All About Glass

You are here

A Recently Discovered Cage Cup

All About Glass

This paper describes and discusses a late Roman cage cup that appeared on the market in 1986 and was acquired by The Corning Museum of Glass in 1987.1 The hemispherical glass cup has metal fittings which indicate that, at the time of burial, it was meant to be suspended. This raises the possibility that hemispherical cage cups (and perhaps other cage cups, too) were hanging lamps rather than drinking vessels.


Figure 1The ensemble has three components: (1) a glass cage cup, (2) a metal collar, and (3) a metal lamp-hanger (Fig. 1).

1. Cage cup (Figs. 2–4). H. 7.1 cm, D. 12.1 cm.  Almost colorless glass with very small bubbles. Blown or cast, wheel-cut, polished.

Hemispherical body; outsplayed rim with ground lip and raised molding slightly above bottom. Decorated with openwork band at base of neck and openwork cage enclosing body. Band consists of ovolo frieze with 59 egg-shaped perforations separated by darts; cage consists of central mesh surrounded by two rings of meshes and concentric border. Central mesh, evidently circular, was attached to body by seven bridges. Inner ring has seven heart-shaped meshes, and outer ring has 14 oval meshes. In both rings, small cruciform element at junction of each pair of meshes conceals bridge; serrated bar at upper end of each heart-shaped mesh also conceals bridge, as does V-shaped element at upper extremity of each oval mesh, where it meets border. Around border, between each pair of meshes, one additional bridge.

Body intact. Extensive damage to openwork decoration: on frieze, short section extending across three perforations is missing, with broken edges ground smooth; frieze also chipped, with many small cracks; most of central mesh and five bridges missing; in inner ring, six meshes and most of their bridges partly or wholly missing; in outer ring, one-half of one mesh missing. Thin film of grayish weathering and slight iridescence; surface, however; has been cleaned.

2. Collar. D. 11.0 cm, H. 0.5 cm, TH. 0.1 cm. Presumably copper alloy.

Single strip with one end terminating in perforated tongue and opposite end terminating in two perforated tongues, one above the other; tongues fastened with pin with diamond-shaped head, bent in position. At regular intervals around collar, three horizontal flanges (L. 1.1 cm, W. 0.8 cm, TH. 0.1 cm), each with circular perforation.

Covered with grayish green corrosion products, which firmly attach pin to collar.

3. Lamp-hanger. L. (hook-and-loop) 17.0 cm, (looped elements) 18.8 cm. Presumably copper alloy.

Object comprising hook-and-loop and three looped elements. Hook-and-loop has upper end bent into hook, long shank (D. 0.2 cm), and lower end looped and wound six times around shank. Each looped element has straight rod (D. 0.2 cm) flattened and perforated at ends, six twisted figure-of-eight links, and hook-and-loop.

Dark brown to gray surface with some greenish corrosion.

Figure 2Figure 3Figure 4


The cage cup is the latest addition to a group of late Roman cut and polished glass vessels with openwork decoration, of which the Lycurgus Cup is the most dramatic and distinguished example. In 1959, Harden and Toynbee divided these objects into two groups: (A) with figured decoration, with or without cages and inscriptions, and (B) without figured decoration.2 Group B includes at least four forms: beakers, some of which have an openwork inscription, usually a toast; a bucket, the cage cup from Soria (ancient Termes) being the only known example; a shallow bowl (see below); and hemispherical bowls, including the cage cup in question.

The "new” cage cup, therefore, is not without parallel. The closest analogy is the Constable-Maxwell cage cup, at present on loan to The Corning Museum of Glass.3 Indeed, the only differences between these objects are attributable to the difference in size, the Constable-Maxwell cage cup (D. 18.4 cm, H. 10.0 cm) being half again as large as the “new” cage cup. As a result, the ovolo frieze has 103 perforations instead of 59, and the cage has three concentric rings instead of two, the outermost ring containing 14 circular meshes. In all other respects, the objects are alike. The forms are identical, and in both cases the central meshes are circular and have seven bridges; the inner/innermost rings have seven heart-shaped meshes, and the outer/central rings have 14 oval meshes. Such is the degree of similarity that I am tempted to suppose that the two cage cups were made in the same workshop.

A second parallel is provided by the fragmentary cage cup (now lost, presumably destroyed) from Hohensülzen, near Worms, in the Federal Republic of Germany.4 The fragments came from a hemispherical bowl with a splayed rim (D. 21.0 cm), which Harden and Toynbee described as follows: “The body and base of the vessel are covered by a network cage consisting of four rows of circular meshes with a central openwork 8-petalled rosette.'' However, with the first description of the fragments, which were found in 1869, Wieseler published a lithograph (reproduced by Kisa and others) showing five rows of meshes but without the mesh at the center. Similarly, in 1922-1923, Morin-Jean published a drawing of the fragments restored as a cage Cup with five rows of meshes. The drawing, which the author signed and dated "Sept. 1912," does not attempt to indicate the form of the central mesh.

These early accounts are supported by a plaster copy of the fragments, made by Lindenschmit in the last century and reproduced in plastic at the Römisch-Germanisches Zentralmuseum, Mainz, in 1980 for the exhibition "Gallien in der Spätantike." In fact, the description of the central mesh as a rosette with eight petals appears to be derived not from the fragments themselves, but from a reconstruction published by Behrens in 1925–1926. In any case, the Hohensülzen cage cup had a splayed rim, hemispherical body, and cage. Despite the absence of an ovolo frieze, it is a valid parallel for the "new” cage cup.

Thus, the Constable-Maxwell cage cup is no longer "the only surviving example of its type,”5 and it is now apparent that hemispherical bowls form as distinct a subgroup of Harden and Toynbee's Group B as do the beakers. Concerning their date, the best evidence is still the Hohensülzen bowl, which was found in a stone sarcophagus, together with three other glasses: (r) a two-handled cylindrical flask with a facet-cut Dionysiac scene, (2) another two-handled cylindrical flask with horizontal linear-cut bands, and (3) a pipette-shaped toilet bottle. As Fremersdorf observed, these objects suggest that the sarcophagus was buried about A.D. 300 or slightly later.6

On the “new" cage cup, the metal collar fits closely around the neck and the pin is corroded in position. The collar, therefore, was already attached to the cage cup at the time of burial.7 The three perforated flanges are difficult to explain other than for suspending the object from a lamp-hanger. Thus, the identification of the "new'' cage cup as a hanging bowl is independent of the lamp-hanger itself, whose lightly corroded surface presents a different appearance from that of the collar, hinting at the possibility that (regardless of its age) it may be a recent addition to the ensemble.8

The conclusion that the "new" cage cup was used as a hanging bowl does not, of course, establish that it was made expressly for this purpose. The possibility, however, is attractive. While the majority of beaker-shaped cage cups have inscriptions, usually associated with drinking, none of the hemispherical bowls is inscribed. Moreover, their wide, splayed rims make the bowls unsuitable for drinking or pouring. I suggest, therefore, that the makers of the hemispherical cage cups intended them to be suspended.

Hanging Lamp, Blown, stamping, Roman Empire, 300-499. 61.1.1.Hanging bowls were widely used in late antiquity, and examples made of glass are not uncommon. The Römisch-Germanisches Museum, Cologne, for example, has two such objects: (1) a dark green, cast, cut, and polished bowl (D. 22.5 cm) with applied mold-pressed medallions, from Horrem in Rheinland, and (2) a similar object with medallions, apparently pressed in the same mold, from Cologne itself.9 A hanging bowl in The Corning Museum of Glass also has applied medallions.10 The "new" cage cup and, by implication, the other hemispherical cage cups therefore take their places in a well-known but diverse class of objects: glass hanging bowls.

The function of late Roman glass hanging bowls has not been established. However, descriptions and illustrations in contemporary manuscripts show that similar vessels served as lamps in Byzantine and Islamic contexts.11 It is reasonable to suppose that the Roman examples also were lamps.

It is necessary, however, to add a note of caution, for not all bowl-shaped cage cups had this function. Among the finds from the Kaiserthermen at Trier is one small fragment from a bowl with a short, outsplayed rim (D. about 21 cm).12 The bowl, which was markedly shallower than the objects discussed above, had an openwork inscription at the base of the rim and an openwork cage. The fragment contains the letters AS and part of the border of the cage, including five bridges. The most obvious (but by no means the only) interpretation of the letters is that they were part of a toast, such as that on the Trivulzio cage cup: "[BIBE VIV]AS [MULTIS ANNIS]” ("Drink! May you live for many years.'').13 If the inscription was a toast, the object was presumably used in connection with drinking.

While the inscribed fragment from Trier suggests that not all bowl-shaped cage cups were intended for suspension, the Cagnola or "Masks and Columns” cage cup at Varese suggests that not all beaker-like cage cups were connected with drinking. This unique object has a bell-shaped body with a cage composed of columns alternating with tragic masks. The area above the cage consists of an outsplayed rim with ground lip and raised molding at the bottom of the neck, below which there is a gap of almost three centimeters above the border of the cage.14 When one compares the Cagnola cage cup with the "new” cage cup, it is tempting to suppose that the gap had a function: to receive a metal collar. In other words, I suspect that this, too, may have been a lamp. Indeed, the "new'' cage cup poses a provocative question: not “how many cage cups were used for drinking?'', but "how many cage cups were hanging lamps?"

This article was published in the Journal of Glass Studies, Vol. 30 (1988), 28–33.

1. Accession no. 87.1.1.

2. D. B. Harden and J. M. G. Toynbee, "The Rothschild Lycurgus Cup," Archeologia, v. 97, 1959, pp. 179–212; D. B. Harden, "The Rothschild Lycurgus Cup: Addenda and Corrigenda," Journal of Glass Studies, v. 5, 1963, pp9–17; D. B. Harden et al., Glass of the Caesars, Milan, 1987, pp. 185–186 and 238–249.

3. L151.1.85. The Constable-Maxwell Collection of Ancient Glass, sale catalog, Sotheby Parke Bernet, London, June 4–5, 1979, pp. 38–41, lot 41; Harden et al. [note 2], p. 242, no. 136.

4. Discussion of this cage cup can be found in the following: F. Wieseler, " Römische Gläser gefunden in Hohen-Sülzen,” Jahrbücher des Vereins von Alterthumsfreunden im Rheinlande (now Bonner Jahrbücher), v. 59, 1876, pp. 64–87 and pl. 2; A. Kisa, Das Glas im Altertume, Leipzig, 1908, v. 2, p. 461, fig. 222, and v. 3, pp. 608 and 621; Morin-Jean, La Verrerie en Gaule, Paris, 1922–1923, p. 233, fig. 314; G. Behrens, "Römische Gläser aus Rheinhessen," Mainzer Zeitschrift v. 20/21, 1925–1926, pp. 62–77, especially pp. 75–76 and fig. 29 (reconstruction); W. A. Thorpe, "The Prelude to European Cut Glass," Transactions of the Society for Glass Technology, v. 22, 1938, pp. 5–37, especially p. 34, no. 9; Harden and Toynbee [note 2], pp. 208–209, no. B4; M. Schulze in Gallien in der Spätantike, Mainz, 1980, p. 69, no. 59. I am grateful to Dr. Mechthild Schulze-Dörrlamm of the Römisch-Germanisches Zentralmuseum, Mainz, for information on the copies of the Hohensülzen fragments.

5. The Constable-Maxwell Collection [note 3], p. 41. Two other fragments, both extant, may come from similar objects:
A. Trier, Federal Republic of Germany. Part of outsplayed rim (D. about 15 cm) and openwork ovalo frieze. Because of the small size of the fragment, however, it is impossible to determine whether it came from a beaker or a hemispherical bowl. Rheinisches Landesmuseum, Trier, no. 14588. See Harden and Toynbee [note 2], p. 207 and pl. 69g; W. Reusch, Trierer Zeitschrift, v. 32, 1969, p. 313ff. and pl. 3; K. Goethert-Polaschek, Katalog der römischen Gläser des Rheinischen Landesmuseums Trier (Trierer Grabungen und Forschungen, 9), Mainz, 1977, p. 63, no. 239; W. Binsfeld in Trier. Kaiserresidenz und Bischofssitz, Mainz, 1984, p. 132, no. 45.
B. Tác-Fövenypuszta, Fejér county, Hungary. Part of outsplayed rim, openwork ovolo frieze, and one bridge from border of openwork cage. Again, it is not clear whether the object was a beaker or a bowl. See E. B. Thomas, "Die römerzeitliche Villa von Tác-Fövenypuszta," Acta Archaeologica Academiae Scientiarum Hungaricae, v. 6, 1955, pp. 79–152, especially p. 127 and pl. 31, nos. 3–4; O. Doppelfeld, "Das Kölner Diatretglas und die anderen Netzdiatrete," Gymnasium, v. 68, no. 5, 1961, pp. 410–424; Harden [note 2], pp. 9–17, no. B 17.

6. F. Fremersdorf, Figürlich geschliffene Gläser, einer Kölner Werkstatt des 3. Jahrhunderts (Römisch-Germanische Forschungen, 19), Berlin, 1951.

7. It is possible that the damage to the ovalo frieze (Fig. 3) was sustained when the collar was attached; the subsequent smoothing of the broken edges certainly occurred in antiquity, as the presence of weathering attests.

8. For a recent classification of Byzantine lamp-hangers, see M. V. Gill, "The Small Finds," in R. M. Harrison, Excavations at Saraçhane in Istanbul, Princeton: Princeton University Press and Dumbarton Oaks Research Library and Collection, 1986, p. 239. Our example approximates Gill's Type A: "[Lamp-hangers] of Type A are the most well made, having loops and hooks more precisely shaped, and being constructed of a wire a little more substantial than the other groups (diam. approx, 0.2 cm)."

9. T. E. Haevernick, "Beiträge zur Geschichte des antiken Gläses. IV: Gefasse mit vier Masken," Jahrbuch des Römisch-Germanischen Zentralmuseums, Mainz, v. 7, 1960, pp. 53–56; F. Fremersdorf, Römisches Geformtes Glas in Köln (Die Denkmäler des römischen Köln, 6), Cologne, 1961, pp. 69–70, pl. 139; F. Fremersdorf, Die römischen Gläser mit Schliff, Bemalung und Goldauflagen aus Köln (Die Denkmäler der römischen Köln, 8), Cologne, 1967, pp. 106–107, pl. 107; Harden et al. [note 2], pp. 204–205, no. 113.

10. Accession no. 61.1.1. "Recent Important Acquisitions," Journal of Glass Studies, v. 4, 1962, pp. 139–149, especially p. 141, no. 14.

11. In his description of the lighting of Sancta Sophia, Paul the Silentiary noted both polycandela and single hanging lamps: "Near the aisles, too, alongside the columns, they have hung in order single lamps apart one from another; and through the whole length of the far-stretching nave is their path. Beneath each they have placed a single vessel, like a balance pan, and in the centre of this rests a cup of well-burning oil"(W. R. Lethaby and H. Swainson, The Church of Sancta Sophia, Constantinople, London and New York, 1894, p. 50). Two vessels in the Treasury of San Marco, Venice, retain the metal collars with which they were suspended from lamp-hangers: a fourth-century rock crystal bowl decorated with sea creatures, to which a silver-gilt collar was added sometime between the 10th and the 12th centuries (H. R. Hahnloser et al., Il Tesoro di San Marco, Florence, 197I, v. 2, no. 50), and a shallow glass bowl with relief-cut disks and bosses, of uncertain date, with an 11th-century silver-gilt collar (Ibid., no. 67). For Byzantine and Islamic hanging lamps in general, see J. Philippe, Le Monde Byzantin dans l'histoire de la verrerie (Ve–XVIe siècle), Bologna, 1970, p. 77, fig. 41; C.J. Lamm, Mittelalterliche Gläser und Steinschnittarbeiten aus dem Nahen Osten, Berlin, 1929, v. 1, pp. 93–94, nos. 16–20, v. 2, pl. 28, and many others.

12. Binsfeld [note 5], p. 129, no. 44.

13. G. D'Adda, Ricerche sulle arti e sull'industria romana. Vasa vitrea diatreta, Milan, 1870, p. 28; Kisa [note 4], v. 3, p. 607; M. Bertolone, "La tazza vitrea diatreta Cagnola," Rivista archeological della antica Provincia e Diocesi di Como, v. 128–129, 1947–1948, pp. 31–35, especially p. 34, figs. 4–5; Harden and Toynbee [note 2], p. 209, no. B6; Doppelfeld [note 5], pp. 410–424, especially p. 421; L. Cassani, Repertorio di antichità preromane e romane rinvenute nella provincia di Novara, Novara, 1962, p. 133, fig. 24; A. M. Tamassia and M. Mirabella Roberti, Mostra dei vetri romani in Lombardia, Milan, 1964, p. 16; G, M. Facchini, La diatreta Trivulzio, Civico Museo Archeologico di Milano, Scheda no. 11, Milan, 1979; E. Roffia, "Le Verre en Italie. Région: Lombardia, Milano," Bulletin de l’AIHV, v. 9, 1981–1983, p. 55, fig. 3; Harden et al. [note 2], pp. 238–239, no. 134.

14. D'Adda [note 13]; Kisa [note 4], v. 3, pp. 609–610, 629, fig. 228; Thorpe [note 4], p. 33; Harden et al. [note 2], p. 243, no. 137.

Published on March 20, 2013