithic marble columns identical in design and dimension and Corinthian capitals that crown them.³Height of the columns 3.3 m; bottom diameter 0.6 m; top diameter 0.5 m; height of capitals 0.6 m. The columns lack bases and were placed on a low surface. The column on the left presents streaks or veins typical of Proconnesian marble, which is characterized by dark veins running lengthwise. The column on the right is made from a block of marble—its characteristic veins running crosswise, which makes it more unique and similar to Cipollino marble, which is less common in the Land of Israel. In addition to the special artistic effect that this type of hewing and design created, it gave the architectural elements the advantage of strength. Similar hewing and design are known from major monuments in the Byzantine world, including Hagia Sophia (Fig. 4) in Constantinople (Istanbul) (Sanpaolesi 1978: 161, fig. 224), but also in central churches in provincial cities, such as Hippos (Sussita) (Fig. 5).⁴My thanks go to Dror Segal for his help in identifying the types of marble and for making available to me the photographs that are part of his research work on the marble in Hippos (Sussita). The columns retained their cylindrical shape, but they were attached to each other and to the support pillar, which was originally built of ashlars, with the plaster between remaining visible. The lower part of the pillar has been renovated in recent years and faced with smoothed stone slabs that align with one another perfectly.

The façades of both capitals (Figs. 6–7) were trimmed along their entire height when they were attached to each other. Their backs, on the other hand, were not altered, and no effort was made to match the support pillars, constructed of plum concrete and ashlar stones, to the capitals’ original profile. The two capitals are identical in their general dimensions and in the relations between their details (such as the ratio between the bottom and top rows of the acanthus leaves), apart from the decoration of the upper part of the calathus (the core of the capital). On the capital on the right, the section between the two volutes and the calathus rim is decorated with a row of pseudo-ovolo similar to the wave decoration mainly in the way the sections are separated.⁵The classic ovolo (eggs and darts) consists of egg-like sections with a dart or arrowhead separation between them (Tornheim 1987). In the capital on the left, this part is adorned with a floral decoration. The decoration is unusual in both cases, as it is more reminiscent of the decoration of the Composite capitals, which lack internal volutes (helices). Capitals with a similar design were found in a survey of marble capitals conducted on the Temple Mount, where the capitals were dated to the first half of the fifth century CE, but are all found in secondary use in various Islamic complexes (Wilkinson 1987: 119–121). This is a capital with a standard structure for the end of the imperial Roman period—that is, the end of the third century CE. It is characterized by two rows of acanthus leaves, equal in height, that equally cover the calathus (capital’s body). This style does not conform to the principles of the “classic” capital, where the body of the capital was divided equally between two rows of acanthus leaves, over two-thirds of the capital’s height, and another third was taken up by the stems of the outer and inner volutes. In our case, the outer volutes stem directly from behind the outer leaves (right and left) of the upper row of acanthus leaves. Both they and the volute (spiral) curl are strongly emphasized. As mentioned, the internal volutes are absent from the design and their place is assumed by the band of ornamentation described above.⁶For a comparison between these capitals and the usual Roman classical type, see Fischer 1990; for a group of Corinthian capitals dated to the Late Roman period, see Fischer 1986. The acanthus leaves still bear the stamp of classical Roman art, the leaflets are well defined, and despite the fan shape covering the body of the capital, the leaf is still noticeable. The capitals represent the transition between standard Corinthian capitals, which were customary during the height of marble importation to the Roman Land of Israel, and the creation of uncommon capitals that heralded Byzantine period capitals. This phase dates to the fourth century and the first half of the fifth century CE. It seems to have originated in the workshops of Constantinople, which took advantage of the proximity of the capital of the Eastern Roman Empire to one of the largest marble quarries in antiquity. The example under discussion was attributed to such a workshop and dated to the middle of the fourth century CE (Zollt 1994: 112, pl. 36, no. 287).

The analysis of these capitals, like those that will be described later, is based on a long history of research in this field in general and in the Land Israel in particular. In decorative terms, these are Corinthian capitals, which underwent an extended period of development from their appearance at the end of the fifth century BCE to the end of antiquity (Fischer 1990). In the Land of Israel, this style received its distinct expression in the Roman period, along with the importation of marble to the region (Fischer 1998). With the transition to the Late Roman period (the end of the third century and fourth century), the Corinthian capital was adopted in the Land of Israel as part of the general development in architectural decoration throughout the empire, on the one hand, and unique local developments, on the other hand. As the decorative development departed from the classical line, room was made for the appearance of regional and even local versions—changes which research has struggled to follow. This difficulty stemmed also from the lack of well-dated complexes and especially from the extensive reuse of these architectural decorative elements in late antiquity and later. After a long period of individual studies on monuments of the transition period between the Roman and Byzantine periods, mainly those from the fourth century in the Land of Israel, Greece, and Asia Minor, Kautzsch published his groundbreaking book in 1936. For the first time, Corinthian capitals were presented in their archaeological-architectural context and in a developmental perspective following a timeline. Kautzsch highlighted the similarities between the styles and the regional phenomena. Thus, for example, the capitals of the key sites of the Land of Israel were described as part of the continuation of the tradition that developed in the region during the Roman period but also as part of the Byzantine Ecumene influenced by the major centers, such as the capital Constantinople, or centers such as Thessaloniki and Ephesus. Since Kautzsch’s research, many studies have attempted to focus on the production centers of the capitals, both at the marble quarrying sites (such as Proconnesus) and at sites far from the quarries. Thus, studies about large concentrations of capitals in Asia Minor and Greece have highlighted the centrality of these areas in production and especially the exploitation of the existing marble there (see Kramer 1994; Zollt 1994). Sufficient effort was not invested, however, in the study of architectural decorative elements in Egypt, Syria, Lebanon, and the Land of Israel. Architectural decoration in Syria received great attention in the studies of Christine Strube (1993–2002), who revealed a wide world of activity in the field of architectural decoration using local stone and carried out under the influence of imported marble decoration. Since no comprehensive study has yet been conducted about the development of the Corinthian capital in the Land of Israel during the period in question, most of our conclusions regarding the capitals discussed here are based on the studies that will be mentioned below.
The second pair of columns (the western one, Fig. 8) comprises a Procconesian marble column and a red granite column originating in Aswan, Egypt (Fig. 8, left and right, respectively). This was one of the favorite materials of the Roman world in general, and of the cities of the provincial cities emulating Rome, in particular. In the Land of Israel Caesarea became the leading importer of these columns. The red granite from Aswan earned a respectable place alongside the “white” marble and the gray granite (from Asia Minor). Evidence for this is the large number of red granite columns collected at the site of Caesarea (Fig. 9). ⁷On the characteristics of the red granite stone from Aswan, see http://granites.uk/aswan_red.html The capitals of the two columns from Lod are identical (Fig. 10) but different from the capitals of the eastern pair of columns described above. The former reflect clearly the process of (further) loss of the classical style. As in the previous column pair, here too two rows of acanthus leaves occupy most of the body, with the completely degenerate volutes growing behind the central leaves and ending in a spiral that is not emphasized at all. The leaves and leaflets have lost their definition, their veins are completely shallow, and they adjoin into a tight fan that covers and obscures the body of the capital. This is the last stage before the appearance of the Byzantine capital. The origin of this design was also in the workshops of Constantinople and it is dated to 350–425 CE (Zollt 1994: 118, pl. 38, no. 38).
Identical capitals discovered on the Temple Mount were dated to the fourth century and the beginning of the fifth century CE (Wilkinson 1987: 116–117, nos. 70–71).

2. Fragment of a column with an Arabic inscription (courtyard of the Al-Umari Mosque)

In the courtyard of the Al-Umari Mosque there is about half of a marble column that was sawed at some point to create a surface for an engraved inscription (Figs. 11–12). The inscription is engraved in relief on the smoothed surface, and it is dated to the thirteenth of the month of Ramadan, Hijri year 892 (September 2, 1487), and is a documentation of the market prices and taxes in the Jabal Nablus area (Shechem).

Though the origin of the marble of this column seems to be Proconnesus, it may possibly be Thasos, as indicated by the large crystals evident in it. The marble quarries of the island of Thasos were significant suppliers, mainly in the Byzantine period, when they specialized, among other things, in liturgical furniture for religious buildings (Sodini 2000). The dimensions of the column fragment (length ca. 2.8 m, diameter ca. 0.6 m) indicate that it may have originally been identical to the columns inside the mosque, described above. The practice of cutting or sawing the pillars for new uses was a common one, and was even documented during the nineteenth century and at the beginning of the twentieth century at various sites in the Land of Israel (Fischer 1999: 283, fig. 6).

In the courtyard of the Al-Umari Mosque there is about half of a marble column that was sawed at some point to create a surface for an engraved inscription (Figs. 11–12). The inscription is engraved in relief on the smoothed surface, and it is dated to the thirteenth of the month of Ramadan, Hijri year 892 (September 2, 1487), ⁸These details were given to me by Professor Moshe Sharon, and I thank him for that. After I gave him the details of the column and the inscription, he cataloged the artifact as Economic Text 892. Lydda. My thanks go also to Dr. Tawfiq Da’adli for his assistance in clarifying matters related to the inscription. and is a documentation of the market prices and taxes in the Jabal Nablus area (Shechem).

Though the origin of the marble of this column seems to be Proconnesus, it may possibly be Thasos, as indicated by the large crystals evident in it. The marble quarries of the island of Thasos were significant suppliers, mainly in the Byzantine period, when they specialized, among other things, in liturgical furniture for religious buildings (Sodini 2000). The dimensions of the column fragment (length ca. 2.8 m, diameter ca. 0.6 m) indicate that it may have originally been identical to the columns inside the mosque, described above. The practice of cutting or sawing the pillars for new uses was a common one, and was even documented during the nineteenth century and at the beginning of the twentieth century at various sites in the Land of Israel (Fischer 1999: 283, fig. 6).

We may assume that the artifacts described above were part of the Byzantine church dedicated to Saint Georgios (commonly known as Saint George), the predecessor of the Crusader church named after Saint George, which is adjacent to the Al-Umari Mosque and is even integrated into it (Pringle 1998: 9–27). In his book, Pringle gives a detailed description of the remains of the Byzantine church and of the Crusader church at their various stages, also referring to some architectural decorative elements from the Byzantine stage that were incorporated into the Crusader church and the Al-Umari Mosque. For example, the two pairs of Corinthian capitals described above are mentioned by Pringle as a Byzantine relic in secondary use in the Al-Umari Mosque (Pringle 1998: 25, pl. XVIa). The description of the church from the Byzantine period is based mainly on the impressions of Clermont-Ganneau from the 1870s (Clermont-Ganneau 1896: 105–106). According to these descriptions, it was a basilica structure, about 47 m long and 24 m wide. During the Crusader period, its outline and dimensions were preserved. Remains of the twelfth-century church were incorporated into the front yard of the mosque (Pringle 1998: 19, fig. 3).

According to the Byzantine tradition that took shape at the beginning of the sixth century, the Church of Saint George was the main church in the city of Lod-Diospolis. It was dedicated, at that time, to Georgios, a soldier-saint born in Lod who died a martyr’s death in nearby Emmaus (Schwartz 1991: 123–126). We must assume that this is the church that was described both on the map of Madaba as part of the description of Lod-Lydda-Diospolis (Donner 1992: 54–55) and in the mosaic of the Church of Saint Stephen in Umm ar-Rasas, Jordan (Piccirillo 1993: 219, pl. 345). The description in Umm ar-Rasas, which refers to the cities where bishops lived, indicates Lod’s status among the other principal cities. The church in its present form is based on the remains of Crusader construction with significant additions made in the nineteenth century (Pringle 1998; on the spectacular iconostasis of the church, see Traister’s (2016) article in this volume).

Lod lost its historical status to Ramla, the new capital of the province, declining from “a provincial capital to a marginal town” (Gat 2015). At the same time, it maintained a certain traditional line reflected in the monuments from the Islamic and Crusader periods, which drew influence from the monuments of the Byzantine period. It seems that Lod’s ancient monumental remains suffered the final “blow” during the extensive construction activity in the reign of Mamluk Sultan Baybars. Many architectural elements, including marble artifacts, were incorporated into the new buildings that embellished Lod, including the Jindas Bridge, north of the city (Gat 2015: 146, fig. 1), highlighting its significant place on the famous Mamluk “postal road”. Another building that can be attributed to this period is the Al-Umari Mosque, which we dealt with above. An inscription incorporated into the front door frame records the construction date of the mosque (Gat 2015: 145; for a detailed description of Lod during the Islamic period, see Sharon 2012).

Only a few sections of the Byzantine church survive—possibly only a few wall courses, some of which were incorporated into the mosque walls (Pringle 1998: pls. V–VII). On the other hand, four marble and granite columns and two pairs of twin capitals (Pringle 1998: 24, pl. XVIa) were installed in the central hall of the adjacent Al-Umari Mosque. In the mid-nineteenth century Guérin (1869: 324) had already noted the reuse of Byzantine period artifacts in the Islamic period. He distinguished, from a professional standpoint, between the nature of Byzantine construction and that of later periods. One of the distinct criteria was the use of marble columns in the Byzantine period in comparison with the use of limestone columns based on the tambour technique (column drums) in the Crusader and the Islamic periods.⁹“En penetrant dans ce sanctuaire musulman, j’y ai observe deux piliers, flanqués chacun de deux colonnes monoliths de marbre, à chapiteaux corinthien, qui proviennent de la basilique byzantine; car les colonnes de l’eglise du moyen age, à en juges du moins par celles du choeur, étaient en pierre et formees de tambours cylindriques superposes.” One of the columns bore visible remains of a Greek inscription (now illegible), which was recorded by Clermont-Ganneau (1896: 107–108, fig. 13) and read as follows:¹⁰This is the translation according to Clermont-Ganneau: “The worshipful pastors who sit at the head of this city for a long time past illuminated by Christ (or of this old and illustrious Christian city), having adorned this illustrious temple…” See a transliteration and translation of the inscription also in Schwartz 1991: 124–125, n. 133, but with an incorrect indication of the location of the inscription.

Clermont-Ganneau, followed by other researchers, saw this inscription as proof of the existence of a church from the Byzantine period. Di-Segni, on the other hand, dated the inscription to a later period, mainly on the basis of paleography.¹¹My thanks go to Leah Di-Segni for her help with the epigraphical issues raised by the abovementioned inscription, as well as her reference to the repeated use of columns with inscriptions, which will be discussed below. She offers the following translation: “Those who ruled the city, the honorable shepherds (of the city) whom Christ enlightened, (they are the ones) who will decorate this magnificent house (temple).”¹²Clermont-Ganneau pointed out that the inscription was written in a rhythm poem style, and Di-Segni commented that the small flowers (fleurons), were used as punctuation marks, indeed dividing the text into three groups of 12 syllables. It seems that the inscription was engraved long after the bishops (the shepherds, in the language of the inscription) of Lod, built or decorated the church. We cannot rule out that during the Islamic periods marble artifacts arrived at various sites from more distant places rather than close ones. The marble sites of the Roman and Byzantine Land of Israel, Caesarea and Ashkelon, were main sources of the reuse of marble elements (Fischer 1996). Even columns and capitals found in Lod and Ramla could have originated in these centers. In this context, a noteworthy dedicatory inscription is one in honor of Emperor Commodus that was found engraved on a marble column in Ramla, near the Tower of the Forty Martyrs, among the ruins of the White Mosque. Clermont-Ganneau copied and published it (RAO 7: 174–177; Clermont–Ganneau 1884–1885: 5e no. 1, p. 93, rapport). Di-Segni published this inscription and another inscription with the same wording (Di-Segni 1997: 486–490, nos. 141–142), and they were published again by the editors of CIIP 3 (nos. 275–278, 2337–2338), without reference to the previous publication of Di-Segni. It is interesting to note that, according to the report of the resident of Ramla who sold the first column, its came from Ashkelon. This was corroborated when the second column, bearing the same inscription, as mentioned, was discovered in Ashkelon. On the other hand, in our case it is apparent that the exchange of the artifacts was local and the inscription on the column in the Al-Umari Mosque, which was described above, confirms the existence of a Byzantine church in the place from which the architectural elements were transferred both to the Crusader church and to the mosque.

Notes on the use of marble in the Byzantine period

The presentation of the finds, albeit few, that represent the “classical” architectural decoration that survived in Lod indicates the use of such elements either as reuse of readily available material or as part of private import of monumental marble elements in the Byzantine period. In the past, I have repeatedly referred to the phenomenon of importing marble to the Land of Israel during the imperial Roman period, which culminated in the second and thirdcenturies CE (Fischer 1998). After a certain hiatus in private importation of monumental marbles at the beginning of the fourth century, we witness the renewal of the phenomenon, and even more so in the fifth and sixth centuries, including the appearance of marble craftsmen marmorarii (Fischer 1994: 24–26). It is possible that very heavy architectural elements, especially columns, were taken from imperial Roman sites already during this period (mainly those that had fallen out of use, such as idol temples). But it seems that more unique elements, such as column capitals, were brought from the main quarries in the Mediterranean region. For example, production and export of capitals from the main quarry in Prokonnesos (by the Sea of Marmara near Constantinople [Istanbul]) was renewed (Asgari 1995). Difficulties in the importation of heavy marble elements, especially columns, are reflected in the foregoing of their use, even in prestigious imperial projects, such as the construction of the Nea Church in Jerusalem during the reign of Emperor Justinian (527–565). Justinian obviously wanted to build a very impressive complex in the city of Jerusalem, and this was indeed accomplished, including the design of vaults and infrastructure to support the construction, evident by the finds yielded in the archaeological excavations (Avigad 1980: 229–246). Justinian’s court historian, Procopius, gave a detailed description of the construction of the building and the difficulties that the builders encountered, but also managed to overcome. One of the problems was the supply of proper columns for such a complex. Procopius describes this in his book, Of the Buildings of Justinian (5, 6): “For the church required throughout columns whose appearance would not fall short of the beauty of the building and of such a size that they could resist the weight of the load which would rest upon them.” Here Procopius alludes to the fact that the builders (the emperor himself) intended to bring marble columns from the port cities of the Mediterranean. But, Procopius continues: “… the site itself, being inland very far from the sea and walled about on all sides by quite steep hills… made it impossible for those who were preparing the foundations to bring columns from outside.” Indeed, both the building’s stone and columns were brought from limestone quarries, some of which were discovered near Jerusalem. This description generally fits most of the monumental projects in the Byzantine period, for which local stone was preferred to imported marble. On the other hand, the few artifacts that did survive in Lod, and were described here, point to the city’s capability to recycle marble columns from an earlier period, or even import them when the church was built.

3. Ramla

A construction boom was underway throughout the country after the transition period that followed the establishment of the Muslim rule in the Land of Israel. The largest investment centered on the construction of the new capital of Jund Filastin, the city of Ramla. On the regional level, Ramla took the place of neighboring Lod (Diospolis). But beyond this, it received all the attention it deserved as the capital city of the entire country (Luz 1996), and for that its urban and architectural development is noteworthy (Gat 2003). According to documentation of the period, Ramla gradually and decisively captured the place of Lod as well as its treasures in general and its architectural standing in particular. The archaeological evidence for this is admittedly rather poor, as we will see later.

The confiscation and transfer of columns from the church in Lod by the new rulers—well described in the source with which this article opens—is just one of several similar cases accompanying the period of transition between one rule and another. This is the story of the spolia—various artifacts that were taken out of their original context and reused elsewhere (Brenk 1987). Their importance has been widely discussed in both ancient sources and modern studies (Fischer 1999; Greenhalgh 2009). Many monuments in the Greek, Roman, and Byzantine worlds bear the signs of reuse, sometimes following the traditional format, but in many cases in an unbridled and violent manner. These are the expressions of the language of a victorious government at the expense of the vanquished (Saradi-Mendelovici 1990). Although the sources name the White Mosque as that to which the columns from the church in Lod were supposedly transferred, the remnants of the White Mosque and especially its surviving minaret do not seem support this. This is, of course, with regard to the archaeological and architectural finds available to us for the time being, which point to extensive changes made to the building, which do not enable to reconstruct the magnificent phase described by Al-Muqaddasi (Rosen-Ayalon 2006). On the other hand, the second monumental site in Ramla, the Great Mosque, which took the place of Crusader period Latin church (the Church of Saint John; Fig. 14) offers a glance at the extensive secondary use of artifacts from the Byzantine period. Such are the capitals that adorn some of the columns of this building, representing the variety of capital types of the Byzantine period. The large number of these capitals warrants a study in itself, but this is not the place for it. We will limit ourselves here only to general comments about them, in an attempt to link them to their possible city of origin—that is, nearby Lod-Diospolis. In a study devoted to the origins of Crusader architectural sculpture in the Land of Israel, Nurith Kenaan-Kedar devoted also a concise description to the array of these capitals, but with no reference to their architectural-decorative background (Kenaan-Kedar 2004: 87–92, figs. 4–8).¹³As stated above (see footnote 1), the treatment of these materials began as part of a joint study planned together with the late Nurith Kenaan-Kedar in 1985. The starting point was, and still is, a description of the array of capitals from the Great Mosque, according to Nurith’s format and comments. Here, and also in the study to be published, we maintain this format. It is very possible that the columns in this building were being reused for a third time—the artifacts from the Byzantine period having been used in an earlier Islamic structure (the White Mosque?) and then moved and reused in the Crusader structure, in which they continued to be used also in the framework of the Mamluk mosque that succeeded it.

The Great Mosque (formerly the Latin church) is a basilica structure with three naves separated by six freestanding pillars (Fig. 15). Two additional pillars are attached to the church wall at the western end of the hall. Pringle (1998: 187–195) offers the clearest and most comprehensive description of the building and all its components. The column capitals from the Byzantine period were alternately incorporated into one or two of the columns that make up the clustered columns surrounding the pillars (Pringle 1998: 190, CVa–h).

The capitals described in Plate CV (Pringle 1998) may be those that Kenaan-Kedar published in her article (2004: figs. 4–8; Fig. 16).

Out of twenty-four Byzantine capitals recorded by Kenaan-Kedar, eight belong to a type consisting of a structure of leaves that enclose the body of the capital like a cup, the two volutes grow from the center of the capital and a prominent abacus flower is placed in the center. Eight additional capitals belong to the type of a complete acanthus structure with a volute and helix emerging behind the top row of acanthus leaves—that is, the type most associated with Roman tradition, which dates to the fourth century. A number of capitals belong to a model fully covered with acanthus leaves, two degenerate volutes, acanthus leaves with scrolls and rounded trays, typical of sixth^–century Constantinople craftsmanship. One capital (Kenaan-Kedar 2004: fig. 5) represents a unique design of flat schematic leaves that rise to the edge of the calathus, which has an empty round medallion in its center (sometimes used to mount a cross). It is one of the most common Constantinopolitan pattern at the beginning of the fifth century (Zollt 1994: 127–136, figs. 10–11, no. 344, pl. 39), but also into the sixth century (Peschlow 2011: 91–92, fig. 5). Only one artifact represents the composite capital, which consists of two parts: the lower part follows the Corinthian style (a regular conical body covered with acanthus leaves), and the upper part has two Ionic volutes and between them a row of ovolo (egg-and-dart) ornaments above and an astragal below. This type characterizes the revival of Byzantine art in Ravenna in northern Italy during the time of Emperor Justinian, but it was less common in the eastern provinces (Peschlow 2011: 78, fig. 7; 91–90, fig. 9). Another unique capital has one row of leaves, and another row with only corner leaves wrapped around the volutes from below (Kenaan-Kedar 2004: fig. 6). The central surface of the calathus remains vacant, and only a row of flat leaf-like astragal adorns its upper part below the calathus rim. This is an unusual type, which combines elements of composite capitals with ordinary capitals, reflecting the form changes that took place during the Byzantine period (Wilkinson 1987: 171; Strove 2002: 40). Despite the relative diversity of the capitals in this structure, overall, it seems that the capitals can be grouped into a few types, which may indicate one source, or maybe two. There is no distribution here of capitals types as it is common in other complexes that usually use capitals from many sources.

As for the interrelationship between the localities in Lod and those in Ramla, a comparison between the capitals from the Al-Umari Mosque in Lod, which were described above (presumably originating from the Church of Saint George), and the capitals from the Great Mosque in Ramla rules out any similarity between them. There are several explanations for this: either the capitals found in Ramla were not brought from Lod but from another place, or they were originally from Lod but from a different structure, of which we have no archaeological evidence.

Marble Usage in Lod during the Roman Period

A Lion-Shaped Table Leg (Trapezophoron) (Fig. 17)

After reviewing the later periods, we return to the glorious period of the Roman Empire in the Land of Israel in general and in Lod-Diospolis in particular, and examine the finds made of marble and other imported stones. Despite Lod’s location and its significance as a central city and mainly as a caput viarum within the province of Palaestina (Fischer, Isaac, and Roll 1996: 199–208; for a summary of the history of Lod, see Schwartz 2015), no monumental architectural complexes from the Roman period have been discovered in it so far (Yannai and Erhlich [2015] note this). On the other hand, in recent years the city of Lod has become famous mainly thanks to its magnificent mosaic floors (see summaries in Gorzalczany 2015a), which testify to the existence of extravagant urban residential complexes (villae urbanae) from the imperial Roman period and later in the Late Roman and Byzantine periods.

A marble artifact stands out against the background of the meager Roman decorative architectural findings—a fragment of a table leg (trapezophoron) decorated with a head and a stylized lion’s leg. The item was discovered in archaeological excavations conducted in Neve Yarak neighborhood in Lod in 1991 (Rosenberger and Shavit 1993).¹⁴Height: 45 cm, width: 20 cm. The artifact was found in an assemblage recovered in a vast destruction layer (Locus 108, Basket 515, Square C2, Stratum VI0). According to the excavators and based on the item’s style, it was attributed to Stratum V, dated to the Roman period. Eli Yannai gave me all these details in writing, as well as part of the description of the artifact. I thank him for his collegial courtesy. I was able to examine the item myself after its discovery (1993), but all efforts to reexamine it in the framework of the present research were in vain. Therefore, it remains uncertain whether the item was made of marble or alabaster. This in itself is less important in the present case, since either material would have been imported to the Land of Israel from other regions: the marble, from Prokonnesos, Marmara, and the alabaster, from Egypt. Later excavations conducted About 10 meters of this locality revealed an opulent residence with mosaic floors of the highest quality, as well as one architectural element, part of a cornice (Gorzalczany 2015b: 49, fig. 29, defined as a frieze). This limestone artifact recalls the poorly executed imitations of marble elements of that period (Fischer 1988). In any case, this artifact indicates the existence of a building decorated according to the classical style.

The artifact described here is a part of the leg of a table of a type that was very common throughout the Roman world in general and in Italy in particular. It was probably a leg of a three-legged table (tripodium), but we cannot rule out that it belonged to a one-legged table (monopodium). The item is made of a single block and shows signs of drilling and carving, although the craftsman tried to smooth it properly. The head is slightly elongated, designed with a gaping mouth expressing the lion’s roar. The tongue and sharp teeth, of which only the upper gums were preserved, the sunken eyes under the forehead muscles, and the prominent cheeks, together with the square jaw, all give it a rather natural appearance. We may presume that the pupils of the eyes were made of a different material, but they were not preserved. The lion’s mane, covering both sides of the head below the ears, is formed with simple, shallow lines, lacking any plasticity whatsoever. The ears are symmetrical. The transition to the back part is sharp and angular, not rounded as seen in types from the Early Roman period (see below). As this artifact served as a table support, the continuation of the design follows the form of a stylized, slightly arched leg, which would have terminated in a paw with claws. The back part, starting from the head and neck, is shaped in the center and for most of its length, as a rectangular block—a type of support common in Roman copies (Ridgway 1984). At the lower end of this block is a bracket, which is tail-like or curls like a volute, and it has a hole in the center. It was probably used to connect something (with a ring?) to the rest of the table legs. The upper part of the rectangular block behind the head is broken, so it is difficult to reconstruct the connection to the table. A round table with three legs of this type was very common at sites from the late Hellenistic period to Roman period. In the second century, this type became extremely popular especially in Rome and its surroundings. Comprehensive studies on the subject have noted the continuity of this design from the Hellenistic period to the republican and imperial Roman periods. Cohon has attributed the present type to the first and second centuries CE (Cohon 1984: 112–122, Types 379–397; V), as did Moss (Moss 1988: 20–25, Type 4, 304–305, catalog nos. A91–A189). In his research Moss also referred to the artifact from Lod,¹⁵I thank Christopher Moss, Irene Romano, and Hans Goette for their help in clarifying the cultural background of artifacts of this type. stating that it can definitely be dated to the secondcentury CE (Moss 1988: 154–156, note 109; 489–490, nos. 123A, 155A), and in any case, to a date later than the time of hundreds of artifacts of this type, which were discovered in the cities of Vesuvius, including Pompeii (that is, after 79 CE). According to Moss, the second century saw a mutual influence between the craftsmen that made such architectural decorative elements and those who decorated the cuirassed statues of the Roman emperors. This affords a more precise chronology (Moss 1988: 487, A121). A perfect example of such a table was discovered in Pompeii in the house of P. Casca Longus (Della Corte 1965: 285), and we may assume that such examples were used as prototypes for later models. Many examples similar to the Lod artifact were found in Rome and Ostia, and a large quantity, in the city of Aquileia. One of their hallmarks is the rectangular block on the back of the item and the volute-like bracket at its end. It appears that these items originated in the same workshop or a few workshops at most, and that they were created near the marble quarries in Asia Minor, though there was some preference there for one-legged tables (monopodia) (Feuser 2013: 23, pl. 1, 1–2). Tables of this type were often described in Roman art—in reliefs in general (Reinach 1909: 44, 255 passim) and in burial reliefs in particular. They reflect everyday life; one example is a 175–250 CE sarcophagus from Simpelveld in the Netherlands, which bears a detailed description of the interior of a Roman house on its own interior (Mols 2007–2008: 147, fig. 3).

It can be assumed that the lion-shaped marble leg was in use in one of the urban villas near which it was discovered, adorning it. The large number of mosaics that have been uncovered in recent years indicates a community with refined taste and a penchant for luxury. Notably, the small panel to the east of the northern carpet of a mosaic discovered in 1996 depicts a richly designed table with three legs shaped as lion’s legs(!), laden with elegant glass tableware. The buildings from Roman Lod emulated the wealthy residences in Italian cities. In this case too, the mosaic segments were found close to the entrance doors, and we may presume that they were set there to reflect the status and wealth of the house’s owner (Talgam 2015: 88–87, fig. 62). The item described here is highly significant, mainly as it is one of its kind among the repertoire of the Roman imperial period art found in the Land of Israel.

Summary

In this article I have attempted to trace the presence of marble in Lod in antiquity. I chose to follow the connection between two cases that testify to the use of marble imported to the Land of Israel. One is the repeated use of architectural marble elements from the Roman and Byzantine periods, seen in the Islamic periods. Since the connection between Lod and Ramla began in the eighth century, it was only natural to examine this connection also through the use of marble. Tracing this connection led us from Byzantine Lod to Islamic Lod and from there to the capital Ramla, and to documentation attesting to the supposed transfer of decorative marble elements from Lod to Ramla. The examination of the rich collection of Corinthian capitals from the Church of Saint John (later the Great Mosque in Ramla) revealed that it is difficult to establish a direct connection between the finds from Lod and those from Ramla. The second issue I dealt with was the remarkable presence of an exclusive artifact from one of the Roman villas of Lod-Diospolis—a rare proof of Lod’s splendor.

Footnotes

• 1
  My thanks go to Alon Shavit for inviting me to write this article—a true challenge with regard to the fate of marble in the Land of Israel in general, and in Lod and Ramla in particular. The research took me back a few years, to 1985, when Prof. Nurith Kenaan-Kedar drew my attention to the interrelationship between Lod and Ramla via the architecture and art they yielded, and invited me to engage in a joint study on the reuse of column capitals from the Great Mosque of Ramla. This research did not come to fruition for various reasons and was eventually cut short due to Kenaan-Kedar’s untimely death. This study did however take shape as part of a research project (in preparation) on reuse phenomena (spolia) in late antiquity. All these are dedicated to her memory.
• 2
  In the past I have often referred to the fact that certain construction and decorative materials, such as marble, colored granite, alabaster and more, are not found naturally in the Land of Israel. Their presence in Land of Israel sites is the result of importation and all that it involved. A research organization called Association for the study of Marble & Other Stones in Antiquity (ASMOSIA), was established in 1988 for the purpose of focusing on and researching these phenomena in the ancient world. This article is a continuation of my research devoted to the subject in recent years (summaries in Fischer 1998).
• 3
  Height of the columns 3.3 m; bottom diameter 0.6 m; top diameter 0.5 m; height of capitals 0.6 m.
• 4
  My thanks go to Dror Segal for his help in identifying the types of marble and for making available to me the photographs that are part of his research work on the marble in Hippos (Sussita).
• 5
  The classic ovolo (eggs and darts) consists of egg-like sections with a dart or arrowhead separation between them (Tornheim 1987).
• 6
  For a comparison between these capitals and the usual Roman classical type, see Fischer 1990; for a group of Corinthian capitals dated to the Late Roman period, see Fischer 1986.
• 7
  On the characteristics of the red granite stone from Aswan, see http://granites.uk/aswan_red.html
• 8
  These details were given to me by Professor Moshe Sharon, and I thank him for that. After I gave him the details of the column and the inscription, he cataloged the artifact as Economic Text 892. Lydda. My thanks go also to Dr. Tawfiq Da’adli for his assistance in clarifying matters related to the inscription.
• 9
  “En penetrant dans ce sanctuaire musulman, j’y ai observe deux piliers, flanqués chacun de deux colonnes monoliths de marbre, à chapiteaux corinthien, qui proviennent de la basilique byzantine; car les colonnes de l’eglise du moyen age, à en juges du moins par celles du choeur, étaient en pierre et formees de tambours cylindriques superposes.”
• 10
  This is the translation according to Clermont-Ganneau: “The worshipful pastors who sit at the head of this city for a long time past illuminated by Christ (or of this old and illustrious Christian city), having adorned this illustrious temple…” See a transliteration and translation of the inscription also in Schwartz 1991: 124–125, n. 133, but with an incorrect indication of the location of the inscription.
• 11
  My thanks go to Leah Di-Segni for her help with the epigraphical issues raised by the abovementioned inscription, as well as her reference to the repeated use of columns with inscriptions, which will be discussed below.
• 12
  Clermont-Ganneau pointed out that the inscription was written in a rhythm poem style, and Di-Segni commented that the small flowers (fleurons), were used as punctuation marks, indeed dividing the text into three groups of 12 syllables.
• 13
  As stated above (see footnote 1), the treatment of these materials began as part of a joint study planned together with the late Nurith Kenaan-Kedar in 1985. The starting point was, and still is, a description of the array of capitals from the Great Mosque, according to Nurith’s format and comments. Here, and also in the study to be published, we maintain this format.
• 14
  Height: 45 cm, width: 20 cm. The artifact was found in an assemblage recovered in a vast destruction layer (Locus 108, Basket 515, Square C2, Stratum VI0). According to the excavators and based on the item’s style, it was attributed to Stratum V, dated to the Roman period. Eli Yannai gave me all these details in writing, as well as part of the description of the artifact. I thank him for his collegial courtesy. I was able to examine the item myself after its discovery (1993), but all efforts to reexamine it in the framework of the present research were in vain. Therefore, it remains uncertain whether the item was made of marble or alabaster. This in itself is less important in the present case, since either material would have been imported to the Land of Israel from other regions: the marble, from Prokonnesos, Marmara, and the alabaster, from Egypt.
• 15
  I thank Christopher Moss, Irene Romano, and Hans Goette for their help in clarifying the cultural background of artifacts of this type.

Bibliography