This article investigates the nature of usage, as well as the geographical origin, of a small group of ivory artifacts recently discovered in the earliest exposed cultural depositions at the Early Islamic (650–1100 C.E.) port of Aylah (Aqaba, Jordan). In addition to explaining the finds and the significance of their context for interpreting possible historical implications, the article uses a range of techniques to learn more about the raw material. In combining archaeological, visual, and biomolecular analyses on these ivories, fresh perspectives are provided that shed new light on the infrastructure and geographical scope of late antique and early medieval trade systems. Moreover, it informs us about the economic and commercial roles played by Red Sea ports in this period and highlights the potential of analyzing organic artifacts from sites in the region to reveal new details and characteristics of historical Indian Ocean trade networks.

The modern port of Aqaba (Jordan) has been an important hub in trade and transhipment practices for millennia. The 10th-century geographer, al-Muqaddasi, referred to it as the “port of Palestine on the edge of the China Sea,”1 reflecting a world-view in which continents were bound together by water rather than land (Fig. 1). The history of Aqaba as a port and emporium of trade nevertheless stretches much further back. From the end of the 1st millennium B.C.E. at least, we have solid archaeological and historical evidence to suggest that the settlement here served as one of the nodal interfaces between land-based and maritime exchange systems. The most famous periods are nonetheless those in which trade networks reached far beyond the Red Sea, in particular the late Roman/Byzantine (3rd – 6th centuries C.E.) and early Islamic (7th – 10th centuries C.E.) periods.

FIGURE 1

Map of Aylah and regional connections, made with Natural Earth through Global DigiMap.

FIGURE 1

Map of Aylah and regional connections, made with Natural Earth through Global DigiMap.

Roman and early medieval trade with Asia is a topic that has generated significant academic interest for decades, and consequently these periods have seen dedicated archaeological fieldwork from which we are now beginning to reap concrete results. In Aqaba, American teams have excavated extensive occupational phases from the periods in question, notably under the Oriental Institute excavations of the medieval Islamic town of Aylah (1986–1995),2 and later under the Roman Aqaba Project (1994–2003),3 which explored the Roman townscape (referred to similarly as Aila) through a range of surveys and focused excavations. Most recently, the Danish Aylah Archaeological Project (AAP; 2010–2016) returned to the early Islamic town, dedicating efforts to the untouched southwest quadrant based on a hypothesis that this was an important commercial district.4 These projects have provided important insights not only regarding the specifics of Aqaba's mercantile profile, but also into first millennium proto-global trade systems.

A major step was the preliminary presentation of the imported ceramic corpus.5 Perhaps of even greater influence has been the identification and recent archaeometric analysis of the carrot-shaped Aila amphora, which is one of the most widely circulated vessels of Late Antiquity6 and perhaps the town's most famous export. Aqaba amphorae have tentatively been identified at coastal sites throughout the Red Sea and Indian Ocean. Further light must be shed on the nature and extent of commercial connections. For now, we are left with a sense of Aylah's connectivity that is difficult to anchor in empirical evidence.

This article begins to address this lack of empirical evidence by exploring in detail a small assemblage of ivory artifacts excavated by the AAP in 2010. Along with contextual archaeological information, a range of techniques were applied to learn more about the raw material, state of preservation, and source region of the ivory. This included the use of non-invasive methods such as Fourier transform infrared spectroscopy (FTIR) and scanning electron microscopy (SEM), as well as biomolecular methods such as collagen peptide mass fingerprinting (ZooMS), stable light isotope analysis (δ13C and δ15N), and ancient DNA analysis. Some of the artifacts were directly radiocarbon dated by accelerator mass spectrometry (AMS). The results of the laboratory tests were then analysed within their archaeological context to achieve as complex and composite an understanding of these artifacts as possible.

NOTES ON THE IVORY TRADE

Ivory was a key commodity that was traded throughout the late antique and early medieval periods, especially around the Indian Ocean rim. The geographical catchment areas for ivory were as varied as the types of ivory available, which included elephant, hippo, and warthog from Africa, elephant ivory from India, and dugong ivory from the Indian Ocean.7 From historical sources such as Pliny's Natural History and the anonymous Periplus of the Erythraean Sea (both written in the 1st century C.E.), it is clear that elephant ivory in particular was a treasured and important commodity that was transported across impressive distances (from East Africa and Gujarat to the Mediterranean, e.g.).8 At the Egyptian Red Sea port of Berenike, excavated fragments of molars and tusks are evidence that elephants were transported through this port in Ptolemaic times, though this probably reflects the State's importation of the animals for military purposes rather than a vibrant ivory trade.9 Excavations at Aksum (Ethiopia) have unearthed multiple deposits of ivory across the site. In the Tomb of the Brick Arches, dated to the late 4th century C.E., over 400 pieces of worked ivory fragments were found, including pieces of a throne made of elephant ivory, carved panels, boxes, and a carved female figurine.10 Further evidence of ivory craftsmanship in Aksum was found in workshop deposits from the urban occupation sites dated from the late 5th to the first half of the 7th century C.E.11 In the 10th century, elephant ivory carving flourished across the Mediterranean, and it has been suggested that the raw ivory was being traded from Africa using land routes through the Sahara and coastal cabotage up the Swahili coast.12 At the same time, demand for elephant ivory from consumers in China and India soared, and tusks traveled great distances across the Indian Ocean.13 By the Islamic Middle Ages (1000–1400 C.E.), historical accounts are increasingly supplemented by archaeological discoveries and objets d'art from ivory carving centres such as al-Andalus, Egypt, and the Ottonian Empire.14 

Early Islamic trade networks operated efficiently in northern and eastern Africa, and ivory was but one of many bulk commodities moving out of Africa and into the rest of the Islamic world.15 From the 7th century, the successful expansion of Islam facilitated the establishment of a new and enfranchised political elite with the resources to consume the most valuable commodities that the world had to offer. Documentary and artifactual evidence links emporia in the Middle East to counterparts along the eastern African coast. A range of early medieval Swahili trading sites have been investigated both archaeologically and historically, revealing close ties to communities on the Arabian Peninsula and in the Red Sea region.16 Imported glass beads and ceramics appear on sites along the eastern African coast as well as sites in the interior of southern Africa from the late 7th to the early 8th century C.E.17 At inland centers in southern Africa, there is evidence of ivory craftsmanship and thriving trade networks from the 7th century onwards.18 Presumably, ivory was a commodity that was traded all around the Indian Ocean rim,19 but it is tempting to envision a scenario in which significant quantities of eastern or southern African ivory was shipped to emporia in the Red Sea, and from here re-distributed in a different form to consumers in the Middle East and Mediterranean.

Being both a maritime corridor between the Mediterranean and Indian Ocean, as well as the locus of the Islamic empire's first successful maritime ventures, the Red Sea was an integral link and crossroads for valuable commodities that needed to traverse great distances before finding their end-market. The way commodities moved varied and changed depending on the route and who facilitated transport. In most cases, the commodities that were produced from a range of valuable raw materials ended up in large cities such as Alexandria, Antioch, and Rome. Later, demand shifted to cities like Cairo, Damascus, and Constantinople, but the pull-factor of large cities remained the most powerful driving force. Thus, we learn from Nasir-i Khusraw, an 11th century Persian traveller, that large numbers of Zanzibari tusks were sold in the market in Cairo in 1047 C.E.20 

While of course conditioned by historical specifics and shifting political landscapes, the juxtaposition of urban consumers and distant catchment areas created a risky but potentially very lucrative domain for entrepreneurial forces to occupy, and occupy it they did. However, in order for such a domain to work efficiently, reliable systems of coastal mobility were necessary and intermediate transactions and transformations were common. The stakeholders and agents of such transformative processes were often based in the coastal emporia, providing an initial explanatory framework for the number of new coastal emporia established during the early Islamic period.21 Here, goods were transhipped, combined or separated, processed, and sometimes re-worked to increase value and saleability at their final destination.

These intermediary steps often required specific knowledge or skillsets, which in turn may have caused certain ports to enhance certain specializations,22 drawing merchants working in those particular commodities.23 Such specializations were of course partly contingent on availability and catchment, and need not have focused on a single commodity. The discovery of the ivory fragments in question at early Islamic Aylah not only confirms that processing raw materials appears to have been an integral part of commercial networks, but that processing need not have been in close geographic proximity to the area where the raw material was originally sourced.

THE AYLAH IVORIES: DESCRIPTION AND TECHNOLOGY

The research presented in this article is based on the discovery of 15 fragments of a material believed to be ivory (Table 1). Once analyzed, 13 of these fragments were confirmed as ivory (Fig. 2), while two fragments were identified as camelid bone. Aesthetically and functionally, both fragments of bone fit within a late antique or Coptic tradition of ivory craftsmanship that extends well into the Islamic Middle Ages (ca. 6th to 10th century C.E.). The first, a carved panel fragment (AAP169), has ample parallels in museum and private collections and is also relatively common in excavations.24 The spoon (AAP319), on the other hand, is consistent with bone and ivory paraphernalia seen in late antique assemblages at sites like Manzalah, Hama, and Fustat.25 

TABLE 1

Detailed data of the Aylah ivory and bone fragments.

Project #LOCUSDESCRIPTIONDimensionsVisual IDZooMS IDδ13C(‰)δ15N(‰)IR/SFC/P14C DATECalendar date C.E.
 105 carved piece, part of pixus 28 × 9 × 5 mm; 1.7 g  elephant −17.6 12.8   1975 ± 35 BP 16 ±38 
0034 106 carved piece, part of pixus 35 × 11 × 6 mm; 2.5 g         
0022 106 large curved piece, part of pixus 43 × 27 × 10.5 mm; 9.1 g elephant ivory  −16.5 6.2   1635 ± 30 BP 439 ±52 
0029 106 curved piece, part of rim, incised edges, slight lip on interior 38 × 11 × 4 mm; 2.6 g elephant ivory  −20.5 4.7 3.7 0.19 1475 ± 35 BP 587 ±30 
0031 106 small ivory finger ring 4 × 19 mm; 0.5 g         
0028 103 large ivory finger ring 5 × 33 mm; 2.5 g  elephant       
0023 103 worked ivory cut outs 15.4 g total for all cut outs elephant ivory    3.5 0.27   
0023 103 worked ivory cut outs  elephant ivory  −18.2 6.8   1555 ± 30 BP 491 ±45 
0023 103 worked ivory cut outs  elephant ivory  −17.8 10.0   1475 ± 30 BP 588 ±27 
0023 103 worked ivory cut outs  elephant ivory  −18.1 10.1   1490 ± 30 BP 576 ±25 
0023 103 worked ivory cut outs  elephant ivory  −20.0 6.3   1350 ± 40 BP 676 ±30 
0023 103 worked ivory cut outs  elephant ivory elephant       
0304 IM2/22 worked ivory cut outs  elephant ivory elephant       
0319 IM3/49 bone spoon/scoop 73.5 × 16 × 4.5 mm bone camelidae   3.5 0.29   
0169 110 bone plaque, decorated with floral and vine motif 50 × 33 × 3 mm bone camelidae   3.6 0.25   
Project #LOCUSDESCRIPTIONDimensionsVisual IDZooMS IDδ13C(‰)δ15N(‰)IR/SFC/P14C DATECalendar date C.E.
 105 carved piece, part of pixus 28 × 9 × 5 mm; 1.7 g  elephant −17.6 12.8   1975 ± 35 BP 16 ±38 
0034 106 carved piece, part of pixus 35 × 11 × 6 mm; 2.5 g         
0022 106 large curved piece, part of pixus 43 × 27 × 10.5 mm; 9.1 g elephant ivory  −16.5 6.2   1635 ± 30 BP 439 ±52 
0029 106 curved piece, part of rim, incised edges, slight lip on interior 38 × 11 × 4 mm; 2.6 g elephant ivory  −20.5 4.7 3.7 0.19 1475 ± 35 BP 587 ±30 
0031 106 small ivory finger ring 4 × 19 mm; 0.5 g         
0028 103 large ivory finger ring 5 × 33 mm; 2.5 g  elephant       
0023 103 worked ivory cut outs 15.4 g total for all cut outs elephant ivory    3.5 0.27   
0023 103 worked ivory cut outs  elephant ivory  −18.2 6.8   1555 ± 30 BP 491 ±45 
0023 103 worked ivory cut outs  elephant ivory  −17.8 10.0   1475 ± 30 BP 588 ±27 
0023 103 worked ivory cut outs  elephant ivory  −18.1 10.1   1490 ± 30 BP 576 ±25 
0023 103 worked ivory cut outs  elephant ivory  −20.0 6.3   1350 ± 40 BP 676 ±30 
0023 103 worked ivory cut outs  elephant ivory elephant       
0304 IM2/22 worked ivory cut outs  elephant ivory elephant       
0319 IM3/49 bone spoon/scoop 73.5 × 16 × 4.5 mm bone camelidae   3.5 0.29   
0169 110 bone plaque, decorated with floral and vine motif 50 × 33 × 3 mm bone camelidae   3.6 0.25   
FIGURE 2

Ivory and bone fragments found at Aylah; images from left to right, top to bottom follow the order listed in Table 1. One block of the scale in each photo equals 1 cm. Copyright Aylah Archaeological Project.

FIGURE 2

Ivory and bone fragments found at Aylah; images from left to right, top to bottom follow the order listed in Table 1. One block of the scale in each photo equals 1 cm. Copyright Aylah Archaeological Project.

Finding ivory at Aylah is not unique. Fragments of a similar material had surfaced both in the Danish and American excavations, but in most cases these were found in contaminated Fatimid fills, and were not unequivocally identified as ivory (camel bone being the most common alternative). A more remarkable find was made in the late 1990s at the Byzantine/Umayyad site of Humaymah, some 60 kilometres north of Aqaba. Here, remnants of three furniture panels with ornately carved depictions of male figures in military dress were excavated and identified as elephant ivory, dated to the mid-8th c. C.E. (Fig. 3).26 The fragments of ivory found at Aylah, however, were not carved with the same level of artisanal competence, nor was aesthetics necessarily the highest aspiration in their making. Rather, these are discards from more utilitarian objects. This demonstrates that even costly raw materials, such as ivory, were worked to differing degrees of perfection, for different purposes, and intended for a quite variegated market.

FIGURE 3

The Humaymah ivory plaques, each approximately 30 x 10 x .03-.05 cm. See: Rebecca Foote, “An Abbasid Residence at al-Humayma,” in Byzantium and Islam. Age of Transition 7th -9th Century, ed. H.C. Evans and B. Ratliff (New Haven & London: The Metropolitan Museum of Art, 2012), 221-23. Photo courtesy of Rebecca Foote.

FIGURE 3

The Humaymah ivory plaques, each approximately 30 x 10 x .03-.05 cm. See: Rebecca Foote, “An Abbasid Residence at al-Humayma,” in Byzantium and Islam. Age of Transition 7th -9th Century, ed. H.C. Evans and B. Ratliff (New Haven & London: The Metropolitan Museum of Art, 2012), 221-23. Photo courtesy of Rebecca Foote.

Other than the Humaymah panels, the bulk of late antique and early Islamic ivory comes from Egypt.27 Ivory appears to have been prevalent in the cities of Lower Egypt—primarily Alexandria and its hinterland28—again suggesting little geographic correlation between catchment area and craft specialization.29 The dearth of excavated late antique/early Islamic urban space in Upper Egypt is also significant, as there is little evidence of how ivory was procured and crafted in these centers of trade. In general, there has been little concrete evidence of ivory craftsmanship outside the source regions and this is, in part, what makes the discovery at Aylah so important.

The Aylah ivories are all fragments, most of which were discarded deliberately. Table 1 provides an overview with information on each individual fragment, including their radiocarbon dates and information on the archaeological context from which the artifacts derived. The largest piece found (AAP 20) together with three other pieces (AAP 33, 34, 29) resemble fragments of a small pyxis (a small storage vessel, typically with a lid), the likes of which are known from Aksum, but also from North African and Iberian contexts. Two fragments (AAP 28 & 31) are rings of ivory, which could have been for personal ornamentation. The final seven fragments (AAP 23 & 304) are what we describe as cutouts or discards: fragments of ivory left over from cutting out circular objects such as buttons. These pieces come from the hollow beginning of the tusk, inside the mouth of the proboscidean. The hollow part of the tusk can be useful for making cylindrical objects such as pyxides, whereas blocks of ivory from the part of the tusk nearest the tip are necessary for making solid ivory objects, such as gaming pieces.30 It is possible that these discards are the result of making other objects, or are the pieces left over from solid sheets of tusk that might have come through the port of Aylah and shipped on to further trading ports. One of the ways we tried to investigate these pieces in more detail was by studying the remaining traces of their manufacture in order to understand how they had been crafted.

During the 2014 season of the AAP, materials technologist Bénédicte Khan (CNRS) studied our objects—in particular the cut-outs.31 Using a hand-held microscope, she focused on the micro-traces, or stigmata, that were left on the surface of the artifacts during their manufacture. Khan identified only a few stages of known manufacturing techniques on the objects in question, and these seem mostly to be the result of their original extraction from a larger piece. Traces of scraping were present on the front and back faces of the discards, but these patterns were in some places breached by traces of transversal sawing (Fig. 4). This suggests that larger sheets or panels of ivory were cut and smoothed into flat panels before the circular elements were extracted. Extraction appears to have been accomplished unifacially with the help of a wheel or a tool moved by a bow, such as a bow lathe or bow drill, as the plaques were transversally grooved in a very regular way. Interestingly, similar ivory cut-outs were found in craft-working deposits at Aksum. Phillipson suggests that the ivory waste fragments found at Aksum indicate the presence of skilled turners living in the town, working ivory for a variety of purposes.32 The cut outs found at Aksum are approximately the same size and shape as the cut outs found at Aylah. This indicates a shared practice for the creation of objects such as pyxides made from elephant ivory.

FIGURE 4

Schematic of possible extraction techniques used to work the Aylah ivory fragments and image of scraping tool mark found on Aylah ivory piece AAP 23. Schematic and image by Bénédicte Khan (see Bénédicte Khan, “L'exploitation artisanale des matières dures d'origine animale au Proche-Orient entre le IIIe s. av. J.-C. et le VIIe s. apr. J.-C. : une approche techno-économique” [PhD diss., Université Paris I Panthéon-Sorbonne, 2019]).

FIGURE 4

Schematic of possible extraction techniques used to work the Aylah ivory fragments and image of scraping tool mark found on Aylah ivory piece AAP 23. Schematic and image by Bénédicte Khan (see Bénédicte Khan, “L'exploitation artisanale des matières dures d'origine animale au Proche-Orient entre le IIIe s. av. J.-C. et le VIIe s. apr. J.-C. : une approche techno-économique” [PhD diss., Université Paris I Panthéon-Sorbonne, 2019]).

On some of the Aylah ivory pieces, there is also evidence of the cementum of the tusk still present. Cementum is the outer, enamel-like layer of the tusk. In most carved ivory, this cementum layer is carefully removed, as it is often ridged and has a different density and working property compared to the softer dentine below. Cutler, for example, describes how this outer layer was stripped with an adze in workshops across Europe around the 10th century C.E..33 Fragments with cementum present are thus often regarded as discards, such as the ivory processing waste with cementum found at the sites of KwaGanadaGanda, Wosi, and Ndondondwane (South Africa), dating from the 7–13th centuries C.E..34 What is interesting about the Aylah material is that the discards found on site are large and suggest that ivory—or at least the extremities of the tusk—was not always regarded as a luxury material, but could also be used for more utilitarian objects.

THE ARCHAEOLOGICAL CONTEXT OF THE AYLAH IVORIES

This article discusses thirteen fragments of ivory discovered in a deep probe that yielded a considerable amount of archaeological material and information. Not only do these contexts represent landmark political changes with the transition from Byzantine to Islamic rule in the 7th century, but they offer a glimpse into one of the earliest examples of a distinctly Islamic manifestation of urbanism.

Reaching the earliest occupation strata at Aylah is difficult due to the shallow water-table and significant number of overlying cultural deposits.35 The 7th-century strata from which the ivory fragments were recovered were identified ca. 8 meters below the current surface (Fig. 5). The excavated area was small (1 x 2 meters, later reduced to 1 x 1 meter), yet contained an impressive density of artifacts and ecofacts, such as charcoal and bone. The sequence in question began immediately below the foundation of the lowest section of walling exposed in this area of town (IM1/101) and was interpreted as a refuse dump or infilling due to its organic composition. Prior to reaching these strata, surfaces were increasingly waterlogged and discerning stratigraphic change became impossible. Consequently, excavation was continued in arbitrary levels between 200 and 500 millimeters in depth until safety concerns prompted us to halt excavation completely (IM1/112). The ivory artifacts, therefore, were discovered in different loci (IM1/103, 105, 106), but essentially form an assemblage from the same deposit. One clear stratigraphic interface did, however, distinguish the upper layers of this deposit (IM1/103, 104, 105, 106) which contained a high density of both organic (especially bones and charcoal) and artifactual remains, from the lower sequence of loci (IM1/108, 109, 112), which contained less organic material and only a small corpus of ceramics.

FIGURE 5

Stratigraphic complexity of dump layers. Copyright Aylah Archaeological Project.

FIGURE 5

Stratigraphic complexity of dump layers. Copyright Aylah Archaeological Project.

The uppermost parts of the dump (IM1/101) contained an assemblage of predominantly 7th-century transitional ceramics36 with some late Roman wares also present. Among the latter was a fragment of an African Red Slip plate (probably Hayes’ forms 82–8437) with concentric bands and a wild boar impression. The motif is not uncommon, but particularly evocative, since the boar was an emblem for the Legio X Fretensis stationed in Aqaba from the late 4th century. A charcoal sample from an unidentified species of wood was also retrieved from this context and sent for radiocarbon dating. The sample, dated with Archaeolabs in Lyon, provided a conventional age of 1590±45BP, suggesting a calibrated (95.4%) date between 382 and 576cal C.E. The ivory fragments were found in the strata below this level.

We were able to obtain radiocarbon dates for seven of the thirteen ivory fragments38. With the exception of one fragment that showed a considerably earlier date, all samples had calibrated dates between the 5th and 8th century C.E. (See Table 1). These seven dates from the ivory pieces give us a timeline of when the ivory was growing on the elephant, so these dates do not necessarily reflect the chronology of the craftsmanship. To establish a chronological context for this, the dated ivory pieces need to be further contextualised by the archaeological evidence found alongside.

Loci 102–108 of the probe contained rich organic deposits, yielding a large amount of animal bones, metal fragments, and ceramics. Three intact oil lamps and a smashed but largely intact bowl made of the local cream-surface ware were also found in the same layers as the ivory fragments. A number of highly deteriorated building stones (IM1/107), the origin of which we cannot ascertain at the moment, were also found in this deposit. However, the fact that they seem to have been dumped here purposefully—perhaps as part of the initial levelling fill—is suggested by a clear stratigraphic interface between the dump layers (IM1/101–108) and a sequence of sand layers with low artifact density and little variation between them (IM1/109, 112). Organic samples were taken on either side of this interface, in order to date the dump and ascertain the chronological distinction between the two sequences. In addition to the dated sample from the top of the dump (IM1/101), mentioned above, a charcoal sample collected from the lower level of the dump was dated at the radiocarbon labs of Waikato University, New Zealand. This came back with a calibrated date between 420 and 570cal C.E. The chronological proximity of the two samples, combined with the dates of the ivory fragments, suggests a relatively quick deposition process, while the section reveals a complex stratigraphic deposition (Fig. 5). In unison, the calibrated radiocarbon dates align well with the proposal that Aylah was constructed in the early 7th century. The paucity of any sample postdating the 8th century corroborates this notion and the samples generally provide a terminus post quem for their deposition. The dates, however, cannot be taken at face value for purposes of dating the craftsmanship and trade in the commodity. There is thus a wide range of dates from materials found in this deposit (from the 1–7th centuries C.E.), which is perhaps due to the fact that these raw materials (wood and ivory) could have been used and re-used over a long period of time.

The dump sequence yielded a plethora of evidence that included fragments of iron, several of which could be recognised as objects (a buckle, a bronze ring, a key, and a partial iron saw blade, e.g.); slate and mother of pearl inlay pieces; and four copper coins (discussed below). In addition to these, the retrieved pottery provided a significant spectrum of diagnostic sherds that in unison form a credible and verifiable ceramic horizon.

Some of the most diagnostic ceramics retrieved were lamps and lamp fragments. The bodies of at least three wheel-made lamps were found, of which two were executed in a reddish-orange fabric and remained almost completely intact. The ribbed bodies and extended nozzle are highly characteristic, and parallel examples have been identified and published from many sites in southern Palestine and Jordan, as well as in Lower Egypt.39 The identification of these lamps thus places their production securely from the 5th to 8th centuries, with production likely increasing in the later part of this time frame.40 That we are indeed talking about a later context here is corroborated by the presence of a large body sherd from a third wheel-made lamp, which has no ribbing and was executed in a cream fabric, suggesting a higher firing temperature and thus a 7th- to 8th-century technology.41 

The upper part of the dump (IM1/102) also contained an intact lamp, which has few parallels. The closest established lamp form is that of the so-called Samaritan 1 and Samaritan 3 lamps, both of which are characterised by small filler and wick holes, and rounded bodies with projecting nozzles. These have been found across Palestine, as well as in northern Jordan (Pella and Umm Qais, e.g.). Perhaps most evocative for us are the examples excavated at the port of Caesarea Maritima. “Samaritan lamps” are generally dated from the 4th to the early 6th century C.E.,42 but the body decoration of the Aylah example suggests that it belongs to the later part of that chronology.

The dump layers therefore have a 6th- to 7th-century ceramic horizon, albeit with a notable presence of 8th-century wares as well. The majority of the ceramics from the deep probe were red wares and cream-surface ware, but 8th century cream wares are also present. All of these wares are common throughout the occupation of Aylah, but the first two also pre-date the foundation of the Islamic town by several centuries. The substantial presence of cream wares nevertheless suggests that the few clearly late Roman ceramics—as well as the single Roman-period radiocarbon date–-must be considered secondary depositions that do not warrant dating these deposits prior to the 7th century. This notion is echoed in the forms identified in the ceramic assemblage, from which are diagnostic sherds consistent with the 7th-century horizon known from the Aqaba kilns.

The final category of evidence for the chronology of this deposit comes from four coins excavated in loci 103 (AAP6), 104 (AAP7 & 8), and 108 (AAP9). While the lowest deposited coin (AAP9) was unidentifiable due to heavy corrosion, the remaining three are Byzantine folles from the 6th century. These large and heavy copper alloy coins were a common and popular aspect of commercial life throughout the 6th century C.E., and continued to circulate well into the 7th century, potentially down to the major reform of the coinage under the Umayyad caliph ‘Abd al-Malik in the final years of the 7th century.43 The dating of these coins is extremely important, as it cements a 6th-century terminus post quem for the deposition of the upper dump layers, and thus presumably the dump in its entirety.

A brief note should be made on the strata below the dump, which consist of two loci (IM1/109, 112) that also constituted the lowest level reached within the town walls of Aylah. These strata were distinct from the dump sequence due to a change in soil composition, from largely organic to largely mineral. Locus 109 contained a nearly-intact Roman/Byzantine lamp and a ceramic horizon, which generally shared more traits with late Roman traditions than with the 7th century. The lamp is of the South Jordan type, which is usually dated from the late 4th century until the Umayyad period, though in our case both form and decoration suggests an early type.44 The potential pre-Islamic date of the sub-dump strata is further echoed by two radiocarbon-dated samples collected at the bottom of the trench: a mollusc yielded a calibrated date between 190 and 580cal C.E., whereas a charcoal sample from the same context gave us a calibrated date between 85 and 235cal C.E. Again, these dates cannot be taken at face value, but in unison with dump strata are highly indicative that the distinction between the dump and sub-dump layers reflects a 7th-century infilling, probably related to the original construction of the township.

A final validation of a 7th-century date for this strata is evidenced through the animal bones found. While the dump strata contained an abundance of animal bones that are still being analysed, the density of such materials dropped noticeably in locus 109. Among the few animal bones retrieved from locus 112, however, were three unclassified fish bones, the femur of a slaughtered caprid, and the jaw bone of a piglet, aged between 7 and 11 months when it was butchered. The piglet, in particular, suggests a pre-Islamic context. In light of the contextual evidence provided by stratigraphic analysis, ceramic and numismatic horizons, radiocarbon dates and faunal remains, the deposits containing the ivory can be securely dated to the 7th century.

ARTIFACT HISTORIES, RAW MATERIAL STUDIES

We used a range of methods to investigate the ivory and bone finds from Aylah in greater detail. These included scanning electron microscopy (SEM) and Fourier transform infrared (FTIR) to analyse the preservation condition of the material. Following this we used zooarchaeology by mass spectrometry (ZooMS), ancient DNA, and stable isotope analyses to determine the species of the bone or ivory and to characterise the geographic source of the raw material.45 We first used SEM and FTIR to assess whether there was likely to be organic material preserved in the ivory and bone samples to extract for the biomolecular analyses (ZooMS, aDNA, isotope analysis). For each method described below, full details of the procedure and extraction protocol are listed in the Supplementary Material. The data from these methods help reconstruct the biographies of the artifacts from Aylah by understanding where they might have originated. Knowing the source region of these artifacts could help us clarify more precisely how trans-regional networks operated.

Ivory is a specialized tooth that grows outwardly, made primarily of dentine. Many species grow ivory, and for material excavated in Red Sea ports such as Aylah, the ivory is likely to come from elephant (African-Loxodonta africana/cyclotis or Asian-Elephas maximus), hippopotamus, or dugong. If the artifacts are large enough, it is possible to identify the species by eye or microscope on the basis of the growth structure of the dentine. Elephant ivory has a distinctive growth pattern in that it grows in a cone-in-cone pattern which causes an optical effect called Schreger lines. This is visible when viewing elephant ivory in cross-section.46 It is not possible, however, to distinguish between the different species of elephant only through visual analysis of ivory artifacts. The definitive way to identify the species of elephant from which an ivory artifact derives is through ancient DNA analysis.

Nine of the artifacts were visually identified as elephant ivory due to the presence of the Schreger pattern and two were visually identified as bone, although they had previously been categorized as possibly ivory (Table 1).

Scanning electron microscopy

Scanning electron microscopy (SEM) is used to visualize the morphology of the outer surface of ivory, and is useful for assessing degradation, damage, or post-depositional processes on an artifact. The loose structure and darker patches in the image (Fig. 6) of the archaeological elephant ivory specimen from Aylah is similar to what has been shown in other archaeological ivories to constitute a breakdown of collagen in the dentine structure, which in turn suggests that there is some loss of organic material in this sample.47 This is not surprising given the burial conditions from which these samples were excavated. Aylah has a hyper-arid climate, which can lead to cracking and deterioration of the surface of bone and ivory, and in this case the discovery of the artifacts in waterlogged and extremely saline conditions seems to have exacerbated the degradation process. Fluctuations in groundwater as well as evaporation in the burial environment can cause protein loss and/or damage.48 

FIGURE 6

Scanning electron microscope image of sample number AAP 22, an elephant ivory specimen from Aylah.

FIGURE 6

Scanning electron microscope image of sample number AAP 22, an elephant ivory specimen from Aylah.

Fourier transform infrared (FTIR) spectroscopy

FTIR spectroscopy is used to measure the mineral composition of bone or ivory, and in this way, can be used to assess diagenesis, or post-depositional physical and chemical changes that occur as a result of the burial environment.49 Two commonly used indices in FTIR spectroscopy to evaluate preservation in archaeological bone or ivory are the carbonate to phosphate ratio (C/P) and the splitting factor (IR-SF). The C/P ratio is measured by calculating the height of the carbonate and phosphate bands in the spectra for the sample using an established method.50 Well-preserved modern bone has been measured in the range of 0.23–0.34, so if a sample falls outside of this range, then it has likely been diagenetically altered.51 The IR-SF, referred to as the splitting factor, measures the degree of alteration in the crystal size and crystal organization.52 A higher IR-SF value is associated with larger crystal sizes and regular patterning of the crystals, whereas a lower IR-SF value is associated with smaller crystal sizes and irregular patterning typically found in modern bone.32 Values over 3.4 are indicative of large crystal size and a loss of proteins such as collagen that can occur when bone and ivory are diagenetically altered in the burial environment.53 

Results

As can be seen in Figure 7, there is a large range in the IR-SF and C/P values published for modern and archaeological bone and ivory, with modern material (grey markers) plotting on the left side of this graph. Two bone samples and one ivory sample from Aylah cluster towards the measured range for modern bone, along with other values published for archaeological bone and ivory which have been found in a range of burial environments, including ivory from submerged marine environments.54 One sample of ivory from Aylah, however, falls outside this cluster, further towards the range of archaeological bone and ivory material that is less well-preserved. Although it fell just outside of the range for well-preserved material, we were able to extract good quality collagen from this sample for stable isotope analyses and radiocarbon dating. The FTIR results thus indicated that the crystal size, organization, and carbonate to phosphate ratios in the Aylah ivory samples were similar to those measured in well-preserved modern and archaeological ivory rather than archaeological bone or ivory that has undergone significant loss of organic material in the burial environment.

FIGURE 7

Carbonate to phosphate ratio (C/P) and splitting factor (IR-SF) data from published values of bone and ivory compared to values measured on the four archaeological bone and ivory samples from Aylah (see published values in: Melanie M. Beasley, Eric J. Bartelink, Lacy Taylor, and Randy M. Miller, “Comparison of Transmission FTIR, ATR, and DRIFT Spectra: Implications for Assessment of Bone Bioapatite Diagenesis,” Journal of Archaeological Science 46 (2014): 16–22; María Teresa Doménech-Carbó, Milagros Buendía-Ortuño, Trinidad Pasies-Oviedo, and Laura Osete-Cortina, “Analytical Study of Waterlogged Ivory from the Bajo de la Campana Site (Murcia, Spain),” Microchemical Journal 126 (2016): 381–405; I. M. Godfrey, E. L. Ghisalberti, L. T. Byrne, and G. W. Richardson, “The Analysis of Ivory from a Marine Environment,” Studies in Conservation 47, 1 (2013): 29–45).

FIGURE 7

Carbonate to phosphate ratio (C/P) and splitting factor (IR-SF) data from published values of bone and ivory compared to values measured on the four archaeological bone and ivory samples from Aylah (see published values in: Melanie M. Beasley, Eric J. Bartelink, Lacy Taylor, and Randy M. Miller, “Comparison of Transmission FTIR, ATR, and DRIFT Spectra: Implications for Assessment of Bone Bioapatite Diagenesis,” Journal of Archaeological Science 46 (2014): 16–22; María Teresa Doménech-Carbó, Milagros Buendía-Ortuño, Trinidad Pasies-Oviedo, and Laura Osete-Cortina, “Analytical Study of Waterlogged Ivory from the Bajo de la Campana Site (Murcia, Spain),” Microchemical Journal 126 (2016): 381–405; I. M. Godfrey, E. L. Ghisalberti, L. T. Byrne, and G. W. Richardson, “The Analysis of Ivory from a Marine Environment,” Studies in Conservation 47, 1 (2013): 29–45).

Zooarchaeology by mass spectrometry (ZooMS)

ZooMS is a biomolecular method of genus or species identification based on peptide mass fingerprinting of the protein collagen that is found in bone and ivory.55 Collagen is made up of amino acids, and differences in the sequence of these amino acids result in genus and sometimes species-specific peptide profiles. In combination with more general knowledge of the site's archaeozoology, it is possible to use these profiles to identify unknown fragments of bone, tooth, or ivory.56 ZooMS is an effective way to distinguish between elephant, hippopotamus, and dugong because they have distinctive peptide mass fingerprints. It cannot, however, distinguish between elephant species (African forest, savanna or Asian) because the species have the same sequence.57 For this purpose, ancient DNA analysis is necessary.

Results

Four ivory and two bone artifacts from Aylah were identified to family with ZooMS analysis. The ZooMS results confirmed two of the visually identified ivory pieces as Elephantidae (Loxodonta sp. or Elephas), and identified two further pieces that had been too small to visually identify as Elephantidae (Loxodonta sp. or Elephas). The two artifacts made of bone were identified as being from the Camelidae family, and are likely camel species from the Camelus genus, as these are indigenous to the region around Aylah. The ZooMS results are detailed in the Supplementary Material section.

Ancient DNA

As it is not possible to determine the species of elephant with ZooMS or visual analyses, we submitted six of the thirteen ivory fragments listed in Table 1 to the ancient DNA lab at the University of York, UK for DNA isolation and sequencing. These fragments were identified as elephant (Loxodonta sp. or Elephas) through visual inspection and ZooMS, but we wanted to know whether the ivory was from African (Loxodonta africana or cyclotis) or Asian (Elephas maximus) elephants.58 Our primary aim was to determine the source location of the elephant, as there were established commercial connections between the Red Sea region and both Southeast Asia and East Africa during the period in question. Ancient DNA isolation methods were performed in the dedicated ancient DNA laboratory, with appropriate contamination procedures in place. Approximately 100 mg of ivory (dentine) was obtained for three samples (AAP23–1, AAP23–2, AAP304) by drilling. DNA was extracted using a silica-based extraction protocol59. Unfortunately, this procedure indicated that there was no targeted (elephant) DNA present in the samples.

Stable isotope analysis: carbon (δ13C) and nitrogen (δ15N)

We measured the stable light isotopes of carbon (δ13C) and nitrogen (δ15N) from seven of the thirteen ivory artefacts from Aylah using extraction methods developed for archaeological bone samples from similar burial conditions as the Aylah ivories.60 The carbon isotope value measured from elephant tissue reflects the food that the elephant consumed. Elephants are mixed feeders, meaning that they eat a variety of plant foods, from trees to shrubs to tropical grasses. In African elephant habitats, most grasses use the C4 photosynthetic pathway, while shrubs and trees use the C3 pathway. These two pathways result in different 13C/12C ratios in plant tissue, which is transferred to the elephant when it consumes this plant tissue and ends up in the chemical composition of the tissue that forms its ivory tusks. Consequently, it is possible to determine what type of vegetation the elephant consumed by analysing 13C/12C ratios in the ivory, and this in turn reveals the habitat in which the elephant lived. Recently published stable isotope data on modern elephant populations across Africa and Asia reveal that there is a considerable amount of overlap in the isotopic ranges of elephants from different regions of Africa, as well as between African and Asian elephant populations.61 It is often easier to determine what type of habitat an ivory sample did not originate from than it is to pinpoint its origin. Elephants from environments such as closed-canopy tropical rainforests have carbon isotope values (δ13C) that are at the extreme end of the ranges measured across the African continent. For example, elephants from densely forested environments such as habitats found in western and central Africa typically have low δ13C collagen values (more negative than -21‰), which reflects the dense C3 vegetation they consume in their habitat.62 Modern savanna or bush elephants living in eastern and southern Africa have a mixed diet of C3 and C4 vegetation with δ13C collagen values measured between −21‰ and −12‰.63 Modern Asian elephants have δ13C collagen values measured from −24‰ to −11‰, indicating the diversity of habitats in which they range, meaning they consume plants from more open C4 grasslands to more closed-canopy forests dominated by C3 plants.64 

Nitrogen isotope values measured from elephant tissue are more variable, as nitrogen originates from a range of sources in the elephant's diet and environment. Nitrogen in the soil is fixated by plants which are then consumed by animals, and therefore becomes part of the composition of the growing animal's tissue. Plants and animals in more arid environments tend to have higher δ15N values, as has been documented in African elephants living in habitats receiving low annual rainfall, but this has been shown to be inconsistent in modern elephant populations.65 In their study of modern Asian and African elephant populations δ15N values measured in Asian elephants ranged from 5.9–11.2‰ and in African elephants from 4.2–17.2‰, so the range from Asian elephants is fully encompassed within the range for African elephants.66 

Results

The δ13C and δ15N values from the seven pieces of elephant ivory from Aylah are −16.5, −17.6, −17.8, −18.1, −18.2, −20.0, and −20.5‰ for the carbon and 6.2, 12.8, 10.0, 10.1, 6.8, 6.3, and 4.7‰ for the nitrogen, respectively (Fig. 8). The δ13C results indicate that most of the elephants from which the Aylah ivory derived consumed a large proportion of C4 grass, especially the ivory sample with a δ13C value of −16.5‰. The values for these seven elephants from Aylah fall within the range measured from the tissue of modern African elephants which lived in eastern, western, central, and southern Africa as well as Asia.67 It is problematic to directly compare the modern elephant data with archaeological ivory data, as habitats and elephant populations have certainly changed in the last centuries. However, the modern elephant data is useful to demonstrate how difficult it is to characterize elephant populations based on just carbon and nitrogen isotope analyses alone.68 

FIGURE 8

δ13C and δ15N values for archaeological ivory from Aylah (black circles) compared with reference samples from published elephant ivory data from modern and historic elephants. Data plotted with the R package SIAR (Andrew C. Parnell, Richard Inger, Stuart Bearhop, and Andrew L. Jackson, “Source Partitioning Using Stable Isotopes: Coping with Too Much Variation,” PloS One 5, 8 [2010]: e9672) with standard ellipses showing maximum likelihood based on small sample sizes. A correction factor based on the year of publication or year of collection has been added to δ13C values of modern elephants to correct for depletion of 13C in atmospheric CO2 since the Industrial Revolution, due to burning of fossil fuels, to compare with archaeological samples from Ayah (see Helge Hellevang and Per Aagaard, “Constraints on Natural Global Atmospheric CO2 Fluxes from 1860 to 2010 Using a Simplified Explicit Forward Model," Scientific Reports 5 [2015]: 17352). See published data for southern African ivory data (green circles), central Africa (blue circles), eastern Africa (red circles), western Africa (light blue circles), and Asia (pink circles) in: Thure E. Cerling, Patrick Omondi, and Anthony N. Macharia, “Diets of Kenyan Elephants from Stable Isotopes and the Origin of Confiscated Ivory in Kenya,” African Journal of Ecology 45, 4 (2007): 614–23; Jacqueline Codron, Daryl Codron, Matt Sponheimer, Kevin Kirkman, Kevin J. Duffy, Erich J. Raubenheimer, Jean-Luc Mélice, Rina Grant, Marcus Clauss, and Julia A. Lee-Thorp, “Stable Isotope Series from Elephant Ivory Reveal Lifetime Histories of a True Dietary Generalist,” Proceedings Biological Sciences 279, 1737 (2012): 2433–41; Ashley N. Coutu, Julia Lee-Thorp, Matthew J. Collins, and Paul J. Lane, “Mapping the Elephants of the 19th-Century East African Ivory Trade with a Multi-isotope Approach,” PloS One 11, 10 (2016): e0163606; P. L. Koch, J. Heisinger, C. Moss, R. W. Carlson, M. L. Fogel, and A. K. Behrensmeyer, “Isotopic Tracking of Change in Diet and Habitat Use in African Elephants,” Science 267, 5202 (1995): 1340–43; J.C. Vogel, B. Eglington, and J. M. Auret, “Isotope Fingerprints in Elephant Bone and Ivory,” Nature 346, 6286 (1990): 747; Nikolaas J. van der Merwe, Julia A. Lee-Thorp, and Richard H. V. Bell, “Carbon Isotopes as Indicators of Elephant Diets and African Environments,” African Journal of Ecology 26, 2 (1988): 163–72; Rina Rani Singh, Surendra Prakash Goyal, Param Pal Khanna, Pulok Kumar Mukherjee and Raman Sukumar, “Using Morphometric and Analytical Techniques to Characterize Elephant Ivory,” Forensic Science International 162, 1-3 (2006): 144–51; Stefan Ziegler, Stefan Merker, Bruno Streit, Markus Boner, and Dorrit E. Jacob, “Towards Understanding Isotope Variability in Elephant Ivory to Establish Isotopic Profiling and Source-Area Determination,” Biological Conservation 197 (2016): 154–63.

FIGURE 8

δ13C and δ15N values for archaeological ivory from Aylah (black circles) compared with reference samples from published elephant ivory data from modern and historic elephants. Data plotted with the R package SIAR (Andrew C. Parnell, Richard Inger, Stuart Bearhop, and Andrew L. Jackson, “Source Partitioning Using Stable Isotopes: Coping with Too Much Variation,” PloS One 5, 8 [2010]: e9672) with standard ellipses showing maximum likelihood based on small sample sizes. A correction factor based on the year of publication or year of collection has been added to δ13C values of modern elephants to correct for depletion of 13C in atmospheric CO2 since the Industrial Revolution, due to burning of fossil fuels, to compare with archaeological samples from Ayah (see Helge Hellevang and Per Aagaard, “Constraints on Natural Global Atmospheric CO2 Fluxes from 1860 to 2010 Using a Simplified Explicit Forward Model," Scientific Reports 5 [2015]: 17352). See published data for southern African ivory data (green circles), central Africa (blue circles), eastern Africa (red circles), western Africa (light blue circles), and Asia (pink circles) in: Thure E. Cerling, Patrick Omondi, and Anthony N. Macharia, “Diets of Kenyan Elephants from Stable Isotopes and the Origin of Confiscated Ivory in Kenya,” African Journal of Ecology 45, 4 (2007): 614–23; Jacqueline Codron, Daryl Codron, Matt Sponheimer, Kevin Kirkman, Kevin J. Duffy, Erich J. Raubenheimer, Jean-Luc Mélice, Rina Grant, Marcus Clauss, and Julia A. Lee-Thorp, “Stable Isotope Series from Elephant Ivory Reveal Lifetime Histories of a True Dietary Generalist,” Proceedings Biological Sciences 279, 1737 (2012): 2433–41; Ashley N. Coutu, Julia Lee-Thorp, Matthew J. Collins, and Paul J. Lane, “Mapping the Elephants of the 19th-Century East African Ivory Trade with a Multi-isotope Approach,” PloS One 11, 10 (2016): e0163606; P. L. Koch, J. Heisinger, C. Moss, R. W. Carlson, M. L. Fogel, and A. K. Behrensmeyer, “Isotopic Tracking of Change in Diet and Habitat Use in African Elephants,” Science 267, 5202 (1995): 1340–43; J.C. Vogel, B. Eglington, and J. M. Auret, “Isotope Fingerprints in Elephant Bone and Ivory,” Nature 346, 6286 (1990): 747; Nikolaas J. van der Merwe, Julia A. Lee-Thorp, and Richard H. V. Bell, “Carbon Isotopes as Indicators of Elephant Diets and African Environments,” African Journal of Ecology 26, 2 (1988): 163–72; Rina Rani Singh, Surendra Prakash Goyal, Param Pal Khanna, Pulok Kumar Mukherjee and Raman Sukumar, “Using Morphometric and Analytical Techniques to Characterize Elephant Ivory,” Forensic Science International 162, 1-3 (2006): 144–51; Stefan Ziegler, Stefan Merker, Bruno Streit, Markus Boner, and Dorrit E. Jacob, “Towards Understanding Isotope Variability in Elephant Ivory to Establish Isotopic Profiling and Source-Area Determination,” Biological Conservation 197 (2016): 154–63.

The modern elephant data are also used to understand the type of habitats in which the archaeological elephant ivory might have originated. For example, the δ13C values of the archaeological ivory from Aylah suggest that none of the ivory was sourced from closed canopy rainforests, as the δ13C collagen values of elephants from these habitats are typically lower than −21‰.69 The δ15N values of the Aylah ivory samples are a large range of values, and when compared to the range measured in modern elephants, this range (8‰) has been measured across entire regions of Africa in elephant ivory originating from diverse habitats with different rainfall patterns and plant species.52 There is also one sample from Aylah with a relatively high δ15N value (12.8‰), which compared to modern elephant values suggests this sample could be African in origin. The data shown in Figure 8 show that the Aylah ivories are more similar to the isotopic ranges of elephants which lived in eastern or southern African savannas, although we have few data for elephants which once lived in northeastern Africa such as Egypt, Somalia, and Ethiopia. One sample from a study of historic ivory70 which originated from present-day eastern Ethiopia plots closely to three of the Aylah ivory samples in Figure 8. Geographically, northeast Africa would have been connected with the Red Sea region in terms of trade links, and therefore this is an important area from which we do not have sufficient data.

The seven pieces tested include one piece that is part of a pyxis, two carved pieces, and four pieces of worked cut-outs (Table 1). Two of the worked cut outs have similar δ13C and δ15N values, which could mean that the pieces are from the same tusk, or at least from elephants that lived in a similar habitat. Aside from these two pieces, however, the large range of δ15N values (8‰) suggests that the remaining pieces did not originate from elephants living in similar habitats, yet may have been sourced from the same region. Without ancient DNA results to verify that these samples are African savanna (Loxodonta africana), African forest (Loxodonta cyclotis), or Asian (Elephas maximus) elephants, it is difficult to say where these Aylah elephants originated. The results do suggest, however, that these animals were hunted in a range of different habitats, which in turn implies that supplying the trans-regional demand was met by different local groups.71 Although the isotope results do not allow us to pinpoint a particular region, northeastern and eastern Africa is a likely region of origin for these artefacts, as there are a diverse range of habitats from which these elephants could have originated and it fits with the documentary evidence of ivory source regions supplying the Red Sea in the 7th century C.E.

DISCUSSION AND CONCLUSION

This paper set out to accomplish two goals. On the one hand, we wished to demonstrate the potential for a detailed study of individual artifact categories, especially artifacts such as ivory which would have had to have been traded into the site of Aylah. The second goal was to apply a combination of techniques to understand the raw material of these objects in more detail, and to explore the preservation conditions for conducting biomolecular analyses, given the region and environment in which they were found.

Dealing with the second issue first, this study has outlined some of the techniques to assess the state of preservation of archaeological ivory. The use of non-destructive screening techniques (SEM and FTIR) provided useful information on the state of preservation, which can aid in selecting the best pieces for destructive biomolecular analyses from larger assemblages, thus lowering laboratory costs. Additionally, our study has demonstrated some of the limitations of using these techniques on poorly preserved material, in particular the negative aDNA results. There was some success in obtaining ZooMS and stable isotope results, confirming that collagen can be preserved even when DNA is not.72 Furthermore, when combining large data sets on stable isotope values of elephants across Africa and Asia, it is not possible to determine the source region of an archaeological ivory sample that could be African or Asian using δ13C and δ15N data. The only conclusive way to do this is to use a combination of aDNA and isotope analyses.73 

Although this study builds on a relatively small data set, we can draw some general conclusions on the commercial systems in place at Aylah in the transitional period from Byzantine to Islamic rule. The identification of the pyxis, rings, and discards as elephant ivory confirms that raw elephant ivory was being brought to Aqaba, and the isotope values as well as documentary evidence indicate that the most likely source of at least some of this ivory was northeastern and eastern Africa. This is important because the most geographically proximate source of indigenous ivory would have been hippopotamus tusks from Egypt, which our material is not. The archaeological context and dating of these ivory finds suggests they are associated with building activities in the 7th century, though they do not unequivocally confirm a 7th century date for their craftsmanship and trade.

In fact, in those cases where radiocarbon dating was possible, most of the samples indicated a 6th – 7th century date for the ivory. One fragment was dated to the early 1st century, but this is most likely due to the fact that the fragments were retrieved from a dump deposit. The fragment in question could thus have been excavated elsewhere with the intention of using it as fill material. Another fragment appears to date from the mid-5th century, but here we might do well to remember both the lifespan of elephants and the longevity of valued objects such as ivory. Combined, the evidence suggests that a vibrant trade and craft of ivory was present in Aqaba during the 6th and 7th centuries at least, an activity that appears to have continued into the Islamic Middle Ages.

Another important observation underlined in this study is that certain non-indigenous materials were being worked at Aylah. The nature of the ivory fragments suggests that the end products were utilitarian rather than exquisite objects for display. This could in turn suggest a degree of local consumption, but it seems more likely that the crafting of ivory objects hinged on a desire to increase the value of an imported raw material through processing. Even though the circular cutting of the discards suggests specific objects being produced (buttons and boxes, e.g.), the notion that these could have been scraps leftover from cutting the tusks down to different raw sizes must also be entertained.74 Also worth noting in that regard is the size of the discarded pieces, which are large enough to be used for a range of other purposes (like inlay work). Whether this is suggestive that African ivory, at least in some periods, was relatively easy to come by in Aqaba is hard to say.

Whatever the motivation for processing the ivory at Aylah, the fragments retrieved from our deep probe unequivocally link the trade in ivory to the port of Aylah. The presence of this commodity reveals the port to be an integral node in the intercontinental trade networks of Late Antiquity and the early Middle Ages. The finds prove that ivory was actively utilized and worked by craftsmen in emporia along the Red Sea, though it does not necessarily tell us who these craftsmen were. One option is itinerant craftsmen, although there is no direct evidence for this. Perhaps craftsmen came to Aylah from elsewhere, such as Coptic Egypt or Aksum, places which by the 7th century had engaged in professional ivory carving for centuries. Similar carving traditions were also present in southern Palestine, where objects carved from camel bone circulated ubiquitously and the skillset needed to execute this work would have been easily adaptable to the more pliable elephant ivory.75 Whatever the case, experimentation on these fragments has allowed us to gain new and concrete insights into the nature, extent and chronology of regional trade patterns in the Red Sea and beyond.

SUPPLEMENTARY MATERIAL: METHODS

Fourier Transform Infrared (FTIR) spectroscopy Methods

Infrared spectroscopy (IR) measures how much radiation is absorbed when a powdered sample of bone or ivory is subject to a small amount of radiation. The spectrum that results is a measurement of the wavelength at which the radiation was absorbed. The resulting absorbance bands can be ascribed to certain molecular groups such as phosphate and carbonate, and therefore this analysis is a way to roughly measure bone composition.76 We used attenuated total reflectance (ATR) sample preparation, which is comparable to the traditional transmission FTIR method.77 The ATR method analyses the bone or ivory powder directly rather than forming it into a potassium bromide (KBr) pellet as in the traditional method. It is thus less destructive to the sample. We measured four samples of ivory powder with a Bruker Alpha Platinum ATR and diamond crystal attachment to record the spectral range of 4000–400 cm-1. The spectral resolution of 4cm-1 was used with 24 scans, and measurements were made in triplicate. We used OPUS 7.5 software for processing the spectra and doing baseline corrections.

Zooarchaeology by Mass Spectrometry (ZooMS) Methods

A small amount of ivory/bone powder (approximately 10–20 milligrams) analysed for FTIR was used for ZooMS by first heating in 50 mM ammonium bicarbonate buffer at 65 °C for 1 h to solubilize a small fraction of the collagen. This soluble fraction was trypsinated overnight at 37 °C to break the protein into peptide fragments, which were then concentrated and eluted using C18 Zip Tips. Mass spectra (Figure 9) were measured using a calibrated Ultraflex III (Bruker Daltonics) MALDI-TOF MS instrument in reflector mode and compared with reference spectra to identify the samples.78 

FIGURE 9

Zooarchaeology by Mass Spectrometry spectra from samples AAP0169, AAP0319, AAP0023, AAP0033, AAP0028, and AAP 0304. Samples AAP0169 and AAP0319 were identified as belonging to the Camelidae family and AAP0023, AAP0033, AAP 0028, and AAP0304 were identified as elephant. Identifications were based on reference spectra from: Michael Buckley, Matthew Collins, Jane Thomas-Oates, and Julie C. Wilson, “Species Identification by Analysis of Bone Collagen using Matrix-Assisted Laser Desorption/Ionisation Time-of-Flight Mass Spectrometry,” Rapid Communications in Mass Spectrometry 23 (2009), 3843–54.

FIGURE 9

Zooarchaeology by Mass Spectrometry spectra from samples AAP0169, AAP0319, AAP0023, AAP0033, AAP0028, and AAP 0304. Samples AAP0169 and AAP0319 were identified as belonging to the Camelidae family and AAP0023, AAP0033, AAP 0028, and AAP0304 were identified as elephant. Identifications were based on reference spectra from: Michael Buckley, Matthew Collins, Jane Thomas-Oates, and Julie C. Wilson, “Species Identification by Analysis of Bone Collagen using Matrix-Assisted Laser Desorption/Ionisation Time-of-Flight Mass Spectrometry,” Rapid Communications in Mass Spectrometry 23 (2009), 3843–54.

Ancient DNA Methods

We submitted six of the thirteen ivory fragments listed in Table 1 to the ancient DNA lab at the University of York, UK for DNA isolation and sequencing. Ancient DNA isolation methods were performed in the dedicated ancient DNA laboratory, with appropriate contamination procedures in place. Approximately 100 mg of ivory (dentine) was obtained for three samples (AAP23–1, AAP23–2, AAP304) by drilling. DNA was extracted using a silica-based extraction protocol.79 Unfortunately, this procedure was unsuccessful as it indicated that there was no targeted (elephant) DNA present in the samples.

One extraction blank (mock extraction) was also included and treated exactly as the extracts throughout the whole process. Five sets of primers were designed to target regions of variability between Loxodonta and Elephas mitochondrial genomes. Unfortunately, gel visualization of PCR products indicated that there was no targeted DNA present in samples (no band present, or multiple bands of incorrect length, indicative of sample contamination) so sequencing was not attempted.

Stable carbon (δ13C) and nitrogen (δ15N) isotope analysis

In brief, small chips of ivory weighing approximately 50–100 milligrams each were first cleaned in an ultrasonic bath with distilled water, dried, and subsequently placed in a solution of 0.5M EDTA (PH 8) for a period of 1 week until the ivory chip was completely demineralized. The samples were then rinsed seven times with distilled water, rinsed once with 0.1M NaOH, then placed in a refrigerator overnight in distilled water. The samples were then rinsed five times with distilled water, frozen and then lyophilized for analysis. Collagen samples weighing 1 milligram were analyzed in triplicate by EA-IRMS in a GSL analyzer coupled to a 20–22 mass spectrometer (Sercon, Crewe, UK) at the BioArCh laboratories, University of York. The analytical error for both δ13C and δ15N values, calculated on repeated measurements of homogenous standards was 0.2‰ or better. The extracts had C:N ratios of between 3.1–3.4, %C of between 39–44%, and %N of between 14–16%, all within the acceptable range for good quality collagen.80 Carbon isotope ratios are expressed as δ13C values, where δ13Csample = ((13C/12Csample/13C/12Cstandard - 1)*1000 ‰) and are measured against the standard of Pee Dee Belemnite (PDB). Nitrogen isotope ratios are expressed as δ15N values in the same way as δ13C values, and are measured against the standard of atmospheric N2 or AIR. Because we are comparing the carbon isotope values of archaeological ivory with modern elephant ivory values in the discussion of the results, we have taken into account that the δ13C value of the atmosphere has changed with the combustion of fossil fuels. This means that modern ivory measured in 2017 is offset by approximately 1.7‰ in the measured δ13C value as compared to archaeological ivory.81 We have taken this into account, and used the approximate correction for each of the published analyses of elephant isotope values in Fig. 8 based on the date of the publication or date of collection if known.

NOTES

NOTES
This research is the result of a fruitful collaboration between the Aylah Archaeological Project (University of Copenhagen) and the research project ENTREPOT (University of Aarhus and University of York), which between 2012 and 2014 explored the expansion of maritime communication and trade in the early medieval period (500–1200 C.E.). The authors wish to express their gratitude to Professor Søren Sindbæk, leader of the ENTREPOT project, for providing the opportunity for this collaborative research. The authors would like to thank the peer reviewers and editorial staff of Studies in Late Antiquity for their valuable feedback, which improved the quality of this article. Warm thanks are also due to Dr. Bénédicte Khan, associate researcher at the CNRS, who analyzed the physical traces on the ivory and provided key insights and to Dr. Alan Walmsley, who provided the numismatic analysis. Finally, we are thankful to the Department of Antiquities in Jordan. Particular thanks are extended to forme Director-General, Dr. Monther al-Jamhawi, to Aktham Oweidi and Ahmad Lash, for their support in securing the permission to extract and export samples of the ivory for laboratory testing and to Manal Basyouni, director of DoA Aqaba office, for her support. Several funding bodies provided absolutely crucial support, without which this research would never have been possible. We are particularly indebted to the C.L. David Foundation, The Danish Institute in Damascus, and the Elisabeth Munksgaard Foundation under the National Museum of Denmark.
1.
Basil Collins, trans., The Best Divisions for Knowledge of the Regions. Ahsan al-Taqâsîm fî Ma'rifat al-Aqâlim by Al-Muqaddasî (Reading: The Center for Muslim Contribution to Civilization & Garnet Publishing, 2001), 149.
2.
For a full bibliography on this work see Donald Whitcomb, “Ayla at the Millennium: Archaeology and History,” in Studies in the History and Archaeology of Jordan X, ed. F. al-Khraysheh (Amman: Department of Antiquities, 2009), 123–32.
3.
J.L. Meloy, “Results of Archaeological Reconnaissance in West Aqaba: Evidence of the Pre-Islamic Settlement,” Annual of the Department of Antiquities in Jordan 35 (1991): 397–414; S. Thomas Parker, “The Roman Port of Aila: Economic Connections with the Red Sea Littoral,” in Connected Hinterlands. Proceedings of the Red Sea Project IV, ed. L. Blue, J. Cooper, R. Thomas, and J. Whitewright (Oxford: Archaeopress, 2009), 79–84; S. Thomas Parker and Andrew M. Smith, The Roman Aqaba Project Final Report. Volume 1 - The Regional Environment and the Regional Survey (Boston: American Schools of Oriental Research, 2014), 14–22.
4.
Kristoffer Damgaard and Michael Jennings, “Once More unto the Beach: New Archaeological Research into Jordan's Port on the China Sea,” Annual of the Department of Antiquities in Jordan 57 (2013): 477–502.
5.
Donald Whitcomb, “Glazed Ceramics of the Abbasid Period from the Aqaba Excavations,” Transactions of the Oriental Ceramic Society 55 (1991): 43–65.
6.
Michael Raith, Radegund Hoffbauer, Harald Euler, Paul Yule, and Kristoffer Damgaard, “The View from Zafar: An Archaeometric Study of the Aqaba Pottery Complex and its Distribution in the 1st Millennium CE,” Zeitschrift für Orient-Archäologie 6 (2013): 318–48; Elisabeth V. Holmqvist, “Ceramics in Transition: A Comparative Analytical Study of the Late Byzantine-Early Islamic Pottery in Southern Transjordan and the Negev” (PhD diss., University College London, 2010).
7.
Timothy Insoll, “A Cache of Hippopotamus Ivory at Gao, Mali, and a Hypothesis of Its Use,” Antiquity 69 (1995): 327–36; A. Reid and A. K. Segobye, “An Ivory Cache from Botswana and Its Implications for Trade,” Antiquity 74 (January 2000): 326–31. In Northern Europe, ivory from walrus and whale was traded. A vivid example of this is the Norwegian captain Ohthere's gift of walrus ivory to King Alfred in the late 9th century: Janet Bately, “Text and Translation: the Three Parts of the Known World and the Geography of Europe North of the Danube According to Orosius' Historiae and Its Old English Version,” in Ohthere's Voyages. A Late 9th-Century Account of Voyages along the Coasts of Norway and Denmark and Its Cultural Context, ed. J. Bately and A. Englert (Roskilde: The Viking Ship Museum), 40–50. For a discussion on the distinctive uses of walrus versus elephant ivory, see Matthew Elliott Gillman, “A Tale of Two Ivories: Elephant and Walrus,” Espacio tiempo y forma. Serie VII, Historia del arte 5 (2017): 81–105.
8.
Lionel Casson, The Periplus Maris Erythraei: Text with Introduction, Translation, and Commentary (Princeton: Princeton University Press, 1989): 15–27, 51–53, 61, 81, 85.
9.
Steven E. Sidebotham, Martin Hense, and Hendrikje M. Nouwens, The Red Land. The Illustrated Archaeology of Egypt's Eastern Desert (Cairo: American University of Cairo Press, 2008), 190–93 and Steven E. Sidebotham, Berenike and the Ancient Maritime Spice Route (Berkeley: University of California Press, 2011), 48–50.
10.
David W. Phillipson, Archaeology at Aksum, Ethiopia, 1993–7, Volume II (London: British Institute in Eastern Africa, 2000).
11.
David Phillipson, “Aksum,” Azania: Archaeological Research in Africa 38.1 (2003): 1–68 and David Phillipson, “Aksum, the Entrepot, and Highland Ethiopia, 3th-12th Centuries,” in Byzantine Trade, 4–12th Centuries, ed. Marlia Mundell Mango (Surrey, UK: Ashgate Publishing, 2009), 353–68.
12.
Sarah M. Guérin, “Forgotten Routes? Italy, Ifrīqiya and the Trans-Saharan Ivory Trade,” Al-Masaq 25.1 (2013): 70–91 and Mark Horton, “The Swahili Corridor,” Scientific American 257.3 (1987): 86–93.
13.
In the ninth century, the Chinese writer Tuan Ch'eng-Shih cites the East African coast as a source of ivory, ambergris, and slaves. By the early 10th century, al-Mas'udi describes how ivory was exported from East Africa to Oman and thence to India and China for making dagger handles, chess pieces, and bangles for Hindu brides: G. S. P. Freeman-Grenville, The East African Coast: Select Documents (Oxford: Clarendon Press, 1962), 14–16, and Ashley Coutu, Gavin Whitelaw, Petrus le Roux, and Judith Sealy, “Earliest Evidence for the Ivory Trade in Southern Africa: Isotopic and ZooMS Analysis of Seventh–Tenth Century AD Ivory from KwaZulu-Natal,” African Archaeological Review 33.4 (2016): 411–435.
14.
Avinoam Shalem, “Trade in and the Availability of Ivory: The Picture Given by the Medieval Sources,” Journal of the David Collection 2.1 (2005): 25–36, and Sarah M. Guérin, “Avorio d'ogni ragione: The Supply of Elephant Ivory to Northern Europe in the Gothic Era,” Journal of Medieval History 36. 2 (2010): 156–74.
15.
See for example Timothy Insoll, The Archaeology of Islam in Sub-Saharan Africa (Cambridge & New York: Cambridge University Press, 2003) and Tim Power, The Red Sea from Byzantium to the Caliphate AD 500–1000 (Cairo: The American University in Cairo Press, 2012).
16.
S. Wynne-Jones & A. Laviolette, The Swahili World (London: Routledge, 2018); Neville Chittick, Kilwa: An Islamic Trading City on the East African Coast (Nairobi: British Institute in Eastern Africa, 1974); Neville Chittick, Manda. Excavations at an Island Port on the Kenya Coast (Nairobi: British Institute in East Africa, 1984); Mark Horton, Shanga. The Archaeology of a Muslim Trading Community on the Coast of East Africa (Nairobi: British Institute of East Africa, 1996); Mark Horton, “Islam, Archaeology, and Swahili Identity,” in Changing Social Identity with the Spread of Islam. Archaeological Perspectives, ed. D. Whitcomb (Chicago: Oriental Institute of the University of Chicago, 2004), 67–88; Stéphane Pradines, “L’île de Sanjé ya Kati (Kilwa, Tanzanie): Un mythe Shirâzi bien réel,” Azania 44 (2009): 49–73; Stéphane Pradines, “The Rock Crystal of Dembeni, Mayotte Mission Report 2013,” Nyame Akuma 80 (2013): 59–72.
17.
Paul Sinclair, Anneli Ekblom, and Marilee Wood, “Trade and Society on the South-East African Coast in the Later First Millennium AD: the Case of Chibuene,” Antiquity 86. 333 (2012): 723–37; Marilee Wood, Laure Dussubieux, and Peter Robertshaw, “The Glass of Chibuene, Mozambique: New Insights into Early Indian Ocean Trade,” South African Archaeological Bulletin 67. 195 (2012): 59–74; James Denbow, Carla Klehm, and Laure Dussubieux, “The Glass Beads of Kaitshàa and Early Indian Ocean Trade into the Far Interior of Southern Africa,” Antiquity 89. 344 (2015): 361–77. Mark Horton, Alison Crowther, and Nicole Boivin, “Ships of the Desert, Camels of the Ocean,” in Trade in the Ancient Sahara and Beyond, ed. D. J. Mattingly, V. Leitch, C. N. Duckworth, A. Cuénod, M. Sterry, and F. Cole (Cambridge: Cambridge University Press, 2017), 131–55.
18.
Nicole Boivin, Alison Crowther, Richard Helm, and Dorian Q. Fuller, “East Africa and Madagascar in the Indian Ocean World,” Journal of World Prehistory 26, 3 (2013): 213–81; Coutu et al., “Earliest Evidence for the Ivory Trade in Southern Africa,” 411–435; Freeman-Grenville, The East African Coast; Horton, Shanga; Paul Sinclair et al., “Trade and Society,” 723–37; Marilee Wood, Serena Panighello, Emilio F. Orsega, Peter Robertshaw, Johannes T. van Elteren, Alison Crowther, Mark Horton, and Nicole Boivin, “Zanzibar and Indian Ocean Trade in the First Millennium CE: The Glass Bead Evidence,” Archaeological and Anthropological Sciences 9. 5 (2017): 879–901.
19.
Jason D. Hawkes and Stephanie Wynne-Jones, “India in Africa: Trade Goods and Connections of the Late First Millennium,” Afriques 6 (2015): http://dx.doi.org/10.4000/afriques.1752.
20.
W. Thackston, Nasir-I Khusraw's Book of Travels: Safarnamah (Costa Mesa, California: Mazda Publishers, 2001).
21.
For a discussion of new ports along the Red Sea in the Early Islamic period see K.Damgaard, “Bahr al-Hijaz: Muslim Expansionism and Formation of an Arabian Mercantile Complex in the Red Sea (7th to 9th century CE),” in The Early Islamic Economy, ed. Hugh Kennedy and Fanny Bessard (Oxford: University of Oxford Press, 2019).
22.
Examples hereof can be seen in the early Islamic ports of the African Red Sea coast (Dahlak, Badi, Aydhab, and Suakin, e.g.) that may have specialised in slaves: G. Puglisi, “Alcuni vestigi dell'isola di Dahlac Chebir e la leggenda dei Furs,” in The Proceedings of the Third International Conference of Ethiopian Studies, ed. Taddese Beyene (Addis Ababa: Institute for Ethiopian Studies, 1969); see also Kristoffer Damgaard, “Modelling Mercantilism: An Archaeological Analysis of Red Sea Trade in the Early Islamic Period (650–1100 CE)” (PhD diss., University of Copenhagen, 2011), 220–36; or the Hijazi port of al-Hawra, where the steatite quarried in its immediate hinterland was processed for further export: Ahmed Kisnawi, S. de Jesus Prentiss, and Baseem Rihani, “Preliminary Report on the Mining Survey, Northwest Hijaz, 1982,” Atlal: Journal of Saudi Arabian Archaeology 7 (1983): 76–83.
23.
Archaeological examples of this are found in the fields of rock-cut cisterns outlying early Islamic emporia along the African coast; e.g. at Badi: J.W. Crowfoot, “Some Red Sea Ports in the Anglo-Egyptian Sudan,” The Geographical Journal 37 (1911): 544; H.E. Hebbert, “El-Rih - A Red Sea Island,” Sudan Notes and Records 18 (1935): 308–13; Mutsuo Kawatoko, “Preliminary Survey of ‘Aydhab and Bādi' Sites,” KUSH 16 (1993): 209); or at Aydhab: J. Theodore Bent, “A Visit to the Northern Sudan,” The Geographical Journal 8 (1896): 336; G.W. Murray, “Aidhab,” The Geographical Journal 68 (1926): 235–40; and at Dahlak: Timothy Insoll, “Dahlak Kebir, Eritrea: From Aksumite to Ottoman,” Adumatu 3 (2001): 39–50). Due to their number and consistent placement outside settlements, these may be related to trade in African slaves.
24.
For private collections, see: Jere L. Bacharach and Elizabeth Rodenbeck, “Bone, Ivory, and Wood,” in Fustat Finds. Beads, Coins, Medical Instruments, Textiles, and Other Artifacts from the Awad Collection, ed. J.L. Bacharach (Cairo: American University in Cairo Press, 2002), 32–43; Helen C. Evans, “Classical Survival,” in Byzantium and Islam. Age of Transition 7th - 9th Century, ed. H.C. Evans and B. Ratliff (New Haven & London: Metropolitan Museum of Art, 2012), 18–26; Gabriele Mietke, “Vine Rinceaux,” in Evans and Ratliff, ed., Byzantium and Islam, 175–82; Shalem, “Trade in and the Availability of Ivory,” 26; and for excavations, see Bernard Bruyère, Fouilles de Clysma-Qolzoum (Suez) 1930–32 (Cairo: Institut Francais d'Archeologie Orientale, 1966), 115, Plate XIX; Elzbieta Rodziewicz, Bone Carvings from Fustat - Istabl 'Antar. Excavations of the Institut francais d'archéologie orientale in Cairo (Cairo: Institut francais d'archéologie orientale, 2012).
25.
For Manzalah, see Flinders Petrie, Objects of Daily Use (London: British School of Archaeology in Egypt, 1927), 43–45, Plate 39; For Hama, see Gunhild Ploug, Evenlyn Oldenburg, E. Hammershaimb, R. Thomsen, and F. Løkkegaard, Hama. Fouilles et recherches de la Fondation Carlsberg 1931–1938. Les petits objets médiéveux sauf les verreries et poteries (Copenhagen: National Museum, 1969), 66–71, 110–135; for Fustat, see Elzbieta Rodziewicz, Bone Carvings from Fustat, 27, 224–26, 410.
26.
Rebecca Foote, “Frescoes and Carved Ivory from the Abbasid Family Homestead at Humeima,” Journal of Roman Archaeology 12 (1999): 423–428; Rebecca Foote, “An Abbasid Residence at al-Humayma,” in Byzantium and Islam, ed. Evans and Ratliff, 221–23. Ivory artifacts have also been found in Umayyad contexts at al-Fudayn (Mafraq) in Jordan: Jean-Baptiste Humbert, “El-Fedein-Mafraq,” Liber annuus 36 (1986): 356; Anna Ballian, “Al-Fudayn,” in Byzantium and Islam, ed. Evans and Ratliff, 213–14.
27.
Henri Stern, “Quelques oeuvres sculptées en bois, os et ivoire de style omeyyade,” Ars orientalis 1 (1954): 119–31; Avinoam Shalem, “Trade in and the Availability of Ivory,” 25–36.
28.
Flinders Petrie, Objects of Daily Use, 43–45, Plate 39.
29.
Many of the smaller objects may nevertheless have been produced from locally occurring hippopotamus ivory rather than elephant. Some studies suggest that hippopotamus ivory may have played a larger role in ivory craftsmanship than hitherto imagined: Insoll, “A Cache of Hippopotamus Ivory at Gao,” 327–36; O. Krzyszkowska, “Ivory in the Aegean Bronze Age: Elephant Tusk or Hippopotamus Ivory?” Annual of the British School at Athens 83 (1988): 209–34.
30.
Anthony Cutler, “Ivory,” in Late Antiquity: A Guide to the Postclassical World, ed. G. W. Bowersock, P. Brown, and O. Grabar (Harvard: Harvard University Press, 1999), 521–522; Mariam Rosser-Owen, Ivory: 8th to 17th Centuries. Treasures from the Museum of Islamic Art, Qatar. (Doha: The National Council for Culture, Arts and Heritage, 2004).
31.
A more detailed description is given in her PhD thesis, Bénédicte Khan, “L'exploitation artisanale des matières dures d'origine animale au Proche-Orient entre le IIIe s. av. J.-C. et le VIIe s. apr. J.-C.: une approche techno-économique” (PhD diss., Université Paris I Panthéon-Sorbonne, 2019).
32.
Laurel Phillipson, “Ivory-Working Techniques” in Archaeology at Aksum, Ethiopia, 1993–7, Volume II, ed. David W. Phillipson (London: British Institute in Eastern Africa, 2000), 460–468.
33.
Anthony Cutler, “Carving, Recarving, and Forgery: Working Ivory in the Tenth and Twentieth Centuries,” West 86th: A Journal of Decorative Arts, Design History, and Material Culture 18.2 (2011): 182–195.
34.
Coutu et al., “Earliest Evidence for the Ivory Trade in Southern Africa,” 411–435.
35.
Damgaard and Jennings, “Once More unto the Beach,” 477–502; Donald Whitcomb, “Evidence of the Umayyad Period from the Aqaba Excavations,” in The Fourth International Conference on the History of Bilâd al-Shâm during the Umayyad Period, ed. M.A. Bakhit and R. Schick (Amman: Yarmouk University, 1989), 164–184.
36.
Identifiable as such due to the discovery and excavation of two 7th-century kilns producing many of the vessels for which Aqaba has become famous. Ansam Melkawi, Khairieh 'Amr, and Donald S. Whitcomb, “The Excavation of Two Seventh-Century Pottery Kilns at Aqaba,” Annual of the Department of Antiquities in Jordan 38 (1994): 447–468; Donald Whitcomb, “Ceramic Production at Aqaba in the Early Islamic Period,” in La céramique byzantine et proto-Islamique en Syrie-Jordanie (IVe-VIIIe siècles apr. J.-C.). Actes du colloque tenu à Amman les 3, 4 et 5 décembre 1994, ed. E. Villeneuve and P.M. Watson (Beirut: Institut Francais d'Archaeologie du Proche-Orient, 2001), 296–303.
37.
John W. Hayes, Late Roman Pottery (London: British School at Rome, 1972).
38.
The collagen samples were dated in 2018 at the Poznan Radiocarbon Laboratory under Dr. Tomasz Goslar, to whom we extend our sincere thanks. Dating the collagen was suggested by one of the anonymous reviewers of this manuscript, and we wish to express our appreciation for pointing this omission out to us. We further extend heartfelt thanks to The Danish Institute in Damascus and Elisabeth Munksgaard Fonden for providing the financial support to conduct these additional tests.
39.
Among the published sites at which this lamp is common are Petra/Jabal Harun (Elisabeth V. Holmqvist, “Ceramic Lamps at Jabal Harun: Typo-Chronology and Chemical Composition (ED-XRF),” in Petra - The Mountain of Aaron. The Finnish Archaeological Project in Jordan Volume II. The Nabatean Sanctuary and the Byzantine Monastery, ed. Z.T. Fiema, J. Frösén, and M. Holappa (Helsinki: Societas Scientiarum Fennica, 2016), 244–65; Petra/ez-Zantur (I. Zanoni, “Tonlampen,” in Petra. Ez Zantur 1, ed. A. Bignasca [Mainz: Phillip von Zabern, 1996], 311–36); Upper Zohar (Richard P. Harper, Upper Zohar - An Early Byzantine Fort in Palaestina Tertia: Final Report of Excavations in 1985–1986 [Oxford: Oxford University Press, 1995], Fig. 19); Nessana (H. Dunscombe Colt, Excavations at Nessana Volume 1 [London: British School of Archaeology in Jerusalem, 1962], 63–64); Rohovot (R. Rosenthal-Heginbottom, “The Pottery,” in Excavations at Rehovot-in-the-Negev, Volume 1: The Northern Church, ed. Y. Tsafrir, J. Patrich, R. Rosenthal-Heginbottom, I. Hershkovitz, Y.D. Nevo, J. Zias, and R.H. Brill [Jerusalem: Israel Exploration Society, 1988], 88); Clysma/Qulzum (Bernard Bruyère, Fouilles de Clysma-Qolzoum [Sue]) 1930–32 [Cairo: Institut Français d'Archeologie Orientale, 1966]), 111–15; and of course Aqaba (Donald Whitcomb, Ayla. Art and Industry in the Islamic Port of Aqaba [Chicago: The Oriental Institute of the University of Chicago, 1994], 25–27; Whitcomb, “Ceramic Production at Aqaba in the Early Islamic Period,” 296–303).
40.
The earliest known contexts to have contained wheel-made lamps come from Petra/ez-Zantur, where six examples were retrieved from deposits dating between the earthquakes of 363 and 419 C.E.: I. Zanoni, “Tonlampen,” in Petra. Ez Zantur 1, ed. A. Bignasca (Mainz: Phillip von Zabern, 1996), 329–30.
41.
Raith et al. “The View from Zafar,” 318–48.
42.
Varda Sussman, “The Oil Lamps,” in Archaeological Excavations at Caesarea Maritima. Areas CC, KK and NN Final Reports. Volume 1: The Objects, ed. J. Patrich (Jerusalem: Israel Exploration Society, 2008), 207–300.
43.
The numismatic identification and analysis was conducted by Alan Walmsley, and we are most thankful for his contribution. Walmsley notes that despite being unidentifiable, coin AAP9 could well be a small-denomination Byzantine coin also datable to the 6th century, but that it may also be an earlier Roman coin. For more on the use of late Roman/Byzantine coinage in the early Islamic period see Alan Walmsley, “Coinage and the Economy of Syria-Palestine in the Seventh and Eighth Centuries CE,” in Money, Power and Politics in Early Islamic Syria: A Review of Current Debates, ed. J.F. Haldon (Farnham: Ashgate, 2010), 21–44.
44.
Identified at Deir ’Ain Abata and contextualized by Kate da Costa in her dissertation (Kate da Costa, “Byzantine and Early Umayyad Ceramic Lamps from Palestine/Arabia [ca. 300–700 A.D.]” [PhD diss., University of Sydney, 2003], 129–38, cat. no. 236–39).
45.
Michael Buckley, Matthew Collins, Jane Thomas-Oates, and Julie C. Wilson, “Species Identification by Analysis of Bone Collagen Using Matrix-Assisted Laser Desorption/Ionisation Time-of-Flight Mass Spectrometry,” Rapid Communications in Mass Spectrometry 23 (2009): 3843–54; Yasuko Ishida, Nicholas J. Georgiadis, Tomoko Hondo, and Alfred L. Roca, “Triangulating the Provenance of African Elephants Using Mitochondrial DNA,” Evolutionary Applications 6.2 (2013): 253–65; S.K. Wasser, L. Brown, C. Mailand, S. Mondol, W. Clark, C. Laurie, and B. S. Weir, “Genetic Assignment of Large Seizures of Elephant Ivory Reveals Africa's Major Poaching Hotspots,” Science 349, 6243 (2015): 84–87; N.J. van der Merwe, J. A. Lee-Thorp, J. F. Thackeray, A. Hall-Martin, F. J. Kruger, H. Coetzee, R. H. V. Bell, and M. Lindeque, “Source-Area Determination of Elephant Ivory by Isotopic Analysis,” Nature 346 (1990): 744–46; Ashley N. Coutu, Julia Lee-Thorp, Matthew J. Collins, and Paul J. Lane, “Mapping the Elephants of the 19th-Century East African Ivory Trade with a Multi-Isotope Approach,” PloS One 11. 10 (2016): e0163606; Stefan Ziegler, Stefan Merker, Bruno Streit, Markus Boner, and Dorrit E. Jacob, “Towards Understanding Isotope Variability in Elephant Ivory to Establish Isotopic Profiling and Source-Area Determination,” Biological Conservation 197 (2016): 154–63.
46.
Michael Locke, “Structure of Ivory,” Journal of Morphology 269. 4 (2008): 423–50; Erich Johann Raubenheimer, J. Dauth, M. J. Dreyer, P. D. Smith, and M. L. Turner, “Structure and Composition of Ivory of the African Elephant (Loxodonta africana),” South African Journal of Science 86.4 (1990: 192–93.
47.
María Teresa Doménech-Carbó, Milagros Buendía-Ortuño, Trinidad Pasies-Oviedo, and Laura Osete-Cortina, “Analytical Study of Waterlogged Ivory from the Bajo de la Campana Site (Murcia, Spain),” Microchemical Journal 126 (2016): 381–405 at 389; Dounia Large, Katharina Müller, and Ina Reiche, “Approche analytique pour l’étude des ivories archéologiques. Les defenses d’éléphant du site de Jinsha (1200–650 BC, Sichuan, Chine),” ArcheoSciences, revue d'Archéométrie 35 (2011): 167–177; X.W. Su and F. Z. Cui, “Hierarchical Structure of Ivory: From Nanometer to Centimeter,” Materials Science and Engineering: C 7. 1 (1999): 19–29; Yin, Zuowei, Pengfei Zhang, Quanli Chen, Chen Zheng, and Yuling Li, “A Comparison of Modern and Fossil Ivories using Multiple Techniques,” Gems and Gemology 49.1 (2013): 16–27.
48.
Robert E. M. Hedges and Andrew R. Millard, “Bones and Groundwater: Towards the Modelling of Diagenetic Processes,” Journal of Archaeological Science 22.2 (1995): 155–64; Clive N. G. Trueman, Anna K. Behrensmeyer, Noreen Tuross, and Steve Weiner, “Mineralogical and Compositional Changes in Bones Exposed on Soil Surfaces in Amboseli National Park, Kenya: Diagenetic Mechanisms and the Role of Sediment Pore Fluids,” Journal of Archaeological Science 31. 6 (2004): 721–39; Doménech-Carbó et al., “Analytical Study of Waterlogged Ivory,” 381–405.
49.
Melanie M. Beasley, Eric J. Bartelink, Lacy Taylor, and Randy M. Miller, “Comparison of Transmission FTIR, ATR, and DRIFT Spectra: Implications for Assessment of Bone Bioapatite Diagenesis,” Journal of Archaeological Science 46 (2014): 16–22.
50.
Lori E. Wright and Henry P. Schwarcz, “Infrared and Isotopic Evidence for Diagenesis of Bone Apatite at Dos Pilas, Guatemala: Palaeodietary Implications,” Journal of Archaeological Science 23. 6 (1996): 933–44.
51.
Beasley et al., “Comparison of Transmission,” 16–22; Wright et al., “Infrared and Isotopic Evidence for Diagenesis,” 933–44.
52.
Stephen Weiner and Ofer Bar-Yosef, “States of Preservation of Bones from Prehistoric Sites in the Near East: A Survey,” Journal of Archaeological Science 17. 2 (1990): 187–96.
53.
R. E. M. Hedges, “Bone Diagenesis: An Overview of Processes,” Archaeometry 44. 3 (2002): 319–28; C. I. Smith, C. M. Nielsen-Marsh, M. M. E. Jans, and M. J. Collins, “Bone Diagenesis in the European Holocene I: Patterns and Mechanisms,” Journal of Archaeological Science 34. 9 (2007): 1485–93.
54.
Doménech-Carbó et al., “Analytical Study of Waterlogged Ivory,” 381–405; I. M. Godfrey, E. L. Ghisalberti, L. T. Byrne, and G. W. Richardson, “The Analysis of Ivory from a Marine Environment,” Studies in Conservation 47.1 (2013): 29–45.
55.
N. L. van Doorn, “Zooarchaeology by Mass Spectrometry,” in Encyclopedia of Global Archaeology, ed. C. Smith (New York: Springer, 2014), 7998–8000.
56.
Also used on hair: Klaus Hollemeyer, Wolfgang Altmeyer, Elmar Heinzle, and Christian Pitra, “Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry Combined with Multidimensional Scaling, Binary Hierarchical Cluster Tree and Selected Diagnostic Masses Improves Species Identification of Neolithic Keratin Sequences from Furs of the Tyrolean Iceman Oetzi,” Rapid Communications in Mass Spectrometry 26. 16 (2012): 1735–45; and eggshell: John R. M. Stewart, Richard B. Allen, Andrew K. G. Jones, T. Kendall, K.E.H. Penkman, B. Demarchi, T. O'Connor, and M. J. Collins, “Walking on Eggshells: A Study of Egg Use in Anglo-Scandinavian York Based on Eggshell Identification Using ZooMS,” International Journal of Osteoarchaeology 24. 3 (2014): 247–55.
57.
M. Buckley, N. Larkin, and M. Collins, “Mammoth and Mastodon Collagen Sequences; Survival and Utility,” Geochimica et cosmochimica acta 75. 7 (2011): 2007–16.
58.
Alfred L. Roca, N. Georgiadis, J. Pecon-Slattery, and S. J. O'Brien, “Genetic Evidence for Two Species of Elephant in Africa,” Science 293. 5534 (2001): 1473–77.
59.
N. Rohland, H. Siedel and M. Hofreiter, “A Rapid Column-Based Ancient DNA Extraction Method for Increased Sample Throughput,” Molecular Ecology Resources 10 (2010): 677–683.
60.
Cheryl A. Makarewicz, “Winter Is Coming: Seasonality of Ancient Pastoral Nomadic Practices Revealed in the Carbon (δ13C) and Nitrogen (δ15N) Isotopic Record of Xiongnu Caprines,” Archaeological and Anthropological Sciences 9.3 (2017): 405–18; N. Tuross, M. L. Fogel, and P. E. Hare, “Variability in the Preservation of the Isotopic Composition of Collagen from Fossil Bone,” Geochimica et cosmochimica acta 52 (1988): 929–935; Noreen Tuross, “Comparative Decalcification Methods, Radiocarbon Dates, and Stable Isotopes of the Viri Bones,” Radiocarbon 54. 3–4 (2012): 837–844; Cheryl Makarewicz and Noreen Tuross, “Finding Fodder and Tracking Transhumance: Isotopic Detection of Goat Domestication Processes in the Near East,” Current Anthropology 53. 4 (2012): 495–505.
61.
Coutu et al., “Mapping the Elephants;” Ziegler et al., “Towards Understanding Isotope Variability,” 154–63.
62.
Nikolaas J. van der Merwe, Julia A. Lee-Thorp, and Richard H. V. Bell, “Carbon Isotopes as Indicators of Elephant Diets and African Environments,” African Journal of Ecology 26.2 (1988): 163–72; Ziegler et al., “Towards Understanding Isotope Variability,” 154–63.
63.
P. L. Koch, J. Heisinger, C. Moss, R. W. Carlson, M. L. Fogel, and A. K. Behrensmeyer, “Isotopic Tracking of Change in Diet and Habitat Use in African Elephants,” Science 267. 5202 (1995): 1340–43; Thure E. Cerling, John M. Harris, and Meave G. Leakey, “Browsing and Grazing in Elephants: the Isotope Record of Modern and Fossil Proboscideans,” Oecologia 120. 3 (1990): 364–74; Thure E. Cerling, Patrick Omondi, and Anthony N. Macharia, “Diets of Kenyan Elephants from Stable Isotopes and the Origin of Confiscated Ivory in Kenya,” African Journal of Ecology 45. 4 (2007): 614–23; Jacqueline Codron, Daryl Codron, Matt Sponheimer, Kevin Kirkman, Kevin J. Duffy, Erich J. Raubenheimer, Jean-Luc Mélice, Rina Grant, Marcus Clauss, and Julia A. Lee-Thorp, “Stable Isotope Series from Elephant Ivory Reveal Lifetime Histories of a True Dietary Generalist,” Proceedings of the Royal Society B: Biological Sciences 279. 1737 (2012): 2433–41.
64.
Ziegler et al., “Towards Understanding Isotope Variability,” 154–63; Rina Rani Singh, Surendra Prakash Goyal, Param Pal Khanna, Pulok Kumar Mukherjee, and Raman Sukumar, “Using Morphometric and Analytical Techniques to Characterize Elephant Ivory,” Forensic Science International 162. 1–3 (2006): 144–51; R. Sukumar and R. Ramesh, “Stable Carbon Isotope Ratios in Asian Elephant Collagen: Implications for Dietary Studies,” Oecologia 91.4 (1992): 536–39.
65.
R. Amundson, A. T. Austin, E. A. G. Schuur, K. Yoo, V. Matzek, C. Kendall, A. Uebersax, D. Brenner, and W. T. Baisden, “Global Patterns of the Isotopic Composition of Soil and Plant Nitrogen,” Global Biogeochemical Cycles 17. 1 (2003): 1031; J. N. Aranibar, L. Otter, S. A. Macko, C. J. W. Feral, H. E. Epstein, P. R. Dowty, F. Eckardt, H. H. Shugart, and R. J. Swap, “Nitrogen Cycling in the Soil-Plant System Along a Precipitation Gradient in the Kalahari Sands,” Global Change Biology 10. 3 (2004): 359–73; Brett P. Murphy and David M. J. S. Bowman, “Kangaroo Metabolism Does Not Cause the Relationship Between Bone Collagen δ15N and Water Availability,” Functional Ecology 20. 6 (2006): 1062–69; Brett P. Murphy and David M. J. S. Bowman, “The Carbon and Nitrogen Isotope Composition of Australian Grasses in Relation to Climate,” Functional Ecology 23. 6 (2009): 1040–49; Coutu et al., “Mapping the Elephants;” van der Merwe et al., “Source-Area Determination,” 744–46
66.
Ziegler et al.,“Towards Uunderstanding Isotope Variability,” 154–63.
67.
Coutu et al.,, “Mapping the Elephants;” Singh et al., “Using Morphometric and Analytical Techniques,” 144–51; Sukumar and Ramesh, “Stable Carbon Isotope Ratios,” 536–39; van der Merwe et al., “Source-Area Determination,” 744–46
68.
Oxygen and strontium isotope anlayses (δ18O and 87Sr/86Sr) were not considered due to the degradation of the mineral in archaeological dentine and the small amount of the samples. Oxygen isotope analysis on the collagen was not considered due to the variability measured in incrementally growing tissues such as ivory: I.C.C. von Holstein, M. von Tersch, A.N. Coutu, K.E.H. Penkman, C.A. Makarewicz, and M.J. Collins, “Collagen Proteins Exchange O with Demineralisation and Gelatinisation Reagents and also with Atmospheric Moisture,” Rapid Communications in Mass Spectrometry 32.6 (2018): 523–534.
69.
Van der Merwe et at., “Source-Area Determination,” 744–46; Ziegler et al., “Towards Understanding Isotope Variability,” 154–63.
70.
Coutu et al., “Mapping the Elephants.”
71.
Coutu et al., “Earliest Evidence for the Ivory Trade in Southern Africa,” 411–435; Denbow et al., “The Glass Beads of Kaitshàa,” 361–77.
72.
M. Buckley and C. Wadsworth, “Proteome Degradation in Ancient Bone; What Ancient Proteins Can Tell Us,” Palaeogeography, Palaeoclimatology, Palaeoecology 416 (2014): 69–79.
73.
A. De Flamingh, A. Coutu, S. Chirikure, J. Sealy, R. Malhi and A. Roca, “Using DNA to Determine the Species and Geographic Origins of Elephant Ivory Discovered in a 16th-Century Portuguese Shipwreck,” 88th Annual Meeting of the American Association of Physical Anthropologists 168 (2019): 55.
74.
Later sources tell us that ivory was sold in raw pieces of various sizes for use in different and often quite particular markets (bangle manufacture in India from African ivory, e.g.); see Freeman-Grenville, The East African Coast, 14–16.
75.
Peter M. Fischer and Georgina Herrmann, “Carved Bone Object from Tell Abu al-Kharaz in Jordan: A Palestinian Workshop for Bone and Ivory?,” Levant 27 (1995): 145–63; Haim Goldfus and Kim Bowes, “New Late Roman Bone Cravings from Halusa and the Problem of Regional Bone Carving Workshops in Palestine,” Israel Exploration Journal 50 (2000): 185–202.
76.
Melanie M. Beasley, Eric J. Bartelink, Lacy Taylor, and Randy M. Miller, “Comparison of Transmission FTIR, ATR, and DRIFT Spectra: Implications for Assessment of Bone Bioapatite Diagenesis,” Journal of Archaeological Science 46 (2014): 16–22.
77.
Beasley et al, “Comparison of Transmission,” 16–22; H. I. Hollund, F. Ariese, R. Fernandes, M. M. E. Jans, and H. Kars, “Testing an Alternative High-throughput Tool for Investigating Bone Diagenesis: FTIR in Attenuated Total Reflection (ATR) Mode,” Archaeometry 55, 3 (2013): 507–32.
78.
Michael Buckley, Matthew Collins, Jane Thomas-Oates, and Julie C. Wilson, “Species Identification by Analysis of Bone Collagen Using Matrix-Assisted Laser Desorption/Ionisation Time-of-Flight Mass Spectrometry,” Rapid Communications in Mass Spectrometry 23 (2009): 3843–54.
79.
N. Rohland, H. Siedel and M. Hofreiter, “A Rapid Column-Based Ancient DNA Extraction Method for Increased Sample Throughput,” Molecular Ecology Resources 10 (2010): 677–683.
80.
Stanley H. Ambrose, “Preparation and Characterization of Bone and Tooth Collagen for Isotopic Analysis,” Journal of Archaeological Science 17 (1990): 431–51; G. J. van Klinken, “Bone Collagen Quality Indicators for Palaeodietary and Radiocarbon Measurements,” Journal of Archaeological Science 26, 6 (1999): 687–95.
81.
Helge Hellevang and Per Aagaard, “Constraints on Natural Global Atmospheric CO2 Fluxes from 1860 to 2010 Using a Simplified Explicit Forward Model,” Scientific Reports 5 (2015): 17352.