Glass in Roman York

An Inventory of the Historical Monuments in City of York, Volume 1, Eburacum, Roman York. Originally published by Her Majesty's Stationery Office, London, 1962.

This free content was digitised by double rekeying. All rights reserved.


'Glass in Roman York', An Inventory of the Historical Monuments in City of York, Volume 1, Eburacum, Roman York, (London, 1962), pp. 136-141. British History Online [accessed 16 June 2024].

. "Glass in Roman York", in An Inventory of the Historical Monuments in City of York, Volume 1, Eburacum, Roman York, (London, 1962) 136-141. British History Online, accessed June 16, 2024,

. "Glass in Roman York", An Inventory of the Historical Monuments in City of York, Volume 1, Eburacum, Roman York, (London, 1962). 136-141. British History Online. Web. 16 June 2024,


The following account is contributed by Dr. D. B. Harden; the accompanying drawings are by Mrs. Marion Cox and Dudley Waterman; the photographs are the Commission's. The identity numbers prefixed by H. or H.G. refer to pieces in the Yorkshire Museum (Y.M.).

The story of glass in Roman York corresponds extraordinarily well with the historical development of the fortress and town sketched in the Introduction (p. xxix). It reflects, in particular, first the period of primarily military occupation from A.D. 71 to the closing years of the 2nd century; then two centuries or more during which York's prosperity as a civilian and military centre was more or less unbroken and during which the city twice became an imperial residence, first from 208 to 211 and secondly at the beginning of the 4th century.

There is a distinct group of late 1st and 2nd-century material to represent the earlier, military phase, but the great majority of the glass belongs to the 3rd and 4th centuries and is spread fairly evenly over that period, though with a probable predominance of late over early pieces. Unfortunately the exact find-spots of over half the material are no longer known and the problem has been further complicated since certain non-local specimens have over the course of years been mixed up with the local ones in the Yorkshire Museum and they can only now be sorted out with difficulty, and without great assurance, on a typological basis. Most of the provenienced pieces come from the Railway Cemetery or from tombs on The Mount.

Late 1st and 2nd century

In the south of England in Claudian and Flavian times one of the most prevalent shapes of glass appears to have been the pillar-moulded bowl. Fragments of these bowls occur also in some quantity at York, mainly in transparent green and dark blue glass (see H.G. 176. Fig. 89). It is known that the type died out about the turn of the 1st century. These York fragments, therefore, represent the earliest years of the fortress's existence. There is no record of any of them being found in graves, but one (H.G. 252) is from S. N. Miller's excavations on the N.E. side of the fortress in 1925 (JRS, XV (1925), 182, fig. 96, no. 5) and Raine speaks of pieces of them from 'under the Exhibition Building, 1878' (YMH, 104, item j).

Only eight other fragments are certainly of 1st-century date and none of them unfortunately has an exact find-spot. Two are pieces of mosaic glass, one from a bowl (H.G. 203), the other (H.G. 168) a piece of flat inlay of green and yellow, the commonest and most long-lived variety of this fabric. A third (H.G. 300) is a fragmentary flasklet (Fig. 89) of blue glass of a type found frequently in 1st-century graves on the continent. A fourth (H.G. 273) is a piece of a mould-blown beaker with a pattern of almond-shaped knops alternating with dots (Fig. 88). It belongs to a family well represented at Pompeii and was found by S. N. Miller in 1925. The remainder are all of considerable interest. Two are handles of colourless glass (H.G. 222–3) from different cups of a type of fine ware that was probably cast in a mould and then finished by grinding the surface, handles and all (Fig. 88). Another (H.G. 192) is a stemmed base (Fig. 88) belonging to the same type as these handlefragments, it seems, when we compare a complete piece from Canosa, Italy, in the British Museum. The type has been dated to Hellenistic times and it certainly derives from Hellenistic metal models, yet, if so, it must have been longlived, for besides these York examples there is a fine bowl of the same general class in the Guildhall Museum, London, from a Flavian pit near the Walbrook. The last 1st-century piece is a fragment in the Museum (unnumbered) of a mould-blown cylindrical cup of bluish-green glass belonging to a well-known type of the Claudio-Flavian period depicting a chariot-race. The fragment ((a) on Fig. 88) bears, in the main, part of a four-horse chariot and a sevenfold ovarium on columns; the inscription (above) is missing (YMH, 105). It belongs to a group of nine specimens, all fragments, most if not all of which are from the same mould. Of the others, five are from Vindonissa (Switzerland) and one each from Rottweil (Württemberg), Vézelay (Yonne) and Southwark.

Of fragments of 2nd-century date a rim (H.G. 218) of a shallow bowl of colourless glass (Fig. 88) of Alexandrian fabric of a type found frequently at the key site of Karanis in the Fayum indicates long distance trade in glass. Other similar fragments have turned up on villa sites in Britain, e.g. Ditchley, Oxon., and Great Weldon, Northants., and there is a complete bowl of similar ware with cut decoration from a Roman grave at Girton near Cambridge. Another import from a distance, though perhaps from Italy rather than Egypt, is a fragment of a colourless beaker (H.G. 205.3) with fine, deep facet-cutting of a type that seems to begin in the later 1st century and continues well into the next (Fig. 88). There are complete examples of the type from Locarno (Switzerland) and Curium (Cyprus) and fragments have turned up on other British sites, e.g. Wilderspool (Warrington) and Caerleon. This piece was found 'in or outside the city walls' during the Railway excavations (Y.M., Cook MS., 142, no. 17).

Much of the other early glass at York consists of fragments of cinerary vessels and of the green unguent flasks that so often accompany them in cremation deposits. The only complete cinerary vessels are the bottle (H.G. 53) which contained the ashes of Corellia Optata, aged 13, found on The Mount in 1861 (see Burials, IV Region, (l), Burial vii. Plate 66), and an ovoid jar with cover (Y.M. 1948.3.1) from Clifton, outside Bootham Bar, 1871, containing cremated bones (ibid., III Region, (f). Fig. 89). Among unguent bottles we may note one (H.G. 48) found on The Mount in 1814 'with fragments of a cinerary urn in which it appeared to be enclosed' and several from the Railway Station Cemetery just outside the walls, e.g., H.G. 14 found in 1872 in a stone coffin, H.G. 32 in 1873 (see Burials, IV Region, (f), iii. Fig. 89), H.G. 231 in 1881 (ibid., (c), v. Fig. 89), H.G. 49 in 1874 (Fig. 89), and H.G. 36.1 and 2 in 1872 (Plate 66). These cinerary deposits belong to the later 1st and 2nd centuries, any closer dating than that being very difficult to assess. The types lasted for a hundred years or more with little change.

Sometimes these unguent bottles had inscriptions on the base. One such (Fig. 89) from the Railway Station Cemetery, from near the arches through the City Wall, is inscribed PATRIMONI (H.G. 217; Y.M., Cook MS., 141, figs. 7–8; YMH, 103, item l). Many examples with this same inscription are known, four from Rome, two from Gaul and one from Cologne; in Britain examples come from Lincoln, Chester, Densworth (Sussex) and Bath. Another inscribed piece from York without exact provenience (H.G. 16) bears a winged victory and a V (Fig. 89). These inscriptions and devices on unguent bottles probably refer to the manufacturer of their contents and do not indicate the maker of the glass itself.

For the rest, the 1st and 2nd-century glass consists mainly of fragments of prismatic or cylindrical bottles. One nearly complete four-sided jug with a moulded inscription S/L on its base was found in a grave in the Trentholme Drive cemetery in 1951–2 (Burials, IV Region, (o)), and a small hexagonal jug (H. 34c) was an accessory vessel in a burial found in the Railway Cemetery in 1874 (ibid., IV Region, (i), 2nd Group. Plates 30, 67. Fig. 88). Though the jug itself is most unlikely to have been made later than the 2nd century, the cinerary urn and colour-coated motto beaker found with it are unlikely to be earlier than A.D. 250, so that the glass must have been quite old when buried (YMH, 148, item n). While all this is thoroughly typical, so far as it goes, of glass of the period found on other sites in Britain, it is less plentiful than we might expect, perhaps because many fragments found in 19th-century diggings were not kept. One would, however, have expected the cemeteries to yield more cinerary urns, especially ovoid jars, and one would also have expected some good early jugs and rectangular and cylindrical bottles.

3rd and 4th century

It is convenient to divide this material into two groups: first, glass which almost certainly belongs to the first half of the 3rd century, if not partly to the last quarter of the 2nd; and second, material more probably belonging to the later 3rd and 4th centuries.

The first group is represented only by fragmentary specimens, but they are important since they indicate, as we should expect, that examples of some of the best Rhineland glass of the period—the heyday of the Cologne glass-houses—was reaching York in some quantity.

We look first at some pieces with faceted and linear cutting: the 3rd-century Rhenish descendants of the fine Mediterranean facet-work of the 1st and 2nd centuries. Three examples (Fig. 88) come from the Railway Station Cemetery, H.G. 162, 205.1, 211 (Y.M., Cook MS., 142, figs. 15, 18, 19); all are typical colourless Rhineland ware such as is found in quantity on numerous western sites. The best piece, however, which is perhaps the finest example of Rhenish crystal yet known from Britain, is a vertical-sided bowl (H.G. 210) with flat base, the sides and base being completely covered with geometrical patterns of grooves and facets (Fig. 88). Unfortunately we do not know its exact provenience in York; but its date and fabric are clear, though it is more cylindrical than is normal for this Rhenish crystal, and it deserves a prominent place in any discussion of glass from Britain.

Another shape of Rhenish colourless glass which is particularly common on early 3rd-century sites is a cylindrical bowl with a raised ring or rings on the bottom and often a raised trail on the side as well. Bases of this type are readily recognisable and many exist from York, all unfortunately without provenience, and rims (not so readily recognisable) also occur. The complete shape (Fig. 88, combining base H.G. 202.6 and an unnumbered rim, b) is best exemplified in this country by one from Airlie in Angus, now in the National Museum of Antiquities, Edinburgh. The colourless jugs of like date and fabric from Cologne are represented by an interesting handle (H.G. 197) made of two vertical trails nipped together at intervals 'from the [old] Railway Station' (Y.M., Cook MS., 142, no. 21). Of the famous Cologne snake-thread polychrome glasses York has yielded only one fragment of a beaker (H.G. 153) with opaque blue trail (Fig. 88). But this is paralleled by two other fragments of fine Cologne glass with coloured trails, namely a two-handled jug-neck with a blue trail, from Toft Green (H.G. 184; YMH, 103, item e. Fig. 88) and a tiny bird (H.G. 208; YMH, 103, item e. Fig. 88) in opaque white glass with opaque light blue beak, eyes and wings, which perhaps comes from one of the elaborate Cologne open-work vases with birds sitting in the holes (Fig. 88, lowest right). (fn. 1) All these indicate that the very best Cologne glasses could find their way hither, perhaps to the Severan court itself.

The group of glasses from York belonging to the later 3rd and 4th centuries is larger and much more varied in its composition than any earlier group. It comprises far more than half of the total of glass preserved from York sites and is as representative a selection of late Roman glasses as can be seen anywhere in Britain. A few important and common western types of the period are missing, e.g. mould-blown barrelshaped jugs stamped by Frontinus and others, mould-blown flasks with bodies in the form of human heads, and bowls with late figured cutting, but most of the well-recognised types are represented by fragments, at least, if not by complete examples. One reason for the quantity found is that by the middle of the 3rd century inhumation had superseded cremation throughout the western empire and this gave the funerary goods a greater chance of survival, especially when coffins were used.

Fig. 88. Half actual size.

Fig. 89. Half actual size.

Though the older form of unguent bottle with long neck and triangular body survived in part, it was usually replaced by new shapes. Two of pipette shape are representative of a group of half-a-dozen or more from York graves; the larger one (H.G. 7; YMH, 103, item c) was found with a black pottery vessel in a lead coffin in the Railway Cemetery in 1875 (see Burials, IV Region, (d), Burial x) and the smaller (H.G. 6) comes from The Mount (Fig. 89). This type is frequent both in the west and in Syria during the 4th century A.D. Another late form with four depressions in the body is also found frequently at York; one (H.G. 9. Fig. 89) is from the Railway Cemetery, 1872 (YMH, 104, item o), and another (H. 324.5. Fig. 88), from which the rim is missing, was found in 1874 in a woman's grave with a bottle (H. 324.6) with ring-shaped body (Fig. 88), a rare type anywhere, and the remains of a bronzemounted casket containing jet pins (now missing) and bangles (see Burials, IV Region, (e), Burial vi). The rim of the ringbottle and most of its two little handles, which were of the delphiniform type, are missing. Of two stone coffins found on The Mount in 1872 (ibid., IV, (n), v), one produced the beaker (H.G. 144; YMH, 103, item k) which, with its wheel-incisions and unworked rim, is a typical late 4th-century piece (Plate 66), as found frequently, for instance, in the cemetery at Mayen in the Eifel, while the other yielded a colourless beaker with indented sides (H.G. 180) which, though of poorish material, is perhaps 3rd rather than 4th-century (Plate 66), and an unguent bottle (H.G. 154) of the long-lived variety with triangular body (Plate 66) (YMH, 102, item f; Y.M., Cook MS., 140, 2, 3). Another colourless beaker (H.G. 127), which has lime accretion on one side, indicating that it lay in a coffin or cist, most probably may be the one from the Railway Cemetery, 1873–4, which Raine (YMH, 25, 147, f) called a perfect drinking vessel of glass (see Burials, IV Region, (c), vi. Plate 66). A tiny jar (H.G. 2) found in a stone coffin in the same cemetery in 1873 belongs to another long-lived Roman type (Fig. 89).

Fig. 90. Glass Bottles (Burial, IV Region, (b) ii). Half actual size.

When the N. Midland Railway was building in 1840 a very interesting burial was found 'outside Barwalls' in a lead sarcophagus which, besides a skeleton, contained a lead ossuary with an infant's cremation (see Burials, IV Region, (b), ii). Four glass bottles (Plate 67. Figs. 63, 90), two with hole-mouths, two necked, lay in the coffin in pairs, lip to lip as reported and sketched by C. Roach Smith (Collectanea Antiqua, VII (1880), fig. 175) though it is not apparent whether this position had any special significance. All are of the same colourless fabric with very similar decoration of horizontal wheel-incisions in groups; the necked pieces (H.G. 146. 3–4) are a pair in size, the others (H.G. 146.1–2) are not (YMH, 102, item b, 146, item b). Among a number of other examples of the necked variety from York graves we may mention especially H. 13 (Fig. 89), which was found in a woman's grave in the Railway Station Cemetery, 1874, with a coin of Crispus (A.D. 317–26) and various items of bone and bronze (YMH, 133, item b), thus indicating a sure 4th-century date for the type (see Burials, IV Region, (i), 1st group). The holemouthed variety is much rarer everywhere and no others are known from York. Another burial in a lead coffin found in Walmgate in 1892 (ibid., I Region, (c), i) produced two similar cylindrical bottles (H. 321.7–8) but smaller and of much poorer greenish-colourless glass, full of bubbles, streaks and black specks (Plate 67), which seem very like imitations of the better variety from the other burial. H.G. 109 (Plate 67), of green glass, from the Railway Station Cemetery, with its very small opening at the base of the neck, is a dropper of a type which occurs with snake-thread decoration both in the east and west, often with indented sides. (fn. 2)

A stone coffin in Clementhorpe outside the S.E. angle of the mediaeval walls, and so probably of the colonia, produced a glass jug (H.G. 44) and a late colour-coated pottery beaker with rouletting (see Burials, V Region, (b), Burial ii. Plate 67). The glass is of poor green metal full of bubbles and impurities and a typical late piece. A better-made jug (H. 12), also of the 4th century, is the well-known dark blue one (Plate 67. Fig. 58) from the famous burial in a stone coffin in Sycamore Terrace (ibid., III Region, (b), v), found in 1901, with bone and jet bracelets, a mirror-glass (H. 11) and an openwork bone fillet inscribed SOROR AVE VIVAS IN DEO. The glass jug, with its ring-fold at the top of the handle, is a very frequent Rhineland type of the late 3rd or 4th century, no doubt from the Cologne glass-houses. The mirror-glass (Fig. 58), which might well be passed by, by the unwary, as a mere fragment of a vessel, belongs to a well-known late Roman type of small mirror, originally mounted in a plaster or wooden frame, and, no doubt, carried in the owner's satchel.

A number of other varieties of flask or bottle also occur, mostly, unfortunately, without recorded provenience. A very common late type in the west is the cylindrical bottle with two small delphiniform handles at the base of the neck. One such is H.G. 182 (Fig. 89) which, as it has lost its neck and rim, may be the 'small bottle, slightly imperfect, with miniature handles' which Raine (YMH, 102, d) cites from the Railway Station Cemetery, 1872 (see Burials, IV Region, (f), iv). Fragments of the other delphiniform-handled type of flask, with spherical body, also occurred at York, e.g. H.G. 227 (Fig. 88). A tiny unguent flask (H. 103.1), mainly green, but with a colourless streak, of poor, late metal (Fig. 89), came from a female burial in a stone coffin found on the site of the Station Hotel, 1874 (ibid., IV Region, (c), iv). Of greater interest are two flasks without provenience, both typical of the 4th century in the west. The first (H.G. 33) is a colourless spherical vessel with cylindrical neck, constricted at the base, which has bands of horizontal wheel-incisions on neck and body (Plate 67). The second (H. 2059.2) has a squat globular body and a funnel neck and it, too, has bands of horizontal wheel-incisions (Plate 67). Both are types which are very frequent in Gaul and the Rhineland.

Most of this 3rd to 4th-century glass was undoubtedly made in the Rhineland or north France, though the shapes are in many instances derived from eastern prototypes. The reason for the predominance of eastern shapes and designs in western glass-houses is that they were largely run by eastern workers who were recruited, generation after generation, so it seems, from the Syrian and Alexandrian industrial centres.

There is no evidence that any of this glass was made at York, or even in Britain, though we know that glass-working existed at Colchester, Caistor-by-Norwich and Wilderspool (Warrington) and it may be that some of it, such as the two poorly made Walmgate bottles (see above, H. 321.7–8), are British products. They are so poor that one would not expect them to be accepted by a self-respecting Rhenish druggist as containers for his toilet preparations.

* * *


  • 1. This was Mr. W. A. Thorpe's view in 1935 (English Glass, 23, pl. viii, b). It is possible, however, that the bird formed the finial of an unguent stick of similar, but not the same, type as those figured by F. Fremersdorf, Das naturfarbene sogenannte blaugriine Glas in Koln (1958), 37, pl. 134.
  • 2. See F. Fremersdorf, 'Der römische Guttrolf' in Archäol. Anzeiger (1931), col. 131 ff., for an account of this and later varieties of droppers. It should be mentioned that we have purposely omitted from this account the complete dropper flask in the Yorkshire Museum illustrated and described as Roman by W. A. Thorpe (English Glass, 41, pl. V, c), for we believe it to be Waldglas of very much later, perhaps 17th-century, date.