An Overview of the Underwater Archaeological Evidence for the Maritime Transport of Sculptures in the Ancient Mediterranean

ABSTRACT This paper presents the results of new research on underwater archaeological evidence for the maritime transport of sculptures in the ancient Mediterranean. Through the creation of a Mediterranean-wide database and with a focus on information from surviving archaeological deposits, this study explores ancient Greek and Roman sculptures from under water as a dataset of transported artefacts that had a specific function within the maritime context of their discovery. This documentation, analysis and interpretation of underwater deposits with sculptures provide previously unexplored data regarding the geographical extent, date, reasons and circumstances of maritime movement of sculptural artefacts during Antiquity.


Introduction
The Mediterranean seabed, similarly, to the land around it, has been a depository of material remains of human civilizations inhabiting this region for several thousand years. Submerged port and harbour structures, sunken settlements, wrecks of seagoing vessels with their cargoes and jettisoned objects are some examples of the Mediterranean underwater archaeological record preserved from Antiquity, a historical period during which the Mediterranean Sea was interconnected through extensive and complex maritime networks (Horden & Purcell, 2000).
One type of artefact from this period found under water is ancient Greek and Roman sculpture. Beginning in the 16th century, hundreds of ancient sculptural artefacts, of various types, sizes, dates and materials, have been retrieved from the Mediterranean seabed by early underwater explorers, archaeologists, or simply by fishermen, sponge divers and recreational scuba divers (Diolé, 1957;Bass, 1966;Mattusch, 1997;Arata, 2005;Tzalas, 2007).
The fascinating idea of discovering and recovering ancient sculptural works of art from under water has attracted over the years the attention of both academia and the public, while it has always stimulated local enthusiasm and pride (Rackl, 1978;Stenuit, 2002;Petriaggi, 2005;Queyrel, 2012;Bellingham, 2014). Despite the large number of Mediterranean underwater sculptural discoveries and their popularity, it is difficult for scholars to determine with certainty the exact area, era, reasons and circumstances of the maritime transportation and consequent underwater deposition of this material. This paper presents the results of a recent project examining evidence for the maritime transport of sculptures in the ancient Mediterranean through the study of underwater archaeological deposits. 1 Through both Classical and maritime archaeological data, this article examines ancient sculptures from under water as a distinct group of underwater finds that had a specific function within their maritime setting. Similar to other studies focusing on transported artefacts from underwater deposits, such as amphorae, stone or other ceramics (Parker, 1992;Russell, 2013a;Leidwanger, 2017), the present article focuses on the underwater archaeological context of the sculptures, documenting the geographical and chronological extent of this maritime activity, as well as the reasons and circumstances under which sculptures of different types and materials were carried on board ancient seagoing vessels. Hence, the aim of this paper is to address the question of where, when and why sculptures were transported by sea in the ancient Mediterranean.
The analysis starts with a brief presentation of previous scholarly hypotheses and interpretations.
Following that, the data and methodology of this research are explained before proceeding to the results regarding the geographical distribution, the chronological distribution, the types of transported sculptures and the reasons of maritime transport of sculpture identified for the ancient Mediterranean world. Through this research it is hoped to highlight that ancient sculptures from underwater sites constitute a unified and solid archaeological record and an outstanding dataset of artefacts with high potential for new scholarly conclusions.

Scholarly Interpretations of the Maritime Transport of Sculptures in the Ancient Mediterranean
The hundreds of ancient sculptures found under water indicates that, under specific circumstances, these objects had been carried on ships sailing around the Mediterranean. However, inaccessibility to the seabed to most of the academic community, at least until the middle of the 20th century when underwater archaeology developed, as well as the insufficient recording of underwater archaeological contexts where sculptures have been discovered, have restricted the amount of archaeological information available about these artefacts. Therefore, from early on, scholarly research has turned to different methods of studying and interpreting the maritime transport and underwater deposition of ancient sculptures in the Mediterranean.
Since the earliest discoveries, scholars have based their interpretations of the underwater deposition and maritime transport of ancient sculptures on relevant references in ancient literary sources. The best recorded textual evidence comes from Hellenistic and Roman historians, orators and other authors, and refers to maritime activities taking place during the 2nd century BC and the 1st century AD. This period coincides with the expansion of Rome in the eastern Mediterranean and the subsequent destruction of several Greek cities, including Corinth (146 BC) and Athens (86 BC). According to ancient literary references, the devastation of these cities involved the plunder of many works of art including sculptures, which were then transported to the Italian mainland and Sicily on ships that sometimes wrecked or lost parts of their cargoes at sea (Polybius, Histories 39.2.1-2; Dio Chrysostom, The Corinthian Oration 37.42; Velleius Paterculus, 1.13; Lucian, Zeuxis 3). In addition, a few ancient authors record the maritime transport of sculptures as part of an art collectors' market that developed in the same period, where wealthy citizens ordered sculptural pieces for the decoration of their private houses and villas, necessitating their transport by ship (Cicero, Letters to Atticus 1.8.2; Against Verres, act. 2.4.126). Influenced by these texts, scholars have mostly considered the underwater deposition of ancient sculptures as a result of accidents that occurred during the maritime transport of these looting and art collection activities (Rackl, 1978, pp. 15-36;Wirth, 1994;Tzalas, 2007, pp. 342-363;Bouyia, 2012b;Koutsouflakis & Simosi, 2015).
Art-historical examinations of well-known sculptures from under water have brought to light stylistic features and sculptural details that generally match the period and areas of movement described in the ancient literary sources (Fuchs, 1963;Ridgway, 1967, pp. 329-334;Mattusch, 1997;Hemingway, 2004). Therefore, for years, any ancient sculpture found in the Mediterranean, with or without associated archaeological context, has been straightforwardly interpreted as having been transported due to looting or art collecting activities during the late Hellenistic or Roman Periods, without necessarily pursuing confirmation through further investigation of the find's context (Boardman, 1985, p. 53;Stewart, 1990, pp. 228-229;Spivey, 1996, pp. 134-136, 219-221;Neer, 2010, p. 86).
Concurrently with the research approaches presented above, there have been a few site-specific studies that have examined ancient sculptural material with consideration of their underwater archaeological context. The innovative studies and publications of the Mahdia shipwreck (identified as database entry #67: D.B.67) in Tunisia (Fuchs, 1963;Hellenkemper Salies et al., 1994) and the Porticello shipwreck (D.B.88) in Italy (Eiseman, 1979;Eiseman & Ridgway, 1987) are such attempts that present coherently all archaeological finds retrieved from the respective underwater archaeological deposits. The extensive documentation, study and publication of artefacts from these sites highlight the significance of contextualising archaeological evidence in order to better understand and interpret the sea-borne transport of sculptures as a distinct maritime phenomenon.
Despite the obvious benefits of studying ancient sculptures from underwater sites within their archaeological contexts, the separate development of the disciplines of Classical and maritime archaeology means that dividing lines still exist between scholars of each field (Bass, 1966, pp. 13-19, 70-80). Combined with the wide geographical distribution of the relevant archaeological evidence and the unique research traditions and socio-political circumstances indicative of these regions, a selective understanding of the maritime transport of sculptures in the ancient Mediterranean has developed. Therefore, there has not yet been a broad scholarly documentation, interpretation and understanding of where, when and why ancient Greek and Roman sculptures were transported on ancient seagoing vessels in Antiquity. This can only be achieved through the collective examination of the available sources and the holistic study of the existing underwater archaeological record, an approach that the present research has adopted.

Data and Methodology
In the present research the subject of maritime transport of sculptures in the ancient Mediterranean was revisited through the study of already available underwater archaeological evidence. As explained previously, the main focus of the research was the underwater archaeological context of the sculptures with the aim to document the geographical and chronological extent of this maritime activity, as well as the reasons and the circumstances under which sculptures of different types and materials were carried on board ancient seagoing vessels.
To accomplish that, a large Mediterranean-wide database was created, summarised in Table 1, recording any known underwater deposits containing ancient sculptures that have been lost under water, probably while in transit during Antiquity. The database was inspired by other archaeological studies that have recorded large number of sites and artefacts from the Mediterranean, for example A.J. Parker's, 1992 inventory in Ancient Shipwrecks of the Mediterranean and the Roman Provinces (Parker, 1992, p. ii) and the Oxford Roman Economy Project (Wilson, 2011, pp. 33-60;Oxford Roman Economy Project, 2019) including both Parker's, 1992 shipwreck database update by J. Strauss (Wilson, 2011, p. 34) and Ben Russell's research on the 'Economics of Roman Stone Trade' (Russell, 2011(Russell, , 2013a(Russell, , 2013b(Russell, , 2015. The database includes underwater deposits that had ancient freestanding sculptures of any size (small-scale figurines or statuettes, medium-scale sculptures, largescale and over life-size statues and free-standing sculptural reliefs), material (stone, bronze, terracotta, wood, ivory, wood-and-ivory) or date. The underwater deposits recorded were ancient shipwrecks with sculptures; assemblages of sculptures from an unidentified archaeological context; and single sculptures that have been found out of context as isolated finds (Figure 1).
The geographical extent of underwater deposits with sculptures examined in this study extends throughout the Mediterranean and its neighbouring regions, such as west of the Strait of Gibraltar, and the Black Sea connected through the Dardanelles, the Sea of Marmara and Bosporus. The chronology of this material record extends from the early Archaic Period, commencing approximately in the 7th century BC, to the end of Late Antiquity, approximately in the 7th century AD.
The information of the archaeological record collected in this database was mainly retrieved from academic publications both within Classical and maritime archaeology. However, since many of the known Mediterranean underwater sculptural discoveries have not been researched and/or published, news and popular media articles had to be used, as well as information from private communication with other scholars, information from personal visits in museums and archaeological services and data from the study of archaeological archives. The variability of these sources of information required the systematic recording, in a standardised and homogenous format, of as many Mediterranean underwater deposits with ancient sculptures as possible. This method, used also by Parker (1992, pp. 3-4), increased the chances of comprehending the overall extent of the available archaeological record, as well as the actual geographical and chronological span of this activity. However, as Parker (1992, p. 4) acknowledged, this large corpus of data that includes non-excavated and largely unpublished information raises some implications and constraints for the interpretation of the recorded material. Therefore, due to the varied levels of reliability of the sources, as well as the inconsistent circumstances of discovery of the recorded underwater deposits, the list of finds of sculptures from underwater archaeological contexts in the Mediterranean should not be considered definitive (Table 1). New underwater sculptural discoveries or previously unknown cases from museums and archaeological services of the Mediterranean and around the world might be added in the future. Moreover, re-assessments of already recorded underwater depositions of sculptures could provide adjusted information and improved data for specific entries of the database.
After the creation of this database that includes 110 entries, and with the awareness of its limitations, a macro-scale research approach was employed. The information of each database entry was studied separately and then all data were classified, quantified and analysed comparatively. Through this process, it was possible to record the overall geographical and  chronological extent of the available archaeological evidence. Furthermore, this macro-scale investigation, using maps, graphs and comparative analysis of various archaeological sites enabled the identification of similarities and differences between the various database entries, which resulted in the detection of specific patterns of transport over space and time, presented below.

Geographical Distribution
Sculptures of different materials have been found all around the Mediterranean Sea, from the coast of the Iberian Peninsula to the Black Sea, Asia Minor and the Levantine coast ( Figure 2). 2 This geographical distribution matches the distribution of generally ancient Mediterranean shipwrecks recorded previously by researchers, such as Parker (1992, p. 548) and Wilson (2011, pp. 33-60). This similarity in the distribution of data constitutes the first indication that the maritime transport of sculptures was not an isolated activity with a regional focus, as it was previously thought. On the contrary, it seems to have been geographically incorporated within wider Mediterranean maritime networks of transport and trade of products in several different periods of Antiquity. The interest of previous scholars on underwater sculptural material from specific regions (Arata, 2005;Tzalas, 2007;Koutsouflakis, 2017), as well as the constant reproduction of older hypothetical theories regarding the transport of Greek sculptures from the eastern Mediterranean to Italy by the Romans (Boardman, 1985, p. 53;Stewart, 1990, pp. 228-229), had created the impression of a localised phenomenon and an one-way maritime transport from east to west, starting from Greece, Asia Minor or the Levantine coast heading toward southern Italy or specifically Rome. However, the data reveal that this activity was far more complex geographically than previously considered, and it must have involved multiple routes and directions around the Mediterranean world.
Another detail that is significant to note is that the higher or lower density of underwater evidence detected in specific areas of the Mediterranean does not necessarily indicate places where ancient cargo ships with sculptures were moving more or less intensely. Similar to Parker's (1992, pp. 6-7) shipwreck catalogue, and the updated shipwreck database by Strauss (2013) that was part of the Oxford Roman Economy Project (2019), the view of the geographical distribution of underwater deposits with sculptural finds in the Mediterranean is subject to a series of distorting factors, caused mainly by the availability of academic, archaeological and financial resources in various regions. These results are affected by the modern political systems in each country and the availability of provisions for the documentation and preservation of underwater cultural heritage. Therefore, the areas of the Mediterranean with no underwater sculptural finds, as seen for example, along the coast of North Africa, does not necessarily indicate that the maritime transport of sculptures in Antiquity did not take place there. The case of the Mahdia shipwreck (D.B.67) in Tunisia, for instance, could be considered representative of the currently invisible archaeological reality in the area rather than an exceptional or unusual find.
In the same way, the higher density of underwater sculptural finds in some areas like Greece, Italy and southern France should not be interpreted solely as a result of more intense maritime transport of sculptures in these regions. Some historical circumstances, geomorphological features and weather conditions could have certainly influenced the wrecking of ships in some areas more than others. However, the accumulation of more evidence in specific regions must be considered as a result of modern-day factors, such as large-scale fishing development, recreational underwater activities and underwater archaeological research, all being conducted more in some Mediterranean countries than in others.
Despite these distorting factors, as long as the above issues are taken into consideration and the density of finds is considered to be indicative of a larger archaeological record not yet fully explored, the geographical evidence recorded in the database still has high validity, indicating that the maritime transport of sculptures was not exclusively restricted between geographical areas.

Chronological Distribution
In order to establish the chronological extent in which this maritime activity took place, it was necessary to establish the date of the sites at which sculptures were deposited, rather than the date at which they were produced. Unfortunately, though, due to the fragmentary archaeological record and the lack of direct scholarly research at most of these sites, the date is not easy to reconstruct.
For approximately 64 database entries, more than 58% of the data, it is impossible to assign a date for the deposit. This is due to the sculptures being isolated finds and their recovery from undated and nonarchaeologically surveyed contexts. However, for the remaining 42%, archaeological information has made it possible to assign dates to the deposits. These dates span from the 7th/6th century BC to the 7th century AD (Figure 3), making it clear that the maritime transport and underwater deposition of sculptures covered a broad chronology. 3 In this distribution it becomes obvious that there is an increase of archaeological evidence from the 2nd century BC to the 2nd century AD, with spikes in the 1st century BC and the 2nd century AD. The 5th and 4th centuries BC, as well as the period between the 3rd and 5th centuries AD, show relatively high frequency, too. This chronological distribution is confirmed also in Figure 4, which represents visually the attributed dating of the underwater deposits with known, or at least partially researched, archaeological contexts.
This chronological range resembles also the general chronological distributions of ancient Mediterranean shipwrecks observed by Parker (1992, pp. 8-9;2008, p. 187) and Wilson (2011, pp. 33-39), only with a slight drop in the 1st century AD. The similarity of these data to the chronological patterns of the wider ancient Mediterranean shipwreck evidence provides a second indication that the shipping of sculptures had been largely incorporated to the wider, already known, maritime trading networks of Antiquity.

Types of Sculptures Transported
The sculptures included in the database vary in material, size, date of construction and subject of representation. Analysis of these data can assist in understanding why sculptures were transported by ship in Antiquity.

Material
The underwater archaeological record preserves evidence for all materials known to have been used for sculptural production during Antiquity (Table 1, Figure 5). The majority of recorded archaeological deposits, 67, have solely bronze sculptures. Twenty possess only stone sculptures, seven with only terracotta, six have both bronze and stone sculptures, four have only wooden or ivory sculptures, one has both terracotta and wooden sculptures, and one has bronze and terracotta sculptures. Finally, there are four sites for which the available sources do not report the material of the sculptures.
The predominance of sites with solely bronze sculptures is very notable. However, similar to the geographical distribution and chronology of the documented underwater deposits, this number can illustrate simple coincidence. Since most of the recorded underwater deposits are simply sculptures retrieved as accidental finds, with no secure recorded contextual information, it is easy to understand that bronze as a material is far easier to be caught and lifted from the seabed by a fisherman, sponge diver or scuba diver. This is due to bronze being relatively light in comparison to stone, which is heavier (causing fishermen's nets to break) and more prone to degradation by marine organisms and salt water (Ricci et al., 2019). Additionally, the distinct colour and texture of bronze makes sculptures of this material easier to detect on the seabed, as well as better preserved in comparison to terracotta, wood or ivory, which become more easily coated, disintegrated or absorbed in the marine environment (Ricca et al., 2021). Hence, the larger number of deposits with bronze sculptures does not necessarily imply that more bronze sculptures were carried on ships sailing the ancient Mediterranean.

Size
Since it was not possible to examine in person every single sculpture from the sites listed in Table 1 due to access constraints, it has been hard to determine the exact size of many of the sculptures, except for only a few cases where exact dimensions were provided. Thus, it has only been possible to classify the finds into the broad categories of small-, mediumand large-scale sculptures.
From the 110 underwater archaeological deposits recorded, 26 have sculptures of unknown size, while 11 are reported with under life-size sculptures, referring probably to either small-or medium-scale sculptures ( Figure 6). Of the rest are 16 deposits with definite small-scale sculptures, ten with medium-scale and 39 with large-scale sculptures, but also eight with sculptures of multiple sizes reported.

Dating
The date of when the sculptures were produced is another detail not easily understood from the available sources unless individual sculptures have been studied art historically and dated stylistically. However, from the overall information provided, it is evident that these sculptures date from the Archaic to the Late Roman Periods. It is important to clarify, though, that typological dating cannot be used for the dating of underwater archaeological contexts in which the sculptures were found. That is because the date of sculptural production does not always coincide with the date of the object's transport and deposition under water.
For example, the Antikythera shipwreck (D.B.6) in Greece wrecked in the first half of the 1st century BC while carrying a large cargo of both bronze and marble sculptures, which were constructed in different periods between the 4th and 1st centuries BC (Tzalas, 2007, pp. 342-363;Vlachogianni, 2012, pp. 62-115). Moreover, the Megadim shipwreck (D.B.75), in Israel, which has been dated by its coins to approximately 100/99 BC, carried bronze sculptural fragments with a production date in the 4th and 3rd centuries BC (Syon et al., 2013, pp. 2-4).
The distinction between the date of the sculptures and the date of the transport and underwater deposition has been observed on several occasions and it signifies once more the importance of contexts in the study of sculptures from underwater sites. Additionally, the variation in the dating of the sculptures in comparison to their underwater deposition has been a significant clue for the detection of possible reasons and patterns of the maritime transport of sculptures.

Themes and Subjects Represented
The available sources for the database are not always very thorough in recognizing the represented subjects of the sculptures. Additionally, there are many cases in which the seriously defaced condition of the freestanding sculptures makes the represented theme impossible to identify. However, the overall database indicates what was previously thought: namely that in Antiquity, ships carried all known freestanding sculptural types produced in different periods and regions of the ancient Greek and Roman world (Boardman, 1985;Stewart, 1990;Smith, 1991). These types could be: freestanding anthropomorphic figures (mythical or mortal), such as the dancing satyr of the Mazara del Vallo 1998 sculpture (D.B.73) found in Italy, or the female figure of the Rhodes sculpture (D.B.91) found in Greece; zoomorphic figures (animals and mythical/imaginary creatures), such as the horses from the Antikythera shipwreck (D.B.6); sculptural groups, with combinations of the above, like the sculptural group of the Horse and Jockey from Artemission (D.B.9) in Greece; freestanding sculptural reliefs, such as those retrieved from the Piraeus shipwreck (D.B.83) in Greece.
Moreover, the wide variety in the levels of finish of the transported sculptural products has been very noticeable. There are examples of roughly cut sculptural pieces, such as the colossal statue of a cuirassed emperor retrieved from the Şile shipwreck (D.B.98) in Türkiye, finished sculptural works, such as the statues from the Riace assemblage (D.B.93) in Italy, as well as sculptures that were most probably carried by sea after having been damaged, such as the headless stone statues from the Lixouri shipwreck (D.B.63) in Greece.
A very interesting observation is the identification of copies or reproductions of earlier original sculptural works. The phenomenon of copying and reproducing versions of well-known sculptures made by renowned artists, mainly from the Classical and Hellenistic Periods, was introduced during the late 3rd and early 2nd centuries BC by the Hellenistic kings of the eastern Mediterranean, who wished to collect famous works of art (Stewart, 1990, p. 63;Smith, 1991, pp. 14-16;Ridgway, 1984;. This activity was adopted and developed further during the Roman Period resulting in large numbers of similar sculptural subjects (Ridgway, 1984;Gazda, 2002;Anguissola, 2018).

Reasons and Patterns of Maritime Transport of Sculptures
The macro-scale research with the classification, quantification and careful recording of the data, described above, provided a good overview of the available evidence for the maritime transport of sculptures in the ancient Mediterranean. Additionally, this study made it possible to understand the function of sculptures at several of the sites and therefore detect reasons and patterns of maritime transport of sculptures during Antiquity (Figure 7). For this identification, only sites classified as 'shipwrecks', 41 in number, and some 'assemblages' with contextual archaeological information, 11 overall, have been used. The 58 sculptures found as single sculptures (Figure 1), namely as isolated finds without secure archaeological contexts, have not been included in this part of the study because they cannot provide any transport data.

Shipboard Items and Personal Belongings
A possible pattern is the transport of small-or medium-scale sculptures of different materials as shipboard items or personal belongings. This type of transport, which is not a trading activity, has been recognised in well-surveyed or excavated shipwrecks, providing a known archaeological context for the sculptures and the other discovered artefacts. Most sites that belong to this category are shipwrecks with identifiable, non-sculptural, cargo, that include a small number of under life-size sculptures in a single or multiple materials, dated as contemporary or slightly older than the date of the underwater deposit. The chronological range extends throughout the entire period surveyed in this study (Figure 7) and must have been part of the habit of ancient mariners carrying personal figurines, sometimes protective, during their journeys (Hanfmann, 1962;Brody, 2008;Galili & Baruch, 2015). This type of transport seems to have taken place beginning much earlier.
The small figurine of a female deity retrieved from the Late Bronze-Age Uluburun shipwreck is one of the earliest preserved examples of this pattern of sculptural transport (Pulak, 1998, p. 207, fig. 20). Despite the lack of direct underwater archaeological evidence after the 2nd century AD, the transportation of under life-size sculptural artefacts as shipboard items or personal belongings that sailors individually carried with them must have continued in Late Antiquity, given the widespread use of small-scale statuary in the private sphere, known from terrestrial finds, during the Roman Period (Trego, 2004;Madigan, 2013, pp. 1-38;Stoner, 2015;Papantoniou et al., 2019).
For the identification of this pattern, it has been vital to know the number, size and type of the recovered sculptures, the main cargo of the ship, but also the location of discovery within the wider archaeological context of the shipwreck site. The   , 1985, p. 563;Parker, 1992, pp. 249-250). Similarly, the late 2nd-or early 1st-century BC Pozzino shipwreck with its cargo of amphorae and pottery included fragments of a small-scale wooden or ivory-and-wood sculpture representing probably the god Asclepios, which was found within a wooden chest with medical containers (Yellowlees-Bound & Bound, 1990, p. 255;Gibbins, 1991, p. 241). This sculptural piece has been interpreted as possibly belonging to a passenger, maybe a medical practitioner, travelling on board at the time of wrecking (Spawforth, 1990, p. 9;Gibbins, 1997, p. 2).

Pomey
Additional examples that could belong to this category are the Camarina 1989 shipwreck (D.B.20)

Trade of Stone Sculptures as Part of Stone Cargoes
This maritime movement of stone sculptures for trading purposes has already been identified previously and is considered as part of the wider stone trade that took place during the Roman Period (e.g. Beykan, 1988;Claridge, 1988;Bartoli, 2008;Castagnino Berlinghieri & Paribeni, 2011;Russell, 2011Russell, , 2013aRussell, , 2013bRussell, , 2015.
There are six identified shipwrecks and one assemblage with stone cargoes that include stone sculptures.  (D.B.79) in Greece could be included in this category. These deposits all preserve evidence for stone blocks, architectural parts, and freestanding stone sculptures, transported altogether. The dating of the stone artefacts, and of the stone sculptures in particular, are usually contemporary to the date of the wreck but they appear to be in different levels of finish (Anguissola, 2018, p. 118). The sites range in date from the 1st century BC to the 3rd century AD (Figure 7), a period which coincides with the large development of the Roman stone trade, as presented and analysed previously by Russell (2011Russell ( , 2013aRussell ( , 2013bRussell ( , 2015. Some particularly interesting deposits are the 2ndor 3rd-century AD Punta Scifo shipwreck, found off Italy in the Ionian Sea (Bartoli, 2008, p. 47, 128-131, 261-262) and the early 2nd-century AD Şile shipwreck, found off the Black Sea coast of Türkiye (Beykan, 1988;Parker, 1992, p. 361;Russell, 2013a, fig. 8.4). Both shipwrecks seem to have been involved in the transport of stone artefacts quarried from the island of Proconnesus in the Sea of Marmara. These two sites suggest the existence of two different routes in the shipping and distribution of Proconnesian stone artefacts and sculptures during the 2nd and 3rd centuries AD. Moreover, the distinct levels of finish seen between the smoothly-finished Punta Scifo sculptural group and the roughly-cut Şile portrait sculptures suggest a variation in the carving requirements of the shipped sculptures or the use of different craftsmanship and techniques, possibly based on the shipping circumstances, the type of the order, the ship or the packaging used (Claridge, 1988, pp. 148-151;Anguissola, 2018, pp. 119-125).

Trade of Sculptures with Other Luxury Objects
A third type of sculptural transport are the wrecks of ship that were carrying large numbers of both bronze and stone sculptures of multiple sizes together with other luxury objects (Bouyia, 2012b, pp. 287-292). The Antikythera (D.B.6) and Mahdia (D.B.67) shipwrecks are the two most notable examples. These shipwrecks, dated to the late 2nd and early 1st century BC, transported both bronze and marble sculptures of different sizes, from small-scale to over life-size, but also a variety of other high-quality luxury, objects such as domestic furnishings, architectural parts, fineware pottery, glassware, large decorative vessels (Hellenkemper Salies et al., 1994;Kaltsas et al., 2012, pp. 14-15, 36;Anguissola, 2018, pp. 116-117).
Similar but less studied sites that could fall into this category are the Apollonia shipwreck (D.B.7) in Libya, the Styra shipwreck (D.B.101) in Greece and the Torre Flavia shipwreck (D.B.105) in Italy. The Apollonia shipwreck (D.B.7) dated approximately to the 2nd century BC was found in close proximity to the harbour of Apollonia, containing a collection of bronze furniture and sculptures (Parker, 1992, p. 57). The Styra shipwreck (D.B.101), excavated recently by the Greek Ephorate of Underwater Antiquities, included several fragments of life-size bronze statues found together with bronze furniture pieces and the main cargo of amphorae and tableware and has been dated in the Late Hellenistic Period (Koutsouflakis, 2017). The Torre Flavia shipwreck (D.B.105) was discovered including several architectural members as well as two marble sculptures and one bronze (Parker, 1992, p. 427).
Ships carrying such valuable cargo (Bouyia, 2012a, pp. 36-49) are considered to have been very large in size and they have been associated with the sculptural transport conducted for the needs of the art collection market of wealthy elites during the Late Hellenistic, Late Republican and Early Roman Imperial Periods (Bartman, 1994, pp. 71-88;Anguissola, 2018, p. 55, 121-122). This was the maritime activity alluded to in the orders that Cicero describes in his letter to Atticus (Cicero, Letters to Atticus 1.8.2). This has also been the interpretation that most scholars have suggested as a general explanation of the underwater deposition of ancient sculptures in the Mediterranean. However, as this study illustrates, trade or simply shipping of sculptures with other luxury objects is not the only reason for the maritime transport of sculptures in the ancient Mediterranean. On the contrary, this type of transport is only archaeologically documented in the last two centuries BC (Figure 7).

Trade of Bronze Sculptures and Other Metals as Scrap
Another reason for the maritime transport of sculptures, identified in several shipwrecks, is seen in mixed cargoes of metal objects including a variety of bronze sculptural artefacts.  (Fernandez-Miranda & Rodero-Riaza, 1985, pp. 175-188;Misch-Brandl & Galili, 1985, pp. 12-13;Parker, 1992, p. 62, 176;Mattusch, 1997, pp. 13-14;Arata, 2005, pp. 143-144;Brokalakis, 2016). The sculptures discovered in these sites are almost always considerably older in date than the period contemporary with the shipwreck and usually preserved in a fragmentary condition, since they had fallen into disrepair prior to their maritime movement.
The chronology of this type of transport ranges from the Late Hellenistic Period to Late Antiquity (Figure 7). This maritime activity spreads from the Balearic Islands to the Levantine coast. The similarities in the material found in the above-mentioned sites, despite their chronological and geographical variation, indicate possibly the existence of a consistent Mediterranean-wide maritime pattern for the transport of bronze sculptures in disuse.
Generally, the transport of scrap metal objects by ship has been recorded in the archaeological record and literary sources (Mundel Mango, 2001, pp. 95-102). However, pieces of sculpture from underwater contexts have not been widely incorporated in the study of scrap metal transport. Only McCormick (2001, p. 51), was one of the first to identify the Favaritx shipwreck (D.B.38) as part of this wider Late Roman network for metal extraction, production and distribution.
The identification of this type of sculptural transport is very significant because the archaeological re-examination of many underwater deposits with fragmentary bronze sculptures, which are not yet thoroughly researched, could also belong to this category of scrap metal transported by ship, rather than other types. For example, the bronze statues of the Artemission assemblage (D.B.9) in Greece that were retrieved from the seabed in a very fragmentary condition (Bass, 1966, pp. 72-73;Hemingway, 2004) or the bronze sculptural fragment of an elephant seen in the Mazara del Vallo 1999 sculpture (D.B.74) in Sicily, Italy, analysed most recently by Lapatin (2018, pp. 166-167), were all found without secure contextual data and could have been transported as scrap, too. In the future, a systematic study of additional underwater deposits with all their contextual information will hopefully confirm this hypothesis.
Despite the uniformity of data that these sites present, due to the poor recording of their underwater archaeological contexts and the lack of extensive publication of the archaeological material, it has not been possible to document accurately the number, condition, size and form of the relevant terracotta sculptural artefacts. Thus, it has been unclear why the terracotta sculptures were transported by sea. The possibilities of trade or other religious purposes have been indicated. Unfortunately, though, due to lack of accurate evidence no further conclusions have been produced for the deposits with solely terracotta sculptures at this stage.

Unclear Transport Purposes
Some well-identified shipwrecks with preserved archaeological contexts do not clearly reveal a reason for their transport. For example, for the case of the Porticello shipwreck (D.B.88) in Italy, the regular size of the merchant ship, the distinct chronology in the late 5th or early 4th century BC, a period when there are no comparative sites, and its geographical location in the Straits of Messina, have not permitted the straightforward identification of a specific transport type solely through the macro-scale methodological approach of this research.

Conclusion
Through the documentation, quantification, classification and comparative analysis of available data from already known underwater deposits with sculptures around the Mediterranean it has been possible to first map successfully and present the geographical distribution of underwater evidence for the maritime transport of sculptures in Antiquity. From that process, it has become clear that sculptures of different materials have been found throughout the Mediterranean, from the coasts of the Iberian Peninsula to the Black Sea, Asia Minor and the Levant, making thus the maritime transport of sculptures a Mediterranean-wide activity and a not regionally focused phenomenon, as previously understood.
Secondly, it has been possible to understand the chronological range of the existing underwater archaeological record with ancient Greek and Roman sculptural artefacts. Thus, it has been proven that ancient sculptures of various known types, dates and representations had been transported on ships in all areas of the Mediterranean and throughout Antiquity, from the 7th or 6th century BC to the 7th century AD.
Moreover, through this macro-scale research, it has been possible to provide evidence for different types of maritime transport of sculptures in the ancient Mediterranean. Thus, it has been identified that sculptures were transported by sea as shipboard items and personal belongings, for trading purposes as part of stone cargoes, or as a cargo of luxury objects, as scrap metal with the intention to be recycled, as part of terracotta sculpture assemblages with a ritual or trading purpose but also for other unidentifiable purposes.
The present study brings together and collectively examines all available underwater archaeological evidence for the maritime transport of sculptures in the ancient Mediterranean. Thus, it has been possible to answer the question of where, when and why sculptures were transported in the ancient Mediterranean. Overall, the preserved underwater archaeological record shows clearly that the maritime transport of sculptures was more complex than anticipated by scholars in the past. This maritime activity was not rare or unique. The sculptures lost in the waters of the Mediterranean were not necessarily a special cargo transported alone, nor only during distinct periods and in specific regions of the Mediterranean. On the contrary, the present research provides solid data that suggest that the maritime transport of sculptures was very common and widespread, both chronologically and geographically. The sculptures seem to have been transported by sea for various reasons as mixed cargo, on a regular mercantile basis and within the generally established trading activities and maritime networks of each period. Thus, the maritime transport of sculptures should be considered as a common maritime transport activity, part of the already known commerce and connectivity of Antiquity. This study, with its new research perspective, refutes the traditional scholarly interpretation that considers the maritime transport of sculptures to have been conducted solely during the Late Hellenistic and Roman Periods, from solely east to west and as a result of looting or art collection activities.
By providing these new insights, this study highlights reasonable archaeological observations and improved understanding regarding the maritime transport of sculptures in the ancient Mediterranean. Additionally, the importance of underwater archaeological contexts, as well as the benefits of archaeological revision and re-examination of reports, data and older scholarship, even without direct access to the primary information due to loss or already disturbed archaeological deposits, is stressed.
However, given the limitation of the database creation and the macro-scale research, this study should be considered only as a first step in a field that still has endless possibilities for further study. It is hoped that the present re-evaluation of archaeological evidence will motivate scholars of both Classical and maritime archaeology to reconsider the approach towards Mediterranean underwater deposits with sculptures in order to improve the preservation and interpretation of this rich archaeological dataset. Notes 1. The present paper stems from a doctoral dissertation with the title 'The Maritime Transport of Sculptures in the Ancient Mediterranean' (Velentza, 2020) completed at the University of Southampton. 2. The geographical position of archaeological deposit is not exact. Due to the lack of research and publication of most of the underwater sites, approximate coordinates have been used, based on information provided by Parker (1992), or established by the author according to the descriptions provided in other academic sources. 3. This does not mean that sculpture transport did not happen at other periods. See for example, the Late Bronze Age 'Uluburun shipwreck', off the coast of Türkiye (Pulak, 1998, p. 207) and post-Antique evidence for the maritime transport and underwater deposition of ancient sculptures of various types and dates during the medieval times and the period of the 'Grand Tour' (Johnson, 1925, p. 22;Arata, 2005, pp. 5-6, 148-149;Velentza, 2020).

Disclosure Statement
No potential conflict of interest was reported by the author(s).

Declaration
No potential conflict of interest was reported by the author.