From Petra to Rome: A Nabataean Gold Necklace in the Roman World, Luxury, movement, and identity in the 2nd–3rd century CE

by Mary Harrsch, © 2026

Photographed at the Dallas Museum of Art (Accession No. 1995.26), this elegant gold necklace illustrates the sophisticated jewelry traditions that flourished in the eastern provinces of the Roman Empire during the 2nd and early 3rd centuries CE. The museum attributes the piece to Nabataea, the former kingdom centered on Petra that was incorporated into the Roman Empire in 106 CE.

The necklace consists of a series of gold settings containing polished garnet cabochons, a gemstone particularly favored by Roman jewelers. Garnets were imported through long-distance trade networks extending to India and Sri Lanka, making them both fashionable decorative elements and indicators of access to the empire's extensive commercial connections. Nabatea served as a crossroads between Roman, Hellenistic, Arabian, and Near Eastern artistic traditions. That cultural blending helps explain why the design feels somewhat different from jewelry typically excavated in Italy itself.

Roman Gold Necklace with Garnet Cabochons (2nd–3rd Century CE), Photo by Mary Harrsch, Dallas Museum of Art,

Particularly striking is the elaborate central pendant, composed of multiple garnet settings surrounded by suspended chains terminating in small gold drops. My research revealed these dangling ornaments, known as pendilia, were designed to move with the wearer, creating a dynamic display of reflected light. Roman jewelry was intended not only to be seen but also to animate the body in motion.

The smaller decorative elements distributed around the necklace appear to represent stylized leaves or floral motifs. Such vegetal ornament was common throughout Roman decorative arts, appearing in wall paintings, mosaics, metalwork, and jewelry. The combination of floral forms, richly colored garnets, and intricate goldwork reflects the blending of Hellenistic, Near Eastern, and Roman artistic traditions characteristic of the eastern Mediterranean provinces.

Unlike the more restrained jewelry often associated with early Imperial Rome, eastern Roman luxury jewelry frequently emphasized color, movement, and visual complexity. Pieces such as this one foreshadow the increasingly elaborate aesthetic that would later characterize Byzantine jewelry.

As both a personal adornment and a portable store of wealth, a necklace of this quality would almost certainly have belonged to a woman of considerable means. Today it offers a fascinating glimpse into the craftsmanship, trade networks, and cultural diversity of the Roman Empire at its height.

If you enjoyed this post, never miss out on future posts by following me by email!

From Ningishzida to Hermes: The Near Eastern Roots of the Kerykeion

by Mary Harrsch © 2026

This is one of the finest surviving examples of a Greek kerykeion — the iconic staff of Hermes, messenger of the Olympians and guide of souls between the worlds of the living and the dead that I photographed at the Dallas Museum of Art.

Bronze kerykeion (caduceus), Greek, early 5th century BCE. Dallas Museum of Art. In Greek art, this staff identifies Hermes—herald, traveler's patron, and guide of souls. But its central motif—two serpents twisting around a rod—appears on Mesopotamian cylinder seals and ritual objects more than a millennium earlier, carried by the Sumerian snake-god Ningishzida. Is this visual inheritance or independent invention? The question remains open, but the parallel is impossible to ignore. Photograph by the Author.

Beneath the head of each serpent, the bronze-smith has rendered a distinct wattle or beard—a stylized, fleshy protrusion. This is not a naturalistic detail. No snake native to Greece (not the adder, the rat snake, nor the Aesculapian snake venerated in the healing cults of Asclepius) possesses a beard. Greek artists were perfectly capable of naturalistic rendering when they chose to be. The presence of this artificial feature therefore signals something else: the artist was following an iconographic convention, not observing nature.

That convention—marking a serpent as divine, powerful, or chthonic by giving it a beard or human-like head—originated not in Greece but in Mesopotamia, where the underworld god Ningishzida was depicted as a human-headed serpent or as the horned, bearded Bašmu dragon. The beard on this Hermes staff is a fossil of that forgotten hybrid form, a small but telling clue that the kerykeion's visual language traveled westward across centuries and cultures before a Greek bronze-caster shaped this staff in the early 5th century BCE.

The kerykeion (its Latin equivalent being caduceus) appears in Greek art from the Early Archaic period, carried most often by Hermes, though it is occasionally held by Iris — the messenger of Hera — and by Nike in her role as herald of victory. Its origins remain debated; the staff's source may ultimately lie in the ancient Near East.

The cylinder seal evidence

The most concrete argument comes from material culture. William Hayes Ward (1910) discovered that symbols closely resembling the classical caduceus appear on Mesopotamian cylinder seals, and suggested the symbol originated somewhere in the 3rd millennium BCE, proposing it as a plausible source for the Greek form. These seals predate the earliest Greek representations by millennia, which gives the diffusionist argument its chronological backbone.

Ningishzida and the divine prototype

The more ambitious theoretical claim involves the Sumerian underworld deity Ningishzida. Ningishzida was a messenger god who dwelled in the Underworld for part of the year, and the symbol for this underworld messenger was two entwined serpents on a staff — the same configuration as the kerykeion. Some scholars have proposed that the Greeks adopted this symbol from the Near Eastern context for their own messenger god Hermes, who shares the same chthonic, psychopomp functions.

Ningishzida makes his first appearance in the Fara god list from the Early Dynastic III period, dated to approximately 2600–2350 BCE. This is our earliest documented evidence of the deity by name, though of course the absence of earlier written records doesn't mean the cult didn't predate that.

His origins as a tree god

Although Ningishzida was a power of the netherworld, he appears to have originally been a tree god — his name apparently meaning "Lord Productive Tree," and he was probably the god of winding tree roots, since he was originally represented in serpent form. The snake-and-staff iconography thus likely grew organically out of this root-as-serpent conceptualization of an arboreal deity, rather than being a purely abstract symbol from the outset.

His nature and cult spread

His primary cult center was Gishbanda, a settlement situated between Lagash and Ur in southern Mesopotamia. Ningishzida, like his father Ninazu, is a chthonic deity associated with vegetation, growth and decay, snakes and demons. Associated with his role in agriculture, he was said to travel to the underworld at the time of the death of vegetation — in Mesopotamia, mid-summer to mid-winter.

The Gudea connection and the key visual evidence

A.L. Frothingham extended Ward's argument in 1916, proposing that the Greek god Hermes himself derived from an "Oriental deity of Babylonian extraction"—specifically, the snake god Ningishzida, whose earliest form was as a "messenger" deity.

Ningishzida as "a messenger god" is a slight oversimplification. His primary roles were as a chthonic vegetation deity and "chair-bearer of the underworld." The messenger-god parallel to Hermes is an inference drawn from his appearance as a doorkeeper in the Adapa myth, not a core title.

The long tail of the cult

Ningishzida's worship persisted all the way into the reign of the Persian king Darius I at Uruk, indicating that his cult endured through multiple historical periods.

Walter Burkert, one of the 20th century's leading scholars of Greek religion, characterized the entwined copulating serpents as not just a borrowed symbol but reference to a specific visual concept — paired serpents in a sexual or entwined posture. Burkert privileges cuneiform literature as a source of literary transmission in light of the continuous routes of contact between Mesopotamia and Greek speakers, with these contacts reaching an apogee in the 8th and 7th centuries BCE. He tracks the migrant craftsmen who brought the Greeks new techniques and designs, the wandering seers and healers teaching magic and medicine, and the important Greek borrowings from Near Eastern poetry.

So from roughly 2600 BCE to at least the late 6th century BCE, this deity with his entwined-serpent iconography was continuously present in Mesopotamian religious life — a timeframe that comfortably predates and overlaps with the period of intensifying Greek contact with the Near East in the Archaic period.

The upshot for the kerykeion question: the deity and his snake-staff symbolism were circulating in Mesopotamia for well over a millennium before the Greek kerykeion crystallizes in the Archaic period, which is why the iconographic parallel carries real weight.

The pre-anthropomorphic argument

There is also a more structural argument about how divine symbolism evolves. It has been argued that the staff or wand entwined by two snakes was itself representing a god in the pre-anthropomorphic era — meaning the staff was Hermes, before the Greeks developed a fully human-figured deity. Like the herm or Priapus, it would thus be a precursor to the anthropomorphic Hermes of the classical period.

The dissenting view

It's worth noting this isn't a closed question. Lewis Richard Farnell (1909) argued that the two snakes simply developed out of ribbon ornaments on a herald's shepherd's crook, with no need for Near Eastern derivation at all. This purely Hellenic-origin view has fallen out of favor, but it reminds us that the Near Eastern connection, while widely accepted, rests largely on iconographic analogy and functional parallel rather than a documented transmission route.

The case rests on the very early Mesopotamian parallels, the striking similarities to Ningishzida's symbolism and role, the serpent's deep roots in Near Eastern religious iconography more broadly, and the prestige of Burkert's endorsement — but it remains an inference rather than a proven chain of transmission.

The development of this object reflects close trading relationships between Greece and the ancient Near East.

The Orientalizing Period

The critical window is what art historians call the Orientalizing Period, roughly 750–580 BCE. This is the period when art of the Eastern Mediterranean and the Ancient Near East heavily influenced nearby Mediterranean cultures, most notably Archaic Greece. The main sources of influence were Egypt, Phoenicia, Syria, and Assyria. The DMA kerykeion, dated to the early 5th century BCE, sits at the tail end of this intensive exchange period, which gives the transmission hypothesis real chronological plausibility.

Al Mina as a key conduit

One of the most important physical nodes in this network was the trading settlement of Al Mina on the Syrian coast. Modern research has shown that Al Mina was a significant eastern trading settlement with strong Greek connections dating back to around 825 BCE, and later work has considered it key to understanding the role of early Greeks in the east at the outset of the Orientalizing period. This gave Greeks direct, sustained access to a Levantine milieu that was itself deeply embedded in Mesopotamian cultural traditions.

The Phoenician intermediary problem

One important nuance is that the transmission was rarely direct. The Phoenicians were the great middlemen of the ancient Mediterranean, and much of what the Greeks received as "Near Eastern" came filtered through Phoenician commercial and cultural networks. This means the kerykeion iconography need not have traveled directly from a Sumerian temple context to a Greek bronze-caster — it could have passed through several layers of transmission, accumulating new associations along the way, which also makes it harder to prove a clean line of descent.

The mercenary connection

There's also a less discussed but significant human channel: during this period, the Assyrians advanced along the Mediterranean coast accompanied by Greek and Carian mercenaries, who were also active in the armies of Psamtik I in Egypt. Greek soldiers serving in Near Eastern armies would have had direct exposure to Mesopotamian religious iconography in ways that pure trade alone couldn't replicate.

So, the development of the kerykeion fits neatly within a much broader pattern of sustained, multi-channel contact between the Greek world and Mesopotamia across the 9th through 6th centuries BCE. The staff was made right at the moment when that exchange was at its most intensive.

However, amicable trade relations did not prevent political conflict. Prior to the 5th century, Greek elites spent money imitating eastern / Lydian dress and buying Persian art and goods. That commercial relationship didn't dissolve — it coexisted with, and was eventually overshadowed by, political confrontation driven by entirely different forces.

The structural trigger: Cyrus and the Ionian Greeks

The animosity didn't arise from trade grievances. It arose from imperial expansion colliding with Greek political autonomy. When Cyrus overthrew the Lydian kingdom of Croesus in 546 BCE, the Greek cities of Asia Minor found themselves under the rule of the Great King — a monarch absolute in religion, politics, and war, and the antithesis of city-state liberalism. This is the key structural fault line: Persian imperial governance was fundamentally incompatible with the Greek polis model of civic self-rule.

After the resulting Ionian Revolt and Persian Wars, Greek writers like Aeschylus and Herodotus actively forged a civilizational contrast between Greek freedom and Persian despotism that served Athenian political purposes. Unfortunately, this schism between East and West has continued into modern times.

If you enjoyed this post, never miss out on future posts by following me by email!

After the Eagles Left: The Last Romans at the Edge of Empire

by Mary Harrsch © 2026

I was listening to The Ancients podcast on Amazon Music and Dr. Rob Collins, Newcastle University, stated that not all the Roman armies left with Constantine III in 405-407 CE as evidence by a lack of coinage with his image beyond southern England. The traditional narrative holds that Constantine III took all of the mobile troops from Britain when he crossed to Gaul to confront Germanic invaders who had crossed the Rhine. However, I learned this is increasingly seen as an oversimplification.

A late Roman frontier officer in scale armor at a northern British fort, c. 5th century CE. The ridge helmet, mail armor, and absence of stirrups reflect equipment of the late Imperial period. AI-generated illustration created with Adobe Firefly. I had to change the helmet type, the armor type then remove stirrups, the horse's breast band and change the soldier's caligae to the enclosed boots more prevalent among frontier troops during that period to improve historical accuracy.

The distribution of Constantine III's coins is indeed geographically telling. The majority of late Roman silver siliquae are found in southern Britain, in the lowland zone to the south and east of the Fosse Way, with finds in Devon and Cornwall, Wales, the west Midlands, and the north-west and north-east being rare. Notable examples of later issues of Constantine III come specifically from Richborough and the Patching hoard in West Sussex. This southern concentration suggests that his monetary authority — and by extension his military presence — was felt primarily in the south, consistent with the theory that troops further north were not his to take.

Archaeological evidence supports the idea that not all Roman military presence ended in 407. The British-based units of the Field Army eventually left and are evidenced on the Continent, effectively "hollowing out" the province — but the northern frontier units were left largely untouched, as were tribal militia and foederati. The northern garrisons remained in place with their units and Roman commanders, eventually becoming the core of warbands in the 5th century.

This is further supported by earlier precedent: coins dated later than 383 have been excavated along Hadrian's Wall, suggesting that troops were not stripped from it during Magnus Maximus's earlier usurpation, or if they were, they were quickly returned.

The crucial nuance that Dr. Cottrell appears to be drawing on is the difference between the comitatenses — the mobile field army — and the limitanei, the static frontier garrison troops. What Constantine took was the Field Army led by the Comes Britanniarum, while the Saxon Shore was stripped to a skeleton staff and the northern frontier units were left largely in place. These latter troops, cut off from imperial pay after 409/410, didn't simply vanish — they evolved into the post-Roman military structures of sub-Roman Britain, which some scholars connect to the legendary traditions surrounding figures like Arthur.

So Dr. Cottrell's coinage argument is part of a broader and well-supported revisionist picture: the end of Roman Britain was not a single dramatic event in 407 but a gradual, geographically uneven process of disintegration.

The Physical Evidence: Birdoswald and the Transformation of the Forts

The most compelling archaeological case study is Birdoswald on Hadrian's Wall. At Birdoswald, the only change in the early 5th century appears to have been that the troops of the fort were no longer paid or supplied by central authority — the unit was still there. The old system of official coercion may have been replaced by a symbiosis, whereby the territory from which supplies had been drawn as part of the Roman tax system continued to sustain the fort in return for the assurance of protection in troubled times. The kind of commander-patronus attested by the large commanders' houses in the late forts may have continued to be an important figure as the 5th century went on — men who became imperceptibly more like chieftains in control of warbands than Roman commanders.

This interpretation is supported by excavation. Recent excavations at Birdoswald, led by director Tony Wilmott, suggest that a warband descended from the later Roman garrison may have continued to occupy the fort, supported by the discovery of a large timber hall which may have served as a gathering place for the post-Roman community. The timber hall is significant — it is a classic marker of early medieval chieftainly power, suggesting a direct architectural evolution from Roman fort headquarters to early medieval great hall.

Unlike the elite mobile field army, many of the static frontier troops — the limitanei — with their local loyalties didn't leave after direct Roman rule ended, and there is evidence that several forts were occupied for centuries after.

The transformation from Roman garrison commander to post-Roman warband leader was apparently gradual and organic rather than a sudden rupture. There is no evidence of any kind that the infantry of British leaders could be classed as "regular" after the middle or third quarter of the 5th century at the very latest. In other words, within a generation or two of 407, the disciplined Roman military structure had given way to something more characteristic of early medieval warfare — smaller, more personal warbands loyal to an individual leader rather than to an imperial institution.

The most tantalizing figure in this transition is Coel Hen — "Coel the Old" — who appears to straddle the Roman and post-Roman worlds. The senior military commander in the northern part of Roman Britain in the late Empire was the Dux Britanniarum and Prefect of the Sixth Legion, probably based at York, who commanded the garrisons of the forts on Hadrian's Wall. Some scholars have proposed that Coel Hen held, or inherited, precisely this role.

Historian John Morris in The Age of Arthur suggested that Coel may have been the last of the Roman Duces Britanniarum who commanded the Roman army in northern Britain, and split his lands among his heirs after his death. It must be noted, however, that Morris's broader thesis has been widely criticized by scholars, and Coel Hen himself is a semi-legendary figure whose historicity cannot be fully verified.

What is less disputed is that Coel Hen appears in the Harleian genealogies and the later pedigrees known as the Bonedd Gwŷr y Gogledd (The Descent of the Men of the North) at the head of several post-Roman royal families of the Hen Ogledd. His line, the Coeling, included such noted figures as Urien, king of Rheged; Gwallog, perhaps king of Elmet; the brothers Gwrgi and Peredur; and Clydno Eiddin, king of Edinburgh.

These figures represent the political legacy of whatever military structures survived in the north after 407. The Hen Ogledd — the "Old North" — covered modern northern England and southern Scotland, encompassing the kingdoms of Rheged, Elmet, Bryneich, the Gododdin territory extending well into Scotland, and other kingdoms in the Pennines. Rheged, one of the most celebrated of these kingdoms, was a Brittonic-speaking realm of the post-Roman era whose capital was likely Carlisle — itself a major Roman administrative and military centre.

The post-Roman kingdoms of the north map almost precisely onto the Roman military and administrative geography of the region, strongly suggesting that their rulers were indeed the descendants — biological or institutional — of the Roman garrison commanders who simply never left.

It is worth noting that this entire world of post-Roman northern warbands forms one of the primary historical contexts proposed for the Arthurian legends. The warrior culture of the Hen Ogledd, with its Roman military heritage, its Brittonic language, and its desperate resistance to Anglo-Saxon encroachment, is precisely the milieu that produced poems like Y Gododdin — one of the earliest texts to mention Arthur by name — and the bardic celebration of warrior heroes that would eventually crystallize into the Arthurian tradition.

If you enjoyed this post, never miss out on future posts by following me by email!

Diana's Curse, Meleager's Glory: The Boar as Elite Symbol in Imperial Rome

by Mary Harrsch © 2026

Exploring more of my images from the Dallas Museum of Art icame across this exquisite little bronze spout in the form of a boar's head, a testament to the extraordinary artistry that Roman craftsmen brought even to functional domestic objects. The carefully rendered bristle crest, alert ears, open mouth, and prominent tusks speak to a workshop of considerable skill, and the quality of execution points firmly to an elite household.

A bronze spout in the forma of a boar, Roman, 1st century CE
 📍 Dallas Museum of Art | Loan from the David T. Owsley Collection | Accession no. 156.1994.45

My research revealed the boar was one of the most richly layered animals in the Graeco-Roman symbolic imagination. The great mythological touchstone was the Calydonian Boar Hunt: King Oeneus of Calydon had failed to honor Diana in his annual harvest sacrifices, and the goddess, refusing to overlook the insult, dispatched a monstrous boar to destroy crops and terrorize the land. Meleager then assembled a band of heroes to hunt it and eventually killed it himself, though a quarrel over the spoils led to war and ultimately to his own death. This myth, told memorably by Ovid in the Metamorphoses, was enormously popular in 1st-century Rome and appears across sarcophagi, mosaics, and the decorative arts.

Hunting wild boar was a privilege reserved for the nobility, and representations of heroic hunts from the mythic past served to validate and glorify aristocratic hunting culture in the eyes of contemporaries. The wild boar held a special place in Roman culture as a symbol of strength, courage, and martial virtue — an association that extended to the emperors themselves. The emperor Hadrian is depicted on horseback about to strike a boar with a spear in the famous Hadrianic tondi on the Arch of Constantine, and his passion for the hunt was so celebrated that he founded a town in ancient Mysia — Hadrianotherae, meaning "Hadrian's Hunt" — to commemorate a particularly successful expedition in the neighborhood. For a wealthy Roman household, a boar-headed spout was thus simultaneously a mythological allusion, a celebration of the hunt, and a quiet assertion of aristocratic identity.

The Hercules connection would have added further resonance for any educated Roman viewer. The capture of the Erymanthian Boar was one of the Twelve Labors, and the 1st-century writer Heraclitus the Grammarian interpreted the myth allegorically, arguing that the boar represents human lack of self-restraint — making its conquest a symbol of heroic virtue and rational self-mastery.

As a spout, this piece would most plausibly have served on a fountain fitting in a domestic garden or atrium — Romans were famously fond of elaborate water features — or on a bronze ewer or mixing vessel used in a triclinium (dining room), where the boar imagery would have resonated perfectly with the culture of aristocratic feasting and the hunt. Liquid emerging from the open mouth, especially red wine, would have made the mythological reference delightfully literal.

If you enjoyed this post, never miss out on future posts by following me by email!

Power, Prestige, and Bronze: Elite Horsemanship in Pre-Etruscan Italy

by Mary Harrsch © 2026

This morning I was reviewing images I took at the Dallas Museum of Art back in 2006 and came across this ornate Villanovan bronze horse bit dating between the 8th and 7th centuries BCE. My research revealed this extraordinary bronze horse bit exmplifies the remarkable technical virtuosity and artistic ambition of Villanovan craftsmen working at the dawn of Etruscan civilization in central Italy.

Cast bronze Villanovan horse bit 
📍 Dallas Museum of Art | Gift of Mrs. John Leddy Jones | Accession no. 1969.6

Rather than simple functional forms, the Villanovan smith has populated every surface with cascading openwork zoomorphic figures — small stylized animals clambering over one another in interlocking composition — topped by elegantly curved arms terminating in distinctive disc-headed finials. Below the mouthpiece, pendant elements hang in layered chains, themselves decorated with miniature figural forms that would have caught light and produced subtle movement and sound when the horse was in action.

In 8th–7th century BCE Villanovan and Etruscan society, horses carried enormous prestige as markers of elite status, and this social weight translated directly into the elaboration of the bits themselves — objects that functioned simultaneously as practical tack and as public declarations of wealth and power. The significance of horses is underscored by the frequency with which bronze bits appear as grave goods in the large Villanovan cemeteries situated just outside major settlements.

Technically, the bit displays the hallmark Villanovan snaffle construction with a jointed cannon and large, architecturally complex cheekpieces rendered through sophisticated lost-wax casting. The density of figural decoration — with what appear to be horse, bird, and possibly anthropomorphic motifs woven together — aligns closely with the Geometric artistic vocabulary shared across the central Mediterranean during this period, while remaining distinctly Italic in character.

The Villanovans didn't acquire bronze casting from any one culture, but rather inherited and developed the tradition through several overlapping channels:

Central Italy saw continuous development from the Bronze Age through the Villanovan period (900–700 BCE) into the Etruscan era — local evolution rather than foreign migration or wholesale cultural replacement. However, Bronze-working knowledge was already present in the Italian peninsula before the Villanovan period properly began.

Trans-Alpine and Balkan connections were significant early on. Bronzeworks indicate contact with Sardinia, central Europe, and the Balkans, and these links brought about a more advanced metallurgy.

Greek contact then accelerated sophistication. The culture came into contact with the wider Mediterranean via the arrival of Greek settlers, especially Euboeans, in the mid-8th century BCE, who were eager to exploit the mineral-rich region that would become Etruria.

Elite demand drove innovation. The great variability of formal features and techniques in Villanovan bronze production — from cast fibulae and weapons to hammered sheet bronze armor and vessels — is closely connected to the display expressions of new elites within proto-urban communities. In other words, the social ambitions of a rising aristocratic class pushed craftsmen to ever greater levels of elaboration

This piece represents a cultural threshold: by around 750 BCE the Villanovan culture had transitioned into the Etruscan culture proper, making objects like this living witnesses to that transformation. It is simultaneously a functional tool, a status object, and one of the finest examples of small-scale sculpture produced in pre-Roman Italy

📍 Dallas Museum of Art | Gift of Mrs. John Leddy Jones | Accession no. 1969.6

If you enjoyed this post, never miss out on future posts by following me by email!