D Home Publikationen Glastöpferei Techniken Kameoglas Diatretglas Hedwigsbecher Craqueléglas Werdegang  D

E EnglishHome Publications G lassPottery T echniques CameoGlass CageCups HedwigBeakers Craquele Vita E

Rosemarie Lierke

Antike Glastechnologie / Ancient Glass Technology


Scratched, ground and cut -
the evolution of early glass cutting

Translation of: Geritzt, geschliffen und geschnitten. Die Evolution der frühen Glasschneidekunst. In: Antike Welt 34 (2003) H.4, 345-356

Any corrections concerning the English are appreciated. For citations, you may mention this summary and website, but please, for literal quotes refer to the German original.


Cut glass is always something special because of the risk involved. How began this art? Was its development a steady progress or was it rather created by personalities – not known by name today - which set standards even for today?  According to the written history of glass, this seems to be the case. However, if the sequence of known examples is critically investigated, the picture changes.     

It is a timely topic. Independent investigations recently confirmed that the assumed grinding marks of ancient glass in reality were not caused by grinding.[1] The same conclusion was already presented several years ago in a series of articles in the Antike Welt [L1993b , 1995a, 1995c, 1996a]. With additional contributions of several well-known scholars of different scientific background (Egyptology, archaeology, glass technology, philology and physics) these  articles were condensed in the book L1999. The glass technologist M. R. Lindig discovered that in reality the assumed grinding traces are ‘hot’ scratches. They are caused during the hot working process by molds or tools with tiny flaws or protrusions. Even today – but very seldom - such scratches occur. 

Recently, a cameo-glass fragment (Abb. 1a .Martin von Wagner Museum, Würzburg, Inv. Nr. H 1725., 1b inside with scratches) was identified to belong to the lower part of a cameo-glass balsamarium (compare with Abb. 1c Museo Archeologico, Florenz, Inv. No. 70811,  recent photo by G. Capecchi). This should convince everybody that the typical scratches on the inside of this fragment are no grinding marks.[2] How and why would a cameo vessel with a narrow neck be ground inside? The scratches here as otherwise are full in agreement with the new explanation of the cameo manufacturing process, which was a cunning method where the decor was hot formed, and which is related to the manufacturing of relief ceramics. Scratches as the ones pictured here - on the inside or on the outside - are typical manufacturing traces of this process.[3]. Cameo glasses remain extraordinary masterpieces in the history of glass, however the history of glass cutting has to be rewritten without them.

The error about the true nature of the typical scratches considerably confused the perception of ancient glass working, especially of cutting and grinding. This is, however understandable. The behavior of the glass material and the problems concerning hot and cold working of glass are rather complicated. Melting the ancient raw glass was a process of long duration at low temperatures. This process and its influence upon the working properties of the reheated glass material are not yet fully understood. Today, glass is molten fast, homogeneous and free of bubbles in a controlled atmosphere at high temperatures. It is cooled down in controlled temperature steps. Replicates and experiments with modern glass material therefore are of limited help to clarify ancient techniques. The manufacturing traces of ancient glasses are usually obscured by corrosion, or they are worn off. Sometimes they are ambiguous. Traces of grinding naturally sometimes resemble the typical scratches. Nevertheless, if the scratches cannot be explained as grinding marks any more, some glasses exist which neither were cut nor ground, despite it was formerly assumed. To these glasses not only belong the early cameo glasses – as mentioned already – but, some other ancient glasses have to be removed from the list of cut glass as well.

A brief survey of representative examples will show that the abrasive working of glass – that means the cold treatment by grinding, cutting or engraving – was developed in a logical sequence from simple scratched drawings to complex compositions of cuts (Abb. 2). As one should expect, this development corresponds with the development of tools and hot glass-working.


The cold working of glass and its invisible precondition

Glass is no stone! Just thinking of treating a block of stone or a block of glass with hammer and chisel reveals the difference. Hitting the stone will leave nearly controllable traces, while hitting the glass block will cause chaotic cracks and splinters. The crystalline structure of the stone and the amorphous structure of the glass are the reason for this difference. Glass cannot be ‘shaved’ like wood or metal either. There was neither acid treatment nor sand blasting, neither flexible shaft nor other modern possibilities of cold treatments available in antiquity.

It was always only possible to take cautiously away fine layers of glass by scratching rubbing or grinding. Easiest is the scratching of thin lines or the stippling of points with a stylus made from flintstone. The Romans, according to Pliny (Naturalis Historia 37,61) already used splinters of diamonds too. The surface could be abraded by rubbing with a grinding stone or with grinding powder. Drilling or filing are the same as grinding or rubbing with a grinding stone or a tool of special shape and the addition of grinding powder. From coarse to fine grained minerals with greater hardness than glass are needed for the abrasive effect. This could be for example quartz sand or the quartz crystals in sandstone, or the famous emery from Naxos, which contains corundum crystals of even greater hardness. Easier working and more creative possibilities were offered by the use of different sizes of grinding wheels. With each turn, these wheels remove with edge or side a fine layer of glass. Metal wheels or disks can be used with a slurry of grinding powder. For polishing, successively finer grains of grinding powder and finally dust-like polishing powders are used. To prevent heating up the glass, it has to be cooled by water or oil, otherwise, strain or cracks could occur. This danger dictates also the most important precondition: The glass which is to be cut or ground needs to be perfectly free of strain. This is of tremendous importance, but widely unknown. Therefore, it shall be illustrated by an example.

A practical experience    

In the 70ties, I was permitted for mosaic experiments to search through the glass trash heap of the Hessen Glassworks, which today do not exist anymore. They produced large multi-colored rods for the glass-button makers. Since multi-colored glass cannot be used after re-melting, everything which went wrong or broke during any stage of the production process became trash. My salvaged ‘treasures’ therefore were – independent of their appearance - of very different qualitiy. Some of the fragments had taken the full cooling cycle, others had cooled down slowly or fast without ever entering the annealing (cooling) furnace. This difference was revealed by the grinding action of the diamond saw (Abb. 3: left strain-less, right strongly strained glass). Either a fragment was easily cut into fine slices, or the slices broke. In an extreme case, the whole rod cracked into very sharp shards by the first touch of the saw because of its internal strain.

To still use the ‘strainful’ fragments, they had to go through cautious reheating and a new cooling process. For glasses which are to be cut or ground, this process cannot be dispensed with. The glass needs to be reheated to a temperature just below its softening point and it needs to stay at this temperature until all strain is resolved, which depends upon shape and wall-thickness. After that, it needs to be cooled quickly through a temperature region where the hot glass is apt to crystallize. The further cooling has to be just so slow that any new strain is prevented. An absolutely certain success of this procedure without modern temperature controls is even for an experienced glassmaker of our time a matter of lucky chance. In ancient or medieval times, this must have been always the case. This explains why the grinding and cutting of glass never was as often applied as the same treatment of stone. 

Everything began ‘from scratch’

The first glass most certainly was created as glossy slag during the copper smelting. There cannot be any doubt that there were early attempts to grind this artificial ‚stone’ just as one used to grind other attractive stones. But, this was successful only if the the raw glass by lucky chance was correctly cooled. People learned fast to reheat the glass and to work the molten glass, which is the biggest advantage of this new material compared with natural stone. The first engravings on hot formed beads and vessels appeared in the middle of the second millenium BC as superficially scratched names of kings. Later inlays and amulets were embellished with scratched drawings which still look not very ‘professional’ (Abb. 4 Detail of a coffin from ‚KV 55’, Munic, Staatliches Museum Ägyptischer Kunst).

For centuries, the abrasive glass working was limited to such scratched names and drawings, and to the treatment of the rims of pressed amuletts and inlays. In the literature are also mentioned glass vessels which were ground from a full block of glass [Supplement: this applies especially to clear or transparent greenish glas]. This would have been possible if the cooling of the raw glass block after melting and the cooling during the drilling and grinding was perfect. But it remains rather unlikely. To explain the manufacturing of such vessels, more realistic possibilities should be taken into account. For instance, the winding of a thin or a thick trail around a rod or a core. Since hot glass has a high surface tension, the traces of winding disappear in the fire. Therefore, winding cannot be precluded even if traces of winding are not present any more. Alternatively, pressing could have been applied. Its use is confirmed since the earliest time of glass working. Pressing, as a rule, in antiquity was not a method of mass production as in modern times.  

The first grinding wheels were of astonishing size

Fragments of bowls, excavated in Nimrud and dated to the 8th or 7th c. BC are assumed to be the oldest cut glass.[4] In Abb. 5 (British Museum, WAA 134895, 1966-12-17,3.) is shown the small fragment of a deep bowl with a partly very thin wall and a protruding band of 1,5 cm width, which carries a fine pattern of cross-hatched grooves – the photograph is deceiving with the impression that the grooves could be crossed ridges instead. The bowl seems to be pressed with its band in a mold, but without the pattern. Theoretically, the pattern could have been molded to, but this seems to be unlikely, since it would have required a complicated mold preparation. The question is, were the strikingly straight lines of the pattern carved with a cutting wheel into the protruding band? During the 7th c. BC, for the working of Mesopotamian rock crystal cylinder seals, straight lines were still made by filing, which has a much older tradition.[5] This working method could have been employed here too. Otherwise, a very thin wheel with a large diameter and powerful drive would have been required. In each case is this early dated cross-hatching something special. It appears again at finds from the 4th c. BC, in one case in the same area as the next example.[ 6] As a rule, early cut glass is more rudely worked.  

Beakers and bowls from the 5th and 4th c. BC show cut grooves resembling for instance lotus flowers or a related finely rounded pressed design. In both cases, rotary cut grooves used to supplement the design, which were carved with a stylus, stone or wheel into the outside of the vessel. Also a combination of pressed design and cut grooves seems to appear – but is rather difficult to discern this at the usually badly corroded examples. A very early shallow bowl from Rhodes, dated to the end of the 5th or the beginning of the 4th c. BC (Abb. 6, Rhodes, Archaeological Museum Nr. g 854), unequivocally witnesses with clumsy grooves the use of a cutting wheel with a diameter of about 15 to 20 cm. It is not possible to preclude for the blank of this shallow bowl that it was pressed into a mold which was positioned on a horizontally turning disk. This induces a theory. If such a disk was made from a suitable material, sandstone for instance, the edge of this disk may have been used for the grinding of the décor.  

Certain is in any case that the first cutting wheels for the decorating of glass were of astonishing large size.[6a]   For that reason it makes sense to assume that they rotated horizontally with a drive related to that of a potter’s wheel. Horizontal cut grooves become more frequent in Hellenistic and early Roman times. They do not appear single or in groups on the outside only, but also on the inside. They are usually cut more professionally on the outside. But, this is nothing to be surprised about. The inside could not be worked with the customary large wheels. As a rule, the inside was only accessible for a stylus or stone which was held steady against the rotating bowl [Supplement: which was fixed for instance in a clay bed on the horizontal wheel ]. The outside always is easier to be worked on, one should not forget the necessary cooling with water and the removal of the cutting remains. The enduring cooling problem show frequent strain cracks, which for good reasons appear especially at inside grooves. 

A certain Hellenistic type of glass (Abb. 7 Düsseldorf, museum kunst palast Inv. Nr. P 1966-63) is distinguished by comparatively wide, tightly packed vertical grooves. In view of the extend of cutting, it represents something like the climax of the simple decor with cut grooves. The shape of the rim with a sometimes bulging edge will be similarly repeated much later at the early facet cut beakers. The rotary scratches which were discovered on the inside of this bowl are proof of its manufacturing by rotary pressing similar to ceramics.   

About grooves and ridges

No matter whether a trace in the glass surface is left by a stylus, by filing, or by the action of a cutting wheel, it will always be a more or less pronounced concave, that means deepened groove (Abb. 8 left). Every protruding ornament (Abb. 8 right) needs a laborious removal of the whole surrounding area, and each convex curvature has to be approached step by step by numerous tiny cutting facets. Often a whole array of different cutting wheel profiles would be required – which are not yet accounted for in antiquity. Above all, the grinding and polishing of a perfectly smooth new background of a protruding relief is one of the most laborious tasks of cutting in general. According to the experts J. Matcham and P. Dreiser it can hardly be perfectly achieved. This permits another conclusion: if the background of an ancient glass with high relief does not show any cutting traces, then it was not cut. Are traces to be detected which look like grinding or polishing marks, then these traces need an investigation to clear their cause. These traces may originate from a recent cleaning. A pressed glass may show traces which reproduce the original traces from a model, or they may have been caused by touching the rotating hot glass with a tool, a template or the press mold. 

This reasoning must suffice here to expel from the history of glass cutting not only the early cameo vessels, but as well the early glasses which were assumedly cut in high relief –  they are certainly high relief glasses, but not cut (Abb. 9 A). It is safe to preclude that it was possible or even customary to cut thickwalled blanks to get glasses with simple thick protrusions on a background without cutting traces– and this even before the first cutting of simple intaglio facets. The simple protruding ornaments were either pressed or applied. It is absolutely correct of P. Hameling to decribe the decor of the beaker from Begram (Abb. 9 A rechts) as ‚relief à chaud’.[7 ]  A single protruding rounded ring around a vessel, which was not made by an applied trail, is proof enough to discern a glass vessel to be pressed or blown into a mold – the last difference easily to distinguish on the inside of the vessel by its smooth surface or a ‚negative’ pattern. If a relief was indeed created by cutting, it would look like a wood cut – as for instance the famous late antique cameo dish from Stein am Rhein, where the clearly visible traces of cutting on the background leave no doubt at all about the manufacturing method.[8

Another misconception – the ancient ‘glass turning’

During late Hellenistic and early Roman imperial times, the rotary pressing of glass became the prevailing method of glass working. Even hollow vessels could be made by sagging a glowing hot bowl over a mushroom-shaped removable core and a final narrowing of the neck. Hollow vessels made this way, unsually feature the typical scratches. The erroneous interpretation of these scratches as grinding marks led to the unrealistic assumption, that in antiquity glass vessels could be made by lathe turning. This would mean to fix a pre-shaped blank or a lump of raw glass into an appliance with a horizontal axis, and to shape the rotating glass by grinding it with a hand held tool – a process called ‘drechseln’ in German. This assumedly was the method to create for instance vessels with protruding rings which were mentioned before. However, grinding or shaping a whole vessel this way is only possible in theory. Almost all ancient glasses which theoretically could be explained this way show slight diversions from the rotation symmetry or other irregularities which can be explained by hot forming, not by rotary grinding. By using a normal turning lathe the trace of a central cavity for clamping the glass into the machine and the trace of a second fixing point at the periphery should be noticeable. Neither the first nor the second trace have ever been detected. There are glass lathes on the market today, however these are machines for laboratory glass blowers to handle big pieces in the flame. They have nothing to do with grinding.

The first facet cutting 

In the 2nd half of the first c. AD, the glass blowers were very successful with their mold-blown vessels. Obviously in competition with this success, the glass potters created – as already mentioned – pressed vessels with simple thick protrusions, for example Abb. 9 A left and middle. As own experiments have shown, their making by pressing is basically simple, but each vessel needs a lost model for a one-piece mold, which was made from plaster or a mixture of plaster and quartz sand. Such a mold permits only one pressing. Perhaps this is the reason why the pressed high relief vessels were soon abandoned in favor of simple intaglio cut facets. The attractive facet beakers were created with rows of overlapping grinding olives in quin-cunx pattern. It became a story of success and the foundation of the further development. (Abb. 9 B, a). One or two grooves in the press-mold created a simple or double bulge at the rim. This way a stable rim was created which could – if necessary – easily become ground smooth. This spezial rim, perhaps rotary rings, or the typical circular protrusion in the bottom, prove the relationship of the facet beakers with the high relief glasses. Also the high slim beaker shape is repeated, which seems to be the earliest facet beaker shape.

The beaker in Mainz (Abb. 10 a. Römisch-Germanisches Zentralmuseum Inv. Nr. O.12633) illustrates with two flat bubbles that the typical wide band under the bulging rim of this vessel type was not ground (Abb. 10 b. the rim of the beaker with rotary scratches and two flat bubbles). Its rotary traces were created – just as with other early facet beakers – during the hot working of the blank in an assumedly two- or three-partite mold. The use of multi-part molds was learned by the glass potters from the glass blowers of their time.

Concerning the cutting technique, there is a clear relationship of the facet beakers to the vessels represented by Abb. 7. The vertical grooves of the latter type were cut with a large, probably horizontally turning cutting wheel. The cutting wheel which was used to create the facets of the tall beaker from Mainz (Abb.10 a, b) also was of astonishing size, its diameter was 12 cm. There are other beakers with small and deep facets, which required a substantially smaller wheel. Measuring a fragment of a smaller beaker with facets of about medium size rendered a cutting wheel diameter of 3,5cm.[9] Cutting wheels of this size and smaller hardly were operated like a potter’s wheel. That means at the latest for the making of  the facet beakers, the existence of cutting wheels which rotated vertically around a horizontal axis has to be assumed – just as it is customary today. (Abb. 11 theoretical simple cutting lathe). The wheels probably were made of sandstone or of another suitable stone material. Depending on its grain size, sandstone can achieve a silky gloss without extra polishing. The cut facet beakers were rarely polished. 

From facets to figures

A personal experience was helpful to understand the logic of the following development. The simplest cutting ornament is the circular or olive shaped concave lens, also called facet, which is created by the action of a cutting wheel with a rounded profile. The overlapping of such olives results in the honeycomb pattern of the early facet beakers. What else would be possible with just one cutting wheel? In the year 1984, I was permitted in the Eisch factory in Frauenau to try an engraving lathe - I had never any training before. They prepared for me a comparatively coarse and wide bonded cutting wheel – the equivalent of a small natural grinding wheel – and handed me a beautiful beaker. Holding this beaker mainly in the right hand, the palm of my left hand became my model. The result of adding facets and moving the wheel across the glass surface is shown on Abb. 12. From this experiment it became understandable, how from simple facets arranged in the usual order, by moving the wheel the lively pattern of the beaker from Pompeji was created (Abb. 9 B, a. right) – and it becomes clear what may be necessary in addition for the engraving of a figure: at least a second or third smaller wheel and/or a stylus to add details.

A few years ago, in Albenga, near the Ligurian coast line, in a grave of the early 2nd c. AD, a really great blue translucent plate was discovered. This plate is the oldest known glass with a three-dimensionally engraved figural design (Abb. 13 Soprintendenza Archeologica della Liguria, Genova.). Here indeed the artist was limited to the use of 3 cutting wheels of different size and a stylus.[10] This fits convincingly to the idea that this plate belongs at the outset of figural engraving. In the workshop were this luxurious object was created, probably three simple lathes existed, with a different wheel each. It seems unlikely that in this early stage interchangeable wheels were used which had to be adjusted every time. Unfortunately, the great blue plate from Albenga shares the fate of many other early cut glasses: concentrated in the cut design, many cracks are visible which are in addition affected by corrosion. These cracks most certainly indicate an insufficient cooling of the blank or an insufficient cooling during the cold abrasion.   

Perhaps even older than the blue plate is a two-handled bowl from Siphnos, which is dated to the end of the first c. AD.[ 11] Its décor is made by a simpler technique. Scratched outlines enclose mat areas – a two-dimensional graphic representation. It therefore would make sense to consider an earlier manufacture, but in view of the distant find-spots, this difference could be incidental. The blanks of both vessels with early figural cutting were pressed. This seems to be indicated by the internal streaks of the plate, for the bowl of Siphnos an autopsy revealed the significant scratches on its inside.   

A first change and the further development

Not a new cutting tool, but blown cutting blanks brought a change in the development of glass cutting. In the first half of the 2nd c. AD, the glass pottery came to a preliminary end. This had consequences for the cut glass. The thin-walled facet beakers of the advanced 2nd century show the difference (Abb. 9 B, b). The pressed stabilizing bulbuous rim was not possible any more. Protruding rings were replaced by narrow cut grooves. Often, the extremely shallow facets do not overlap any more. The bottom looses its central circular protrusion. Beakers of this kind are rare. Very popular and widespread became blown hemispherical cups with often quite roughly broken off rim, decorated with rows of different notches, grooves and facets in all kinds of different combinations (Abb. 9 B, c). They are dominating in the 3rd century. 

In the 4th century, the pattern become richer (Abb. 9 B, d) and the selection of shapes for cut decoration more divers. There are shallow bowls with ‚carpet’ decor and rich pattern on beakers, bottles and other big vessels. Tools, melting and cutting technique have reached a high standard. But, this standard still cannot be compared with the standard of our time. According to an expert: „works of glass cutting and engraving of the 4th century are primitively executed and show no great expertise’ [12 ].“ There are indeed very often – not counting some awkward figural designs – stuttered or slipped lines, displaced grooves or facets. One has to consider the ancient working conditions. The precision of the primitive ancient cutting device could certainly not be compared with a modern lathe. The drive was the task of a helper since the glass cutter as a rule needs both hands to hold his piece. This explains already many shortcomings. An easy interchangeability of wheels obviously was not possible yet, and the selection of cutting wheels was extremely limited. Almost exclusively rounded profiles can be found. Even the use of a simple mitre wheel is rare. What still has been achieved under these circumstances may still be admired.

In a large area of central and eastern Europe, the bell-shaped or later the conical facet beakers are the leading finds of cut glass in the 4th and 5th century (Abb. 9 B, e, f). They often feature especially large facets – witnessing a stronger and more effective grinding equipment. It becomes more difficult for the late facet beakers to differentiate between blown or pressed blanks. Pressed glasses re-appeared around the middle of the 3rd century and show – less often than the early pressed glasses – the typical scratches. These were the ‚production marks’ of early pressing, which also could be called ‚production defects’ if the standard is raised. In any case, the typical scratches on the inside of an especially thick-walled facet beaker leave no doubt about its manufacturing (Abb. 9, e right).[ 13] The time- material- and energy-saving method of rotary pressing is best suited to produce the typical bell or upside-down cone shape of the late facet beakers. Thin-walled beakers with cut-off rim certainly were blown. But some new vessels with high relief and the significant scratches proof the re-appearance of the phenomenon that re-heated raw glass obviously could be pressed rather thin-walled too. This must either be caused by the mode of re-heating – for instance directly in the press-mold, or by the working properties of the raw glass – for instance by a high content of gas.  

The triumph over fragility – the cage cups

Mentioning the bell shape reminds of the outstanding cage cups or diatreta (Abb. 14 a. the cage cup from Munic, Staatliche Antikensammlung Inv. Nr. 12129). Surprisingly, despite their uncontested mastership, these fragile pieces are no exception from the generally low standard of glass cutting and grinding of their time. In Cologne, Munic or Trier, an attentive museum visitor would be able to detect for instance the crudely cut struts between cage and beaker which show the still existing problems. The delicate appearance of the cage cups is caused by the use of a double-shell blank, of which the outer shell was opened with comparatively few cuts in a strict rhythm. The edgy struts originally were thick and round bridges between the two shells of the blank. The thin-walled internal beaker often features flat pressed bubbles – just one of several points which proof that these beakers were not cut from a thick-walled blank (Abb. 14 b. Detail of 14a. Showing large flat bubble in the internal beaker).[14 ]

The cradle of these masterpieces has not been found yet. It obviously was not necessarily located were the art of glassblowing was at home. Perhaps, the cage cups were made somewhere in an easterly or south-easterly area at the outskirts of the huge Roman empire. This must have been an area, where the old tradition of pressing reheated raw glass chunks was practiced without interruption, and where new masterpieces were created to withstand the competition with blown glass. The pressed facet beakers of late antiquity and other pressed glasses of this time could have been made in the same workshops.   

The masterpieces of late antique figural glass cutting 

Naturally, the figural glass cutting did not flower everywhere. The availability of suitable blanks, the selection of tools, the craftsmanship and creativity of the artists and the interest of solvent customers were the foundation for the establishing of workshop centers – limited in time or area. With a certain limitation or selection of the resources or tools in each center, a local style was developed. In a try to locate their provenience or workshops, F. Fremersdorf was 1967 the first to sort the extant artefacts according to their stylistic and technological features. There is for instance a group with engraved outlines filled with mat areas, pursuing the two-dimensional drawing style of the bowl from Siphnos. A further group of vessels features a crude style of parallel grooves and typical ‚hedgehog’ hairdos. A group called Wint-Hill-Group by D. B. Harden presents superficially with short parallel strokes engraved drawings – a style which is less demanding concerning the craftsmanship of cutting, but usually still pleasing aesthetically (Abb. 15 Römisches Museum Augsburg, Inv. Nr. 2000,4189.). The majority of these groups did not contribute anything remarkable toward a technological progress, however, they do add color to the picture of the cutting evolution, and they transport their message – for instance the early Christian theme of the Augsburg dish – in an impressive way in a luxurious material.

The decisive impact for a technological progress has to be searched in the succession of the great blue plate from Albenga. With its combination of concave cuts and scratched lines it has shown the range of possibilities. It is forerunner of a whole family of figural engraved vessels, called ‚Lynkeus-Group’ by F. Fremersdorf after a find from Cologne.

The name-giving small cup from Cologne is dated into the 2nd half of the 2nd or the beginning of the 3rd century and reminds indeed with two cut grooves under the rim of the blown facet beakers of its time [15]. There are beakers of similar shape in the same group, signalizing with protruding rings and other features that they have not been made from blown but rather from pressed blanks. This difference may not have been sufficiently regarded so far. But it may be important concerning chronology or provenience – considering the hiatus of the western European pressed glasses from the early 2nd to the middle of the 3rd century.[16

To the Lynkeus-group belongs also the majestic Dionysos bottle from Mainz (Abb. 116 a,  Landesmuseum Mainz Inv. Nr. R 6111. Abb. 16 b. impression), which was found in a grave of the 4th century. With its interesting multi-figured design, it belongs to the top achievements of late antique glass cutting. It surprises, however, that 200 years after the plate fom Albenga even for this masterpiece a handful of cutting wheels with round profile and a stylus were sufficient. More than half a millennium after the shallow bowl from Rhodes are bend lines for the master of this large bottle still a problem. But, the late antique master does not force a large wheel into a bending. He now has smaller wheels which he moves step by step into the desired direction. This can be noticed at the Dionysos bottle by attentive observation. His collegues use this method as a choice of style. At many ancient cut glasses the halos are strikingly ‘stuttered’. 

To the top of the art and craftsmanship of glass cutting of this time belongs the youngest of the cut bowls found in Germany (Abb. 17. Augsburg, Römisches Museum, Inv. 1983, 2325). Unfortunately, only part of its fragments could be salvaged from a grave in Augsburg of the beginning 5th century.[17] A graphic reconstruction give a hint of its artistic quality. An investigation of this object reveals the progress in craftsmanship. Still are almost exclusively rounded profiles of cutting traces visible, however they are more varied in size and placed with superb precision. Beside them are small cuts, which could be explained as traces of a small mitre wheel. The exact number of cutting wheel is hard to discern without a precision measuring instrument, since the traces are usually rather shallow. The same wheel by more or less deep cutting could have generated smaller or larger cutting olives. It does not seem that there have been used substantially more wheels than the master of the Dionysos bottle had at his disposal. What was achieved by these limited means gives an impressive testimony – not only of the labour, progress and success of the craftsmen, but also of the images and ideas of this time, long ago.     

A big jump to the Renaissance

The masterpieces of the late antique glass cutters with their rich pattern, as well as the top selection of figural engraved glass were made with simple cutting wheels of natural stone. The real ‚Glasschnitt’, offspring of the stone cutting of the Renaissance toward the end of the 16th century, uses small copper wheels with a slurry of grinding powder and oil. According to an invention which may go back to Leonardo da Vinci, the glass engraver now can use a crooked shaft and pedal drive, which enabled him to control the speed of the turning motion himself. A battery of at least a dozen different interchangeable cutting wheel profiles, each appearing in different sizes, is at his disposal – permitting to intaglio-engrave minute details of three-dimensional images into the glass surface.

More than 1000 years passed between the intaglio engravings of late antiquity and the first intaglio copperwheel engravings at the outset of modern times. Still 100 years more passed by, til the first laborious baroque high relief was cut, that means a protruding relief as opposed to the ‚negativ’ engraved intaglio. Only since watermills provided the necessary energy, this art began to flower. It is very enlightening to read at F. A. Dreier, how often and quite naturally the protruding relief was cut from especially prepared pressed blanks – and not, as always assumed for the old Romans, from simple thick-walled blanks, were every protrusion had to be cut from. For the cut cameo-replicates of the 19th century for the crude preparation acid was used – which was not available in antiquity.

Who looks again at the survey of Abb. 2 , will notice a simple, understandable evolution from simple to complicated cutting tasks with improved tools and glass material. In the medieval age, the famous Hedwig beakers with their deep relief seem to fall out of the level of craftsmanship which would be expected for a cut glass in their time – a conclusion which seems to be supported by other observations[18 ] – and the same is to be expected for some Islamic glasses which were not investigated for this paper. 

Naturally, the progress of the evolution can not be described simply by ‚few tools + simple technique = early work’ There were masterpieces of glass at all stages in glass history, but they do reflect the technological possibilities of their time. It is our knowledge of modern [assumedly simple] methods, which can lead to a misinterpretation of the traces of a technique which is not used anymore - as in the case of the early cameo glasses.


Selected Literature

R. J. Charleston, Wheel-Engraving and –Cutting: Some Early Equipment, I Engraving, Journal of Glass Studies 6 (1964) 83-100.

F. A. Dreier, Baroque Glass Engraving in Hesse, Journal of Glass Studies, 38 (1996) 11-228.

F. Fremersdorf, Die römischen Gläser mit Schliff, Bemalung und Goldauflagen in Köln,  Die Denkmäler des Römischen Köln Bd. 8, Köln 1967.

A. Grimm, Das Münchner Konvolut aus ‚KV 55’, in: A. Grimm, S. Schoske, Das Geheimnis des Goldenen Sarges. Echnaton und das Ende der Amarnazeit, München 2001, 64-100.

D. B. Harden et al, Glas der Caesaren, Mailand 1988.

R. Lierke mit Beiträgen von M. R. Lindig, A. Locher, H. Mommsen, B. Rütti, B. Schlick-Nolte, E. Simon, C. Steckner, E. M. Stern, C. Weiß, Antike Glastöpferei – ein vergessenes Kapitel der Glasgeschichte,. Mainz, Verlag Philipp von Zabern (1999).

M. J. Klein, Hrsg., Römische Glaskunst und Wandmalerei, Mainz 1999.

J. Matcham, P. Dreiser, The Techniques of Glass Engraving, London, 1983, S. 96.

A. Oliver, Early Roman Faceted Glass, Journal of Glass Studies 26 (1984) 35-58.

E. Straume, Gläser mit Facettenschliff aus skandinavischen Gräbern des 4. und 5. Jhs. n. Chr., Oslo 1987.



[1] The investigations were carried out by Ch. Eckmann, M. Fecht, S. Künzl in the Römisch-Germanisches Zentralmuseum, Mainz; by H. Römich and her team in the Fraunhofer Institut für Silicatforschung, Wertheim/ Bronnbach. Results are published in: L2002b

[2] C. Weiß, U. Schüssler, Kameoglasfragmente im Martin von Wagner Museum der Universität Würzburg und im Allard Pierson Museum Amsterdam. Jahrbuch des Deutschen Archäologischen Instituts 115 (2000), Berlin: Walter de Gruyter, 2001, S. 219-224.   

[3] see more about cameo glass here

[4] A. v. Saldern, Glass, in: M. E. L. Mallowan, Nimrud and ist Remains II, London (1966) 623-671. Supplement: Another glass of this find was published by G. Lehrer in Journal of Glass Studies 16 (1974) 9-13. I do not share the assumption that this was a figural cut glass. 

[5] M. Sax, N. D. Meeks, D. Collon, The introduction of the lapidary engraving wheel in Mesopotamia, Antiquity 74 (2000) 380-87.

[6] Pavlos Triantafyllidis, Rhodian Glassware I, The luxury hot-formed transparent vessels of the classical and early hellenistic periods (in Greek with engl. summary), Athens 2000, S. 128-131 No.1, 2

[6a] Supplement: The famous bow-driven lathe, published also in L1999, 21, obviously has not played the role in early glass cutting which was formerly assumed.

[7] Pierre Hamelin, Matériaux pour servir a l’étude des verreries de Bégram, Cahiers de Byrsa 4 (1954) 174.

[8] Museum zu Allerheiligen, Schaffhausen, Inv. Nr. 23096. H. Urner-Astholz, Die Gläser, in: M. Höneisen (ed.), Frühgeschichte in der Region Stein am Rhein – Archäologische Forschungen am Ausfluß des Untersees. Schaffhauser Archäologie 1, Basel (1993) 134-157, Abb. 111-117; L1999, Abb. 167, 168.

[9] Edith Welker, Die römischen Gläser von Nida-Heddernheim, Schriften des Frankfurter Museums für Vor- und Frühgeschichte III, Frankfurt 1974, T. 9/148. 

[10] Detail photos in: Bruno Massabò, Grande piatto in vetro blu, figurato ad incisione e ad intaglio, da una tomba della necropoli di ALBIGAUNUM (Albenga), Journal of Glass Studies 40 (1998) 25-53.

[11] Nationalmuseum Athen, Inv. Nr. 16275. G. Mackworth Young, The Roman Graves of the First Century AD; in: J. K. Brock, G. Mackworth Young, Excavations in Siphnos. Annals of the British School at Athens XLIV (1949) 80-92, Pl. 33, 34; G. D. Weinberg, Glass Vessels in Ancient Greece, Nr. 76,  p. 112-113.

[12] Josef Welzel, Schleiftechnik der Diatretgläser, Glastechnische Berichte 51 (1978) H. 5, 130-136. W. assumes that stone cutters made the cage cups with the same tools used in a different way. The original quote: „Die Glasschliff- und Gravurarbeiten des vierten Jahrhunderts sind ... primitiv ausgeführt und deuten auf kein großes Können“.

[13] Hermann Günter Rau, Spätkaiserzeitliche Glasbecher vom Typ Sakrau II aus der Moldau, Archäologisches Korrespondenzblastt 5 (1975) 139-141, T. 44/2.

[14] About the manufacturing of cage cups most recently L2001a. In this JGS publication was printed erroneously a flat bubble of the cage cup in Cologne instead of the same detail of the cage cup which is now in Munic, as shown in Abb. 14b here.   

[15] Römisch-Germanisches Museum Köln Glass no. 295. Harden 1988, Nr. 108; Klein 1999, S. 66, 67

[16] British Museum GR 1868.5-1.919. Harden 1988, Nr. 109. The dating to the early 3rd century seems to be questionable.  

[17] A. Rottloff, Spätantike Repräsentationskunst in Süddeutschland, Augsburger Beiträge zur Archäologie 3 (2000) 123-160 with Literature about early cut glass. 

[18] L2003a