Colorless Glasses in Antiquity

Most glass objects of the second and first millennia B.C. were strongly colored, as are the precious stones they were meant to imitate. Those that were not intentionally colored usually had a pronounced greenish tint, owing to the presence of iron impurities introduced with the raw materials. However, there are a few readily recognizable types of luxury vessels, as well as beads and small decorative inlays, that were made of truly colorless, transparent glass.

Such glasses have been found at various sites throughout the Near East, but perhaps the most familiar are the Achaemenid bowls thought to have been made in Iran or Iraq. Gold-glass objects (objects containing gold-leaf decoration encased in glass) were also often made of this kind of colorless glass.

The colorless glass material is itself easily recognized, once one has seen and handled it. It is water-white, that is, free of even the pale greenish or yellowish tints associated with iron impurities. These glasses were decolorized by the intentional addition of antimony, usually at a level of 0.5%-2.0% Sb₂O₅, and, apparently, by the selection of raw materials that introduced somewhat less iron than those used for making more ordinary glasses. From about the second century B.C. onward, manganese gradually replaced antimony as the decolorizer.

Hellenistic or earlier colorless glass objects excavated in Greece are especially rare. The author knows only of the cast pieces from Olympia, customarily associated with the workshop of Phidias;¹ the inlays from the Royal Tomb at Vergina; and rib-shaped pieces of glass (resembling those from Vergina) excavated at the Great Tomb at Lefkadia. Numerous colorless beads and small, flat pieces of glass were also found at the third-century B.C. glass factory site at Kakouli on Rhodes.² Samples of all these glasses have been analyzed chemically.

With the discovery of glassblowing, late in the first century B.C., glass in general became much more common, and so did decolorized glass. At those sites in Greece that yield glass of the Roman and Byzantine periods, colorless objects are sometimes found. We have analyzed many fragments of imported Roman-period vessels from Kenchreai, and of mosaic tesserae³ of considerably later dates from northern Greece. Both include intentionally decolorized glasses.

This paper concerns primarily the small pieces of colorless glass found among the treasures excavated in Tomb No. 2 of the great tumulus at Vergina. Vergina is generally regarded as the site of ancient Aegae, once the capital of Macedonia, and the tumulus is generally regarded as the royal cemetery of Macedonia.⁴ The tomb was opened by the late Prof. Manolis Andronikos in 1977· There is good reason to believe that this tomb was that of Philip II of Macedon, who died in 336 B.C.—although not all experts agree on this. Andronikos himself, in fact, seems always to have reserved some judgment on that point, stressing instead the luxurious quality of the finds in the tomb. On more than one occasion, he expressed this point of view to the author. The contents of the tomb were befitting the possessions of royalty, and the burial certainly dates from the mid- to late fourth century B.C.

The glasses appear to have been bits of ornamentation, decorated themselves with gilding and painting, and bearing the residue of bituminous adhesives. They are well melted, relatively free of imperfections, and moderately weathered.

The glass from Olympia, although known to have been cast and shaped in Greece, may or may not have been actually made in Greece. Similarly, there is a question concerning the origin of the Vergina glass. The glass objects from Vergina (and Lefkadia) could have been made entirely in Greece, shaped in Greece from material made elsewhere, or even imported in finished form. The studies described below were undertaken in the hope of shedding light on the question of which of these three possibilities applies to the Vergina glass. As will be seen below, however, the findings of chemical analyses and lead-isotope analyses are a bit perplexing when the three crucial pieces of evidence are brought together and compared with the little we already know about glasses of this type.

Chemical Analyses

Quantitative chemical analyses were carried out for five representative samples of colorless glass from the Royal Tomb at Vergina. These samples were selected by Professor Andronikos, Konstantin Assimenos, and the author in 1979. They are described under "Sample Descriptions" and illustrated in Figs. 1 through 6. The results of the analyses are reported in Table 1. For comparison, Table 2 gives analyses of some colorless glasses from other sites, and Table 3 lists additional examples of colorless glasses.

There are four things one looks for in chemical analyses of glasses of this period. The first is the general compositional family and the levels of major constituents.⁵ Like almost all ancient glasses, the Vergina glasses are of the soda:lime:silica type (Na₂O:CaO:SiO₂). There is nothing unusual about this, or about the levels of the three major components in the Vergina glasses.

Secondly, one looks at the levels of potassia and magnesia (K₂O and MgO) because two distinctly different levels of these oxides are found in ancient soda-based glasses, depending upon whether the alkali ingredient used was natron or plant ash.⁶

Table 1
Chemical Analyses of Some Glasses from Vergina

	3750 Beveled Rod	3751 Small disk	3752 Flat	3754 Large disk	3755 "Rib"	Mean Composition
SiO2	≈62.4%	≈66.2	≈63.5	≈67.2	≈67.3	65.3
Na2O	20.1	18.7	19.4	17.6	17.6	18.7
CaO	8.39	7.12	9.40	7.12	7.48	7.90
K2O	0.47	0.82	0.53	0.85	0.74	0.68
MgO	0.59	0.63	0.61	0.68	0.36	0.57
Al2O3	2.33	2.37	2.24	2.50	2.48	2.38
Fe2O3	0.81	0.66	0.67	0.63	0.58	0.67
TiO2	0.15	0.15	0.15	0.15	0.15	0.15
Sb2O5	2.63	1.27	1.79	1.24	1.15	1.62
MnO	0.077	0.065	0.068	0.071	0.061	0.068
CuO	0.02	0.02	0.02	0.02	0.02	0.02
SnO2	0.003	nf	nf	nf	nf	nf
Ag2O	0.05	0.02	0.02	0.05	0.02	0.03
PbO	1.48	1.51	1.11	1.46	1.56	1.42
BaO	0.02	0.02	0.02	0.02	0.04	0.02
SrO	0.05	0.05	0.05	0.05	0.05	0.05
B2O3	0.03	0.02	0.02	0.02	0.02	0.02
Cr2O3	0.002	0.001	0.002	0.005	0.002	0.002
ZnO	0.026	0.021	0.021	0.020	0.018	0.021
ZrO2	<0.01	<0.01	<0.01	<0.01	<0.01	<0.01
P2O5	0.30	0.26	0.36	0.28	0.28	0.30
Reduced Compositions
SiO2*	65.62	68.61	65.89	69.57	69.72	67.88
Na2O*	21.14	19.37	20.15	18.23	18.23	19.42
CaO*	8.82	7.38	9.76	7.37	7.75	8.22
K2O*	0.49	0.85	0.55	0.88	0.77	0.71
MgO*	0.62	0.65	0.63	0.70	0.37	0.60
Al2O3*	2.45	2.46	2.33	2.59	2.57	2.48
Fe2O3*	0.85	0.68	0.70	0.65	0.60	0.70
			nd 1.528		nd 1.516

Notes: All analyses by B.A. Rising and co-workers of Umpire and Control Services Inc., West Babylon, N.Y.
Si02 estimated by difference from 100%.
nf = not found
Sought but not found at 0.01% level: Li, Rb, Co, V, Ni, Bi, As.
Refractive indices by Chris Welker of Corning Incorporated.
Precision ±0.002.
Reduced compositions are the seven major and minor oxides normalized to 100%. The oxides are designated by an asterisk (*).

Table 2
Comparisons of Chemical Analyses

	Vergina (n=5)	Rhodes (colorless) (n=5)	Phidias's workshop 395	Great Tomb Lefkadia 501	Sardis 3470	Hellinistic bead 3770
SiO2	≈65.3%	≈69.5	≈69.1	≈66.6	71.3	75.01
Na2O	18.7	16.4	18.5	18.6	18.4	13.7
CaO	7.90	8.56	6.68	8.79	6.97	6.74
K2O	0.68	0.92	0.61	0.60	0.23	1.04
MgO	0.57	0.60	0.93	0.70	0.45	2.07
Al2O3	2.38	1.98	0.82	2.26	0.41	0.38
Fe2O3	0.67	0.37	0.75	1.09	0.29	0.34
TiO2	0.15	0.09	0.08	0.10	0.1	0.10
Sb2O5	1.62	0.71	2.46	1.0	1.56	0.46
MnO	0.068	0.05 (0.68)	0.04	0.005	0.06	0.046
CuO	0.02	0.002	0.005	0.01	0.003	0.003
SnO2	nf	nf	nf	0.05	0.001	0.002
Ag2O	0.03	<0.001	<0.001	<0.001	0.002	0.003
PbO	1.42	nf-0.003	0.005	0.01	0.02	0.01
BaO	0.02	0.06	0.003	—	0.01	0.01
SrO	0.05	0.08	0.005	—	0.03	0.02
Li2O	nf	0.002	0.001	—	0.001	0.01
Rb2O	nf	nf	nf	—	0.01	nf
B2O3	0.02	0.03	0.03	0.02	0.02	0.02
Cr2O3	0.002	nf	nf	nf	0.01	0.01
ZnO	0.021	0.0086		0.052	0.017	0.006
ZrO2	<0.01	<0.01	<0.01	nf	0.01	0.01
P2O5	0.30	ns	ns	—	0.08	ns
Reduced Compositions
SiO2*	67.88	70.68	70.94	67.52	72.72	75.56
Na2O*	19.44	16.68	19.00	18.86	18.77	13.80
CaO*	8.21	8.71	6.86	8.91	7.11	6.79
K2O*	0.71	0.94	0.63	0.61	0.23	1.05
MgO*	0.59	0.61	0.96	0.71	0.46	2.08
Al2O3*	2.47	2.01	0.84	2.29	0.42	0.38
Fe2O3*	0.70	0.38	0.77	1.11	0.30	0.34

Notes: nf = not found
ns = not sought
Sought but not found at 0.01%: Co, V, Ni, Bi, As.
— = insufficient sample

Table 3
Some Glasses That Resemble the Vergina Glasses in Chemical Composition
(All are colorless)

Object	Source and Date	Anal. No.
Nugget	Olympia, 5th c. B.C. Phidias's workshop (natron; contains Sb2O5; no PbO)	3395, 450
Phiale	Persia, 450–400 B.C. (natron; contains Sb2O5; no PbO) CMG 59.1.578*; SMG no. 248.	29
Achaemenid bowl	Persia, 450–400 B.C. (natron; contains Sb2O5; no PbO) CMG 59.1.67*; SMG no. 249	31
Omphalos bowl	Gordion, late 8th c. B.C. (4000; G206) (plant ash; contains Sb2O5; no PbO)	20
Painted plaque	Ft. Shalmaneser, 6th c. B.C. (plant ash; contains Sb2O5 and PbO)	1712(Pb-457)
4 vessel fragments	Kenchreai, 3rd–4th c. A.D. (natron; contain Sb2O5; no PbO)	773, 3715, 3717, 3718
Vessel fragment	Cosa, probably Roman (natron; contains Sb2O5; no PbO)	3030
5 fragments and beads	Rhodes bead factory, 3rd(?) c. B.C. (natron; contain Sb2O5; no PbO)	3503–7
Fragment	Lefkadia, about 300 B.C. Great Tomb (natron; contains Sb2O5; no PbO)	501

* The Corning Museum of Glass accession number.
SMG- Sidney M. Goldstein, Pre-Roman and Early Roman Glass in The Corning Museum of Glass, Corning: The Coring Museum of Glass, 1979.

Natron is a naturally occurring soda found in Egypt.⁷ Its use led to values of K₂O and MgO generally less than 1%, or only slightly greater, in the resulting glasses. When plant ash was used as an alkali, the K₂O and MgO values were about 2%-4% and 2%-6%, respectively. As a rough guideline, glasses from Italy, the Rhineland, Egypt, and (sometimes) the Levant are of the natron type. Glasses from more easterly sources, such as Mesopotamia, Persia, and (sometimes) the Levant are of the plant-ash type. Not surprisingly, the Vergina glasses are low in K₂O and MgO, indicating that they were made from natron, and that they are more likely to have been made in Egypt or elsewhere around the Mediterranean shores than in Persia or Mesopotamia.

A third thing to look for is the presence of colorants or decolorizers. The Vergina glasses contain no colorants, but they are decolorized with antimony. This is to be expected for colorless glasses of the period of the tomb, and it is consistent with their intended use in luxurious surroundings.

Finally, one always looks for unusual compositional features, such as the presence of uncommon trace elements or additives, or especially high or low percentages of the more common trace elements. The most striking feature of the Vergina glasses in this respect is the presence of lead oxide (PbO) at levels of about 1.5%. While this much lead occasionally finds its way into colored glasses, it is surprising to find this much in a colorless glass. Most colorless glasses contain only about 0.0X% PbO or less.

At a concentration level as great as 1.42% (the mean value of the five Vergina specimens), the lead seems to be an intentional additive; but, on the other hand, that much lead would not have had much effect on the physical properties or melting behavior of the glass. Therefore, it is difficult to understand why it would have been added. If the lead came in accidentally, it could have been through the reuse of cullet or, conceivably, as a contaminant of the antimony. One possible source of contamination to be considered is the yellow colorant-opacifier lead antimonate (Pb₂Sb₂O₇). However, this would have left some remains of yellowish flakes in the glasses, and such remains definitely are not present. Therefore, that possibility can be ruled out. The glasses analyzed are, in fact, of conspicuously good quality, free of inclusions or seed (an excessive number of bubbles). So the presence of the lead remains puzzling.

There is one interesting precedent for the lead content: a seventh-century B.C. colorless miniature glass inlay plaque from Fort Shalmaneser at Nimrud,⁸ which contains 2.63% PbO. Chemically, that glass resembles the Vergina glasses, except that it was made from plant-ash soda. As with the Vergina glasses, we can offer no explanation of why the lead is present in the Nimrud specimen. It is worth noting that the well-known colorless, hemispherical bowls from Nimrud are also made from plant-ash soda, but do not contain lead.⁹

When an ancient glass is found to contain silver, it is usually assumed to have been introduced into the glass as an impurity of lead, copper, or manganese additives. At a level of 0.06%, the silver in two of the Vergina glasses is noticeably higher than usually found. If it came in only with the lead, that lead would have been remarkably rich in silver. Possibly it is associated with the silver-leaf decoration observed on some of the glass. The glasses were cleaned thoroughly before being analyzed, so there should have been no such contamination, but perhaps some already decorated pieces had once been resoftened to be reworked, and in that case some traces of silver could have been picked up by the glass.

The five Vergina samples are very close, but not identical, to one another in composition. Samples 3750 and 3752 are very similar to each other. They differ slightly from the other three, which, in turn, are very similar to one another. However, the variations are about what are to be expected for day-to-day variations among glasses made in a single small factory in ancient times.

The Vergina compositions resemble those of a sampling of five colorless glasses from the Kakouli factory in a general way, but not so closely as to constitute a case for arguing that they were made there. However, given the fact that a group of five samples could hardly be thought to be representative of the output of the Kakouli factory over an extended period of time, that possibility cannot be ruled out. The one important discrepancy is that the glasses from the factory site do not contain the lead oxide present in the Vergina glasses, but they were decolorized, or fined, with antimony oxide. A comparison with the glass from Phidias's workshop presents about the same sort of match as does that with the Kakouli specimens, but lead oxide is also absent in the Phidias glass. We have analyzed only a single specimen from that site, but we have plans for analyzing more.

Sample 501 in Table 2 is from one of the rib-shaped pieces of glass from the Great Tomb at Lefkadia. This sample was given to the author for chemical analysis by Gladys D. Weinberg, who had obtained it from the excavator, Photios Petsas. It can be seen that this glass has a composition very closely resembling that of the Vergina glasses, except that it contains very little lead oxide, and it contains a trace of tin. Except for the fact that it contains so little lead, this glass could well be considered to have come from the same factory as the Vergina glasses. It would be extremely interesting to have analyses of additional glasses from the Lefkadia tomb, or from contemporaneous sites elsewhere in the area, to see how they compare with the Vergina glasses. In this connection, we were pleased to learn, as this paper was going to press, that D. lgnatiadou and E. Mirtsou of the Archaeological Museum of Thessaloniki are undertaking analyses of other glasses from this region.¹⁰

Table 2 also reports analyses of two other natron-based colorless glasses containing antimony. One is a Hellenistic bead of unknown provenance, and the other is a seventh-sixth-century B.C. piece of cullet from Sardis.¹¹

Figures 1-3 are graphical representations of the data for the Vergina glasses and a selection of other early colorless luxury glasses containing antimony. In addition to the glasses mentioned above, the graphs also contain data for nine conical bowls from Anafa. These samples were provided by David Grose, who dates them to about 150-75 B.C. Their reduced compositions are not very different from those of the Vergina glasses, although they contain manganese rather than antimony.

In summary, then, the Vergina glasses were made from natron, suggesting a "more westerly" origin; they are decolorized with antimony; they resemble the Phidias and Lefkadia glasses, except for the puzzling presence of lead; and they were made in a single factory—unfortunately, one whose location remains unknown.

Lead-Isotope Determinations

The determination of lead-isotope ratios is a useful method of scientific examination that yields information related to the provenance of ancient objects.¹² Lead is extracted chemically from minute samples of any lead-containing material or artifact, and the isotope ratios are determined by mass spectrometry. The isotope ratios are then compared with ratios determined for other artifacts and for galena (lead sulfide) ores from ancient mining regions. The objects can then be classified according to those containing leads that might have had a common geographical origin. Properly interpreted, these findings offer valuable clues as to where the objects themselves might have been made. In the most favorable instances, the actual mining regions from which the leads came can be identified. Two complications of the method are overlapping and mixing. Overlapping refers to the fact that lead ores from different mining regions sometimes have very similar isotope ratios. Mixing means that when leads from different sources are recycled and melted together, the resulting isotope ratios are somewhere between those of the starting leads.

Figure 4 (above) summarizes the results of some 1,200 samples of various types of ancient materials analyzed previously at the laboratories of the National Institute of Standards and Technology in Gaithersburg, Maryland. The caption explains the significance of the groupings.

Table 4
Lead-Isotope Ratios

Sample No.	208Pb/206Pb	207Pb/206Pb	204Pb/206Pb
Glasses
Pb-1144 (3751)	2.0642	0.83261	0.052940
Pb-1145 (3754)	2.0614	0.83165	0.053013
Pb-1147 (3755)	2.0611	0.83184	0.053075
Pb-457 (Ft. Shal. glass)	2.0664	0.8235	0.05231
Pb-543 (Ft. Shal. paint)	2.0721	0.8242	0.05222
Ores from Northern Greece
Pb-884 Pontokerasia*	2.0681	0.83240	0.053143
Pb-885 Olympias*	2.0689	0.83301	0.053169
Pb-886 Madem Lakkos*	2.0705	0.83392	0.053264
Pb-887 Therme Xanthi*	2.0814	0.83701	0.053380
Pb-888 Kirki	2.0773	0.83699	0.053463
Pb-889 Pefka	2.0754	0.83659	0.053407

Deposits known to have been worked in ancient times.

Samples of lead were extracted from three of the Vergina glasses, and their isotope ratios were determined. The results are tabulated in Table 4 and plotted in Figure 5. All three samples were found to contain the type of lead designated as Type L. This is the type of lead known to occur in the mines at Laurion and in its immediate vicinity. This same type of lead has been found in about 60 samples of ancient leads with proven (or reasonable) attributions to Greece. These 60 samples include metallic leads, bronzes, silvers, pigments, and slags dating from Mycenaean to Roman times. We have found only a few archaeological objects that do not have obvious or probable Greek connections, but nevertheless do fall in Type L.

Figure 5 also shows that the lead in the Fort Shalmaneser glass and that in traces of a white pigment adhering to it are of Type M, an isotopic type of lead found in many Mesopotamian glasses and other materials. This proves (to our satisfaction) that both the glass and the paint were made in that region.¹³ A list of such objects is given in Table 5.

The fact that the Vergina glasses contain lead of the Laurion type implies that this lead, which somehow found its way into the glasses, was mined either at Laurion or in some other mining region having an isotopically similar ore. Until now, we have analyzed only one lead ore, other than the Laurion ores, that falls in Type L. This came from the island of Imbros. There may well be other sources of Type L lead in the Aegean region, but we have yet to find any. (The question of whether or not Type L ores occur in the Near East is discussed below.)

Lead-isotope ratios determined recently for six galena ores from mines in northern Greece were found to differ significantly from those of the Laurion ores. These ores were provided by Dr. Michael Nicolaou. At least four of them come from mines known to have been worked in ancient times. The mines are located at Pontokerasia, Olympias, Madem Lakkos, Therme Xanthi, Kirki, and Pefka (Alexandroupolis). All lie to the east of Thessaloniki. These ores are also plotted in Figure 5.

The importance of this finding is that it establishes that at least some of the leads found in northern Greece are readily distinguishable from Laurion lead. Until actual data are available to prove that Laurion-like leads also occur in northern Greece, we are inclined to believe that the lead in the Vergina glasses is most likely from Laurion itself, or from some as yet unidentified Aegean source of Type L lead.

There is some other evidence to be considered. Among dozens of ancient metallic leads, glasses, glazes, and other materials we have analyzed from Mesopotamia and Iran, all are markedly different from Type L lead, except for six samples that fall within Type L or on its border line. These are listed in Table 5. We are reluctant to conclude that these six leads came from Laurion, and we expect that there is, somewhere in that part of the world, an overlapping source of lead isotopically indistinguishable from Laurion lead. Consequently, as far as the lead-isotope data are concerned, there is a possibility that the Vergina glass could have been made in Mesopotamia or Iran. Nevertheless, we regard this as very unlikely. In this instance, chemical analytical evidence must take precedence over lead-isotope evidence. In view of the fact that the Vergina glass was made with natron, it is extremely unlikely that it could have been made in Mesopotamia or Iran because glasses from those regions simply were not made with natron—they were made with soda derived from plant ashes.

Table 5
Some Objects Having Type L Lead
Type L, the Laurion type of lead, contains about 60 samples, comprising galena ores from Laurion and archaeological objects having either proven or reasonable connections with Greece. Those objects date from Mycenaean through Roman times. The objects listed in this table do not have obvious connections with Greece, but they contain Type L lead. Their lead might have come from some mining region closer to Mesopotamia and Iran that produced lead istopically indistinguishable from Laurion lead.

Pb-202	Rimah, 1500 B.C.; lead wire.
Pb-1332	Rimah, 1500 B.C.(?); lead medallion.
Pb-1337	Susa, date uncertain; metallic lead.
Pb-1088	Babylon, Lion Panel, 7th c. B.C.; yellow glaze.
Pb-1099	Babylon, Ishtar Gate, 7th c. B.C.; yellow glaze.
Pb-2138	Hasanlu, about 900 B.C.; red opaque glass.

All of the evidence and reasoning above indicates that the lead in the Vergina glasses probably came from Laurion or possibly some nearby Aegean source still to come to light. It remains, then, not only to explain why the lead is present in the Vergina glass at all, but also to decide what can be inferred from the fact that that lead is Laurion lead. What one really has to do is find a place with a plausible contemporaneous natron-using glassmaking history where Laurion lead would be expected to have been in use. The usual catchall for attributions for such glass is Alexandria, but according to traditional thinking, its glassmaking reputation should not have begun until 332 B.C., postdating the Vergina tomb by a generation. In fact, in terms of the history of glassmaking, the Vergina finds beg the question—not often asked—as to just where pre-Hellenistic luxury glass was made outside of western Asia.

Returning to the Vergina glass specifically, we can only say as follows: If the glass was made in Macedonia, it was apparently made with Laurion lead, not with local lead; but if the Macedonian mines had not yet been opened, then Laurion lead would be one type expected to have been in general use there. If the glass had been made in Athens or the Peloponnesus and brought north, it would almost certainly have contained Laurion lead. If it was made somewhere in Egypt, it probably would have contained Laurion lead; if in Rhodes or along the Levant, it might have; if anywhere else, it probably would not have contained Laurion lead. In order of descending likelihood, then, we suggest that the glass was made (1) either in Macedonia by itinerant glassmakers or in Athens or the Peloponnesus, (2) in some early stage at Kakouli on Rhodes or in some similar factory, (3) in Egypt at some glassmaking center predating the flourishing of Alexandria per se, or (4) at some as yet to be identified glassmaking center along the Levant. We are quite confident that it was not made in Mesopotamia, Iran, or elsewhere in western Asia (because of its natron composition). In any event, the shaping of the glass into its final intended forms is most likely to have been carried out near the site where it was found, possibly by the same glass artists who made the glass material itself in one of the places mentioned above.

We look forward to the results of more comprehensive scientific research on Macedonian glass and, indeed, on the luxury glass of the fourth century B.C. and earlier in general. Such research could place our understanding of the Vergina glass on a more substantial basis, whether or not the speculation above is borne out.

Sample Descriptions

Glass Samples

Six specimens of glass were given to RHB for analysis by Prof. Manolis Andronikos. They were selected by Professor Andronikos, Konstantin Assimenos, and RHB during a visit to Vergina on August 2, 1979. The glasses are representative of the types found throughout the site. All the glass is somewhat weathered, and a few pieces are heavily weathered. Although the glasses have been decolorized, some show a very pale greenish tint in thicker sections. Several pieces have gold leaf or foil attached to them, and also a black resinous material or bitumen. These pieces all appear to have been used as inlays of some sort, and some of them have been shaped, ground, or beveled.

3750
Fragment of glass inlay, rod-shaped, with beveled edges. Colorless glass, moderately weathered, with traces of gold (and possibly silver?). Contains very few bubbles, all of which are small and spherical. (Same as Pb-1142.)

3751
Fragment of circular glass inlay, disk-shaped, with ground edges, ground-flat base, and convex upper surface. Colorless glass, moderately weathered, with traces of gold leaf. Contains one small spherical bubble. Surface has traces of a yellowish or pinkish waxy substance. Apparent D. ≃ 1.5 cm. Described by KA as "a small lens." (Same as Pb-144.)

3752
Fragment of flat glass inlay. Colorless glass, heavily weathered, with gold foil attached and a heavy layer of bitumenous substance or possibly a charred resinous substance. Contains well-formed MnO₂ dendrites. Appears originally to have been flat piece about 2.5 cm square. Several of these were found in groups of nine squares, each covered with gold ornamentation. (Same as Pb-1146.)

3753
Fragment of large, circular glass inlay. Color indeterminate, but probably colorless; completely weathered, with no glass remaining. Similar pieces had a diameter of ≃3.5 cm. Described by KA as "a large lens."

3754
Fragment of circular glass inlay, thought to be identical in shape to 3753. Colorless glass; heavily weathered, but with some glass remaining. (Same as Pb-1145.)

3755
Fragment of "rib-shaped" glass inlay. Colorless glass, moderately weathered. W. ≃2.0-2.5 cm, Th. 3 mm. Shape is similar to larger "rib-shaped" glass pieces from the Great Tomb at Lefkadia, CMG 501. (Same as Pb-1147.)

Ores from Northern Greece

This group of galena ores from six present-day mines in northern Greece was submitted on June 18, 1981, by Dr. Michael Nicolaou of The Hellenic Chemical Products and Fertilizers Company Ltd., 20, Amalias Avenue, Athens 118, Greece. The samples are not of pure galena, but galena is the predominant mineral. The first four samples are from mines known to have been exploited in ancient times. (Ancient slags were found there.) It is not known whether the other two mines—Pefka and Kirki—were worked in ancient times.

Pb-884-Pontokerasia.
Pb-885-0lympias.
Pb-886-Madem Lakkos.
Pb-887-Therme Xanthi.
Pb-888-Kirki.
Pb-889-Pefka (Alexandroupolis).

---

This article was published in the Journal of Glass Studies, Vol. 36 (1994), 11–23.

---

Author's Note. This paper is dedicated to the late Prof. Manolis Andronikos, who provided the samples analyzed and appeared pleased with a preliminary report of the findings, dated Nov. 10, 1980; and to the late Dr. I. Lynus Barnes, who supervised the lead-isotope analyses of the samples at the National Bureau of Standards in Gaithersburg, Md. in 1980.

1.Jürgen Letsch, Walter Noll, and Wolfgang Schiering, "Glasformgebung in Tonmatrizen: Eine meisterliche Technologie der Werkstatt des Phidias in Olympia," Glastechnische Berichte, v. 56, 1983, pp. 96-105.

2. G.D. Weinberg, "Glass Manufacture in Hellenistic Rhodes," Arkaiologikon Deltion, v. 24, 1969, pp. 143-151 and pl. 76-88.

3. Robert H. Brill, "Scientific Studies of the Panel Materials," in Kenchreai, Eastern Port of Corinth, by L. Ibrahim, R. Scranton, and R. Brill, Leiden, the Netherlands: E.J. Brill, 1976, pp. 225-255. The data on the Byzantine tesserae are unpublished.

4. For information on the subject in general, see: M. Andronikos, The Royal Graves at Vergina, Athens: General Directorate of Antiquities, 1978; idem, "The Royal Graves in the Great Tumulus," Athens Annals of Archaeology, v. 10, 1977, pp. 1-39; idem, "The Royal Tombs at Vergina: A Brief Account of the Excavations," The Search for Alexander: An Exhibition, Washington: The National Gallery of Art, 1980, pp. 26-38, passim; and Eugene N. Borza, "The Macedonian Royal Tombs at Vergina: Some Cautionary Notes," Archaeological News, v. 10, 1981 , pp. 73-87.

5. For information on compositional families and interpretations of chemical analyses, see Robert H. Brill, "Scientific Investigations of the Jalame Glass and Related Finds," in Excavations at Jalame, Site of a Glass Factory in Late Roman Palestine, ed. Gladys Davidson Weinberg, Columbia: University of Missouri Press, 1986, chap. 9, pp. 257-294; and idem, "Introduction," in Scientific Research in Early Chinese Glass, eds. R.H. Brill and J.H. Martin, Corning: The Corning Museum of Glass, 1991, pp. vii-ix.

6. For plant-ash analyses, see Robert H. Brill, D. Barag, A.L. Oppenheim, and A. von Saldern, "The Chemical Interpretation of the Texts," Glass and Glassmaking in Ancient Mesopotamia, Corning: The Corning Museum of Glass, 1971, pp.105-128.

7. For natron analyses, see Brill, "Scientific Investigations" [note 5].

8. Robert H. Brill, "Some Miniature Glass Plaques from Fort Shalmaneser, Nimrud. Part II: Laboratory Studies," Iraq, v. 40, Spring 1978, pp. 23-39.

9. Ibid. Interestingly, although the MgO values are quite high for both the bowls and the Fort Shalmaneser inlay, the K₂O values are only about 1.5%. This combination is not unlike some contemporaneous Egyptian glasses.

10. By private communication, July 16, 1993.

11. This fragment was found with the red opaque glass from Sardis described in R.H. Brill and N.D. Cahill, "A Red Opaque Glass from Sardis and Some Thoughts on Red Opaques in General," Journal of Glass Studies, v. 30, 1988, pp. 16-27.

12. For some recent references on the subject relating to applications to glass, see I. Lynus Barnes, Robert H. Brill, Emile C. Deal, and G. Venetia Piercy, "Lead Isotope Studies of Some of the Finds from the Serҫe Liman Shipwreck," in Proceedings of the 24th International Archaeometry Symposium, ed. Jacqueline S. Olin and M. James Blackman, Washington: Smithsonian Institution Press, 1986, pp. 1-12 ; Robert H. Brill, I. Lynus Barnes, and Emile C. Joel, "Lead Isotope Studies of Early Chinese Glasses," in Scientific Research in Early Chinese Glass, eds. R.H. Brill and J.H. Martin, Corning: The Corning Museum of Glass, 1991, pp. 65-83; and Robert H. Brill, "Scientific Investigations of Ancient Asian Glass," Proceedings of the Nara Symposium '91, UNESCO Maritime Route of the Silk Roads, Nara, March 7, 1991, pp. 70-79.

13. For information on Type M leads, see Christine Lilyquist and Robert H. Brill, Studies of Early Glass in Egypt, New York: The Metropolitan Museum of Art, 1993, part 3, pp. 57-73.

Published on July 11, 2013