Diatretbecher

A colorless glass beaker 12.1 cm high, encased in a free-standing network of dark red, yellow and green glass cut so that it appears to float around the cup on short struts, with a Greek drinking toast, Drink, live well, forever, cut in purple-red letters below the rim. It was found in 1960 in a sandstone sarcophagus in a Roman cemetery on the edge of Cologne, one of a documented handful of cage cups, a small and technically extraordinary class of late Roman luxury glass, to survive at all.

How its cage was actually cut or formed is a genuine, still-open dispute in the scholarship: cold-ground from a single solid blank, as the classic theory holds, or shaped with the help of a mould, as a rival theory argues. Neither side has fully replicated the ancient process, and modern glassmakers who have tried have not settled the question. This is a separate object from the more famous Lycurgus Cup in the British Museum, which belongs to the same broad class but is a distinct, dichroic, colour-changing piece; nothing in the sources consulted suggests the Cologne cup shares that optical property.

The find itself raises its own unresolved questions: which of the burials in the cemetery the cup belonged to, why so costly an object sits, on one account, among otherwise modest grave goods, and where the workshop that made it actually stood.

What it is

The object is a small drinking cup, 12.1 centimetres high, blown or worked from colorless glass and then encased in a second, separate-looking layer: an openwork network, or cage, of dark purple-red, gold-yellow and emerald-green glass cut so that it stands proud of the inner cup's surface, touching it only at a series of short colorless struts. Seen from any angle, the cage reads as a basket of coloured glass hovering a few millimetres off the vessel it protects.

Below the rim, in purple-red lettering worked into the same technique, runs a Greek inscription: PIE ZESAIS KALOS AEI, a drinking toast conventionally rendered Drink, live well, forever. The combination of a tricolour network cage with a colour transition worked into the connecting struts is, on the account of the museum that holds it, without an exact parallel among the surviving corpus: the piece is described by the Cologne city museums as the only well-preserved three-colour network diatretum of its kind known anywhere, with a pale green net-basket and a yellow collar shading into the strut material.

How the cage was cut, or cast: a dispute nobody has settled

Cage cups have been known and discussed in print since 1680, and for most of that time the working assumption was straightforward: a glassworker took a single thick blank of glass and, over what must have been weeks of patient cold-cutting and grinding, removed everything except the inner cup and the network of struts and cage that surrounds it. In 1956 the German scholar Fritz Fremersdorf, at the time associated with Cologne's own Roman-Germanic collections, set this account out in its classic form, and it remains the position most handbooks still give.

Rosemary Lierke proposed a rival account in the 1990s: that the vessel was pressed, using a perforated inner mould through which molten glass was forced to form the struts, with cutting and grinding applied afterward to refine the surface rather than to create the whole form from scratch. The case for a mould-assisted or two-part process rests partly on cups, including this one, that show smooth-looking joins where cage meets cup, and partly on the practical difficulty of freehand-cutting struts as fine and as regular as those that survive. Against it stands a fragmentary cup excavated at Corinth in the 1960s, examined under a microscope with no detectable joint at all where its cage meets its body, which reads as evidence for solid cold-working rather than assembly.

The most direct evidence bearing on this specific cup comes from its own excavator. Otto Doppelfeld, who published the Cologne find in 1961, used macro-photography to record what he described as colour transitions, Farbwechsel, at the points where the struts meet the cage, an observation some later writers have read as supporting some form of hot-working or fusion rather than pure cold-grinding of a single blank, though Doppelfeld's photographs alone do not settle the question either way.

David Whitehouse, with William Gudenrath and Paul Roberts, assembled the fullest study of the whole class, published by the Corning Museum of Glass in 2015 after Whitehouse's death from cancer in February 2013; his colleagues completed the manuscript and catalogue from his notes. Whitehouse examined Lierke's pressing theory in detail and was not persuaded by it; Lierke responded at length in a 2016 review. A 2020 review of the book in the Journal of Roman Archaeology was titled, without irony, Diatreta (cage cups): the debate continues. That title is the most accurate one-line summary available of where the question still stands.

Modern attempts to settle the question by replication have not done so either. An early German replica made in 1906 was toasted with champagne on completion and broke at the rim when the maker went to drink from it. The German experimental glass engraver Josef Wenzel has since reproduced several historic cage cups, including this Cologne example, by first shaping a thick-walled cup, applying the coloured glass on the outside as bands and plates, and then carving the network free, a hybrid method consistent with Rhineland gem-cutting traditions rather than a straightforward confirmation of either the pure-grinding or pure-moulding camp. No modern reconstruction has been accepted by all sides as a faithful reproduction of the ancient technique, and this record does not attempt to resolve a dispute the specialist literature has not resolved.

The grave at the edge of Roman Cologne

The cup was found in 1960 during work at the corner of Stolberger Strasse and Eupener Strasse, on the boundary between the Cologne districts of Braunsfeld and Mungersdorf, on the site of a Roman villa rustica, a rural estate, with its own cemetery. Excavation there uncovered five red sandstone sarcophagi, one further inhumation burial, and a separate cremation grave; the cage cup came from one of the sarcophagi, recorded by the photographer whose image accompanies this record as grave 5 of the family cemetery. The excavation was published the following year, 1961, by Otto Doppelfeld, then director of the Romisch-Germanisches Museum, in the Kolner Jahrbuch fur Vor- und Fruhgeschichte; that report, rather than any of the secondary summaries used here, is the primary source a researcher should consult for the full grave inventory and stratigraphy, and this record has not been able to read it directly.

One secondary account states that the rest of the grave goods found with the cup were comparatively mundane, which, if accurate, would make the presence of an object of this cost and technical difficulty in the same burial a genuine puzzle worth further study. A separate account of the associated cremation grave at the same cemetery describes a lead vessel, ceramic jugs and a folded beaker as accompanying goods, which may or may not be the same grouping being described as mundane; this record cannot reconcile the two secondary accounts without the primary excavation report, and flags the discrepancy rather than resolving it. A dedicated scholarly article exists specifically on the problem of assigning individual grave goods to individuals within this late-antique grave, which indicates that the question of whose grave, precisely, and whose cup, precisely, has itself been treated as unresolved in the specialist literature, not merely by this record.

A toast carved in glass

The Greek inscription on the Cologne cup belongs to a well-attested genre: a great many cage cups and other pieces of Roman luxury tableware carry short convivial toasts, wishing the drinker long life, good health or good fortune, meant to be read aloud or simply enjoyed at the table where the object was used. Comparable formulas survive on other cage cups in Latin: the Munich cup, also found in Cologne but a separate and later object, reads BIBE MULTIS ANNIS, drink for many years; the Trivulzio cup in Milan reads BIBE VIVAS MULTIS ANNIS, a fuller version of the same wish; the Pljevlja cup in Montenegro carries VIVAS PANELLENI BONA, and the Autun cup found in France in 2020 carries VIVAS FELICITER, live happily. The Cologne cup's Greek phrasing rather than Latin is itself a small data point in the debate over where such objects, and such customers, sat within the wider Roman world of the Rhine frontier, a frontier province where Greek-language luxury goods were not the ordinary case.

These were not funerary inscriptions composed for the grave. They were dining-room and banqueting-hall phrases, meant for use, which makes their presence in a burial a matter of a valued possession being sent into the ground with its owner rather than an object made for the purpose.

What survives, and how much

Cage cups are one of the smallest well-defined classes of ancient luxury object. The fullest inventory available, David Whitehouse's 2015 catalogue, lists sixty-nine vessels and fragments, with a further thirteen added in an addendum compiled by his colleagues after his death, for a documented total in the low eighties; looser figures circulating in more popular sources put the number at about fifty cups or fragments, and the two figures should be read as different counts by different methods rather than as evidence that one is wrong. Whichever total is used, only a small minority survive complete or near-complete, and the great majority of what is known is fragmentary.

Named comparanda span several countries and several sub-types of the class. Beaker-shaped, drinking-vessel cage cups include this Cologne example; the Trivulzio cup in the Museo Archeologico, Milan, described in the literature as the only example with no damage at all, recovered from a sarcophagus near Novara in 1675; the Munich cup, held by the Staatliche Antikensammlungen but excavated in Cologne, a separate and later Latin-inscribed piece; the Daruvar network beaker in the Kunsthistorisches Museum, Vienna, found in 1785 and incomplete; the Piesport-Niederemmel cup in the Rheinisches Landesmuseum, Trier, found in a sandstone sarcophagus in 1950, another Rhineland grave find; the Pljevlja cup in Montenegro's Heritage Museum, found in 1975; and the Autun cup in the Musee Rolin, discovered as recently as 2020 and reported to have contained traces of ambergris, evidence of what such a vessel actually held. Bowl-shaped cage cups, thought in some cases to have served as hanging lamps rather than drinking vessels, include an example at the Corning Museum of Glass and the Constable-Maxwell cup, which remains in private hands.

A separate, rarer sub-type carries figural, narrative scenes rather than geometric network patterns cut directly into dichroic glass; of that sub-type, the Lycurgus Cup in the British Museum is the only complete survivor, with other figural examples known only from fragments, including a piece from the Begram hoard in Afghanistan showing the Pharos of Alexandria.

A separate object from the Lycurgus Cup, and not established to be dichroic

This record concerns the Cologne diatretum alone. It is not the Lycurgus Cup, and it is not the Portland Vase; all three are distinct objects in Iconotheca's records, made by different techniques or, in the Lycurgus Cup's case, a different sub-type of the same broad cage-cup class, with their own separate find histories.

The Lycurgus Cup is dichroic: lit from the front it reads as an opaque jade green, and lit from behind or through the glass it glows a translucent ruby red, an effect produced by gold and silver metal nanoparticles suspended in the glass, identified through scientific analysis by Ian Freestone and colleagues. Every source consulted for this record describes that property as belonging specifically to the Lycurgus Cup, calling it the only complete ancient glass object known to show it, with only a very small number of other, non-vessel dichroic glass fragments known worldwide. Nothing in the sources consulted claims or implies that the Cologne cup, or any of the network-type comparanda listed above, shares this optical property; its cage is built from opaque coloured glass in discrete zones of red, yellow and green rather than a single dichroic glass body, and this record does not assume otherwise.

Recent scholarship: a possible maker's mark on the cage

In February 2023, the art historian and glassblower Hallie Meredith, examining cage cups in a private collection at the Metropolitan Museum of Art, turned one vessel over and noticed small abstract symbols, diamonds, leaves and crosses, cut into the reverse of the openwork cage alongside its drinking-toast inscription. Her subsequent research, published in the Journal of Glass Studies in April 2025 and in World Archaeology in October 2025, argues that these symbols are workshop identifiers, a shared visual language marking which glass workshop produced a given piece, rather than personal signatures or pure ornament, and that cage cups were the output of organized, multi-person workshops operating across the 4th to 6th centuries CE rather than isolated individual masterpieces.

Press coverage of Meredith's work describes examining a vessel bearing the exact Greek inscription and date range of this Cologne cup, PIE ZESAIS KALOS AEI, dated 300 to 350 CE, among the pieces studied, alongside a separate Cologne-excavated but Latin-inscribed vessel (matching the Munich cup) and a Corning Museum piece. None of the popular-press accounts consulted for this record names the Romisch-Germanisches Museum explicitly as the holder of the Greek-inscribed vessel Meredith examined, so this record treats the identification as probable, given the exact match of inscription and date, but unconfirmed, and recommends checking Meredith's original journal articles directly to establish whether this specific object carries one of the newly identified maker's marks.

What such an object was worth, and what it was for

Cage cups were display and drinking vessels for elite dining, made by a process that, on any version of the manufacturing debate above, demanded weeks of a highly skilled worker's time on a single object that could shatter at any stage of the cutting. That combination of material cost, labour cost and fragility is itself the evidence for what such a cup signalled about its owner's standing, independent of which manufacturing theory is eventually vindicated.

No sale price exists for this specific cup; it has been public property since its excavation in 1960 and has never been on the market. As a point of comparison only, and not as a valuation of this object, a bowl-shaped cage cup of the same broad class, the Constable-Maxwell cup, sold at auction in 2004 for 2,646,650 pounds sterling, a price reported as a world record for a piece of Roman glass; that sale concerns a different, privately owned object and is cited here solely to indicate the order of value this class of object can command when one does reach the market.

In public ownership since the day it was found

Unlike several of its comparanda, the Cologne cup has no market history to record. German law places archaeological finds of this kind under state and municipal heritage protection, and the cup passed directly from Doppelfeld's 1960 excavation into the Romisch-Germanisches Museum's own collection, where it has remained under inventory number 60,1 ever since. There has been no export question, no private ownership, and no auction. The museum itself describes the piece as the outstanding item within what it calls the world's largest collection of Roman glass of the 1st to 4th centuries CE, a claim made by the institution about its own holdings and recorded here as such rather than independently verified against rival collections.

What is still unresolved

Three separate questions about this object remain genuinely open in the published literature rather than merely under-researched by this record. How the cage was actually formed, by freehand cold-cutting of a solid blank or by some mould-assisted or two-part process, has been argued in print since the nineteenth century and was still being argued, explicitly and by name, in a 2020 review of the most comprehensive study yet published. Which burial within the Braunsfeld cemetery the cup truly belonged to, and to which individual within that burial, has its own dedicated scholarly literature on the general difficulty of assigning grave goods to persons, without a settled answer for this grave in the sources this record could reach. And whether this piece was made on the Rhine or carried there from elsewhere in the Roman world remains argued rather than demonstrated on either side.

None of that uncertainty is a defect in the object or in the record of it. A cup this rare, cut by a method its own excavator could not fully explain from the evidence of his own photographs, sitting in a grave whose full inventory has not been re-examined here against the primary report, is exactly the kind of object where stating the open questions plainly is worth more than a smoothed-over story.