This article was downloaded by: [Facsády, Annamária R.] On: 27 July 2009 Access details: Access Details: [subscription number 913404540] Publisher Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Materials and Manufacturing Processes Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t713597284 Analysis of Roman Bronze Finger Rings from Aquincum Annamária R. Facsády a; Anett Verebes a Department of Antiquity, Aquincum Museum, Budapest, Hungary a Online Publication Date: 01 September 2009 To cite this Article Facsády, Annamária R. and Verebes, Anett(2009)'Analysis of Roman Bronze Finger Rings from Aquincum',Materials and Manufacturing Processes,24:9,993 — 998 To link to this Article: DOI: 10.1080/10426910902979942 URL: http://dx.doi.org/10.1080/10426910902979942 PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Materials and Manufacturing Processes, 24: 993–998, 2009 Copyright © Taylor & Francis Group, LLC ISSN: 1042-6914 print/1532-2475 online DOI: 10.1080/10426910902979942 Analysis of Roman Bronze Finger Rings from Aquincum Annamária R. Facsády and Anett Verebes Department of Antiquity, Aquincum Museum, Budapest, Hungary The majority of rings recovered in Aquincum (Budapest), the capital of the Roman province Pannonia Inferior, in today’s Hungary, are made of bronze. Finger rings can be classified in eleven groups by their shapes. In addition, their composition allows the distinction of eight groups. Concerning the correlation between the types and alloys, it can generally be concluded that there is only rarely a close relationship between the deliberate selection of the alloy and the decoration technique of the jewelry. Finger rings from Aquincum made of copper-based alloys always contain some zinc, albeit in variable quantities. Indeed, the most common alloy employed for the largest number of ornaments and with the greatest variety of shapes is brass, the alloy of copper, and zinc. The high number of alloys with a different composition indicates that there was a significantly increased demand for jewelry similar in color to precious metal, but cheaper and easy to produce. Its manufacture did not require a specialized workshop. Downloaded By: [Facsády, Annamária R.] At: 10:16 27 July 2009 Keywords Aquincum; Copper-based alloys; Roman bronze jewelry; Roman finger rings; SEM/EDS analysis; Types of Roman finger rings. used throughout the Roman Empire. However, in Aquincum simpler types seem to have been preferred. They were made of both organic and inorganic materials, such as amber, jet, glass, gold, silver of different composition, iron, and even lead, but the great majority of them, i.e., 80%, were made of copper-based alloys. As mentioned before, in the case of rings, determining the date of the pieces is particularly difficult. Their shape or decoration, the main criteria for their typological classification, tended to be in use for a long time, but was altered and became simpler at the end of the period in which they were fashionable. The earliest finger rings discovered in Aquincum can be dated to the 2nd century A.D. Their design is typically Roman, they are undecorated pieces of circular or slightly oval shape, primarily created to hold and preserve an engraved gemstone. This shape continued to be also used in later times, but from the Severian period on, at the end of the 2nd and the beginning of the 3rd century A.D., there was a considerable increase in number and shapes of finger rings. At the end of the 3rd century, the variety of the shapes became even greater, but at the same time they were simplified both in terms of workmanship and material. Among the funerary gifts of 4th century graves there is a steep rise in the number of undecorated band and wire finger rings. In Aquincum, on the basis of shape, proportions, cross section, and decoration, it is possible to distinguish eleven types of finger rings [3]. All groups include pieces made of copper alloys, but, in the different groups, the number and proportion of pieces made of other materials varies. Characteristic shape and decoration, which define each type, are most evident in the case of pieces made of precious metals, and in particular of gold. The manufacture and decoration of bronze jewelry considerably differed from that made of gold. Even if they had the same shape, the decoration was often produced in a different way and as imitation of the goldsmith technique. This is for example the case with cast bronze finger rings which imitate granulation. Introduction The investigation of Roman metallurgy, the localization of workshops, the classification of their products, and the reconstruction of the production techniques [1, 2] has focused primarily on statues, metal tableware, brooches, and objects made for the army supply. The typological investigation and chemical analysis of objects used mainly as ornaments are pushed into the background. This fact can mainly be explained by the characteristic features of the finds. The production techniques of jewelry do not change much for a long period of time for many types of items (Fig. 1). Minor differences can be normally observed in the quality of the design only. The periods of manufacture, apart from a few exceptions with particular objects, are therefore rather hard to define. Because of this, jewelry pieces cannot be easily dated, and therefore, they have been ignored in most studies. Types of finger rings from aquincum A large proportion, approx. 1/3, of the so-called body jewelry discovered in Aquincum (Budapest), the capital of Pannonia Inferior in Roman times, consists of finger rings. The rings have various names and bear different symbolic significance. They are not only numerous, but there is also a considerable variation in their shape. The variety of finger rings is indeed huge, and goes from the most simple closed rings to the elaborate goldsmith’s work, decorated by different techniques. Analogies to the pieces from Aquincum can be found among finds of remote provinces, dated to different periods. The rings from Aquincum generally follow the typology of rings Received July 10, 2008; Accepted January 27, 2009 Address correspondence to Annamária R. Facsády, Department of Antiquity, Aquincumi Múzeum, Záhony utca 4, Budapest 1031, Hungary; E-mail: facsady@ aquincum.hu 993 Downloaded By: [Facsády, Annamária R.] At: 10:16 27 July 2009 994 A. R. FACSÁDY AND A. VEREBES Figure 1.—Types of finger rings from Aquincum. Some decoration techniques were mainly or exclusively typical for bronze jewelry, such as for instance chiselling, enamel, and gilding. The composition of finger rings Fifty-seven pieces out of the 78 finger rings inventoried as “bronze” rings in the collection of the Aquincum Museum have been analyzed in the Central Research Institute for Physics in Budapest by Dr. Attila Tóth. The samples were prepared and the results evaluated by Vecsey Ádám (Aquincum Museum). The amount of sample required for the analysis was a maximum of 1 mm3 , sawed with a 0.3 mm wide blade. The samples, placed on a 0.8 mm thick and 15 × 15 mm wide aluminium plaque of great purity, were analyzed by electron probe X-ray microanalysis in the scanning electron microscope (SEM/EDS). Only the main elements (copper, zinc, tin, lead, iron, and silver) were considered in the evaluation of the analysis results (Table 1). Further elements were also detected; however, the low results indicate that they were not deliberately added to the alloy. The table does also not include elements, which are due to metal corrosion or to conservation by organic materials (S, Cl, C, etc.). The aim of the analysis of bronze rings by SEM/EDS was to get the answers to several questions. In addition to the possible identification of traces that would reveal certain phases of the production technique (for instance the identification of soldering material) and coating, we were in the first place interested to see if composition and shape could be correlated. A second point was the possibility to date the use of the alloys, and if they were restricted to certain periods. Further we tried to find out if it is possible to distinguish pieces that were certainly worn as jewelry from artefacts with different functions, but similar shape (e.g., rings, bands). As these are objects primarily used as ornaments, it is also important to evaluate the color of the alloys. Was the proportion of the mixed metals the result of deliberate choice, and was the colour of the alloy similar to that of precious metals? Only by an extended investigation, i.e., by analyzing more pieces belonging to other groups of jewelry it would be possible to determine differences in composition among different types of jewelry, and if jewelry was manufactured in specialized workshops. Workshops which produced bronze objects are also known from Pannonia [4] and workshops for repairing bronze objects have also been discovered in the territory of the Aquincum municipium. The questions related to workshops might be answered in the future, when we will widen our investigations. The majority of the analyzed objects were made of one single metal: no coating was observed. However, during the various conservation treatments in the past we cannot exclude that the coating was removed. Some identified elements originated from intercrystalline corrosion. No traces of soldering were determined. As no soldering traces were found, and the cross-sections are regular all around, we infer that the models for the rings were made of wax cylinders. The iron detected in the analyzed objects has to be considered as impurity only or perhaps due to corrosion. The iron content is not significant and, with the exception of four pieces, it is below 1%. Three pieces also contain arsenic and silver impurities. All specimens contain some zinc. In the case of 14 objects it was identified only at trace level. In six items the zinc content was found to be between 3 and 5%, while in 5 further pieces the Zn results were between 5 and 10%. The majority of the analyzed objects (i.e., 22 pieces) contain 10 to 20% Zn. In the case of 4 rings, the zinc content is above 20%, with max. 25.4 and 25.7%. In these pieces, the alloy is pure brass, without any deliberate addition of Sn or Pb. The tin content shows variable results: 995 ANALYSIS OF ROMAN BRONZE FINGER RINGS FROM AQUINCUM Table 1.—The results of the SEM/EDS analysis by types of finger rings. Downloaded By: [Facsády, Annamária R.] At: 10:16 27 July 2009 Aquincum Museum Inv. No. 40017 54102 56.161.65 56.19.42 54431 40020 54415 53847 56.23.8 56.185.215 56.185.182 55.36.1 56.200.103 40016 60004 40001 56.19.45 54101 40007 66.6.166 55274 56.185.183. 56.40.20 55.36.2 40013 55.38.20 40021 56.117.4 54411 56.185.207 56.185.66 56.185.61 40082 47/13508 54434 40088 56.185.102 40083 40084 40090 40097 40095 40087 40091 54104 (a) 54104 (b) 62.2.4 54.17.17 56.184.42 56.185.70 56.185.68 86.7.480 56.171.3 53751 54100 64.8.6 54103 Ring Type I/a. I/b. I/c. II/a. II/b III/a. III/c. III/c IV/a IV/c. IV/c. IV/c IV/e. IV/e. V/b. VI/a. VI/a. VI/b. VI/b. VI/c VII. VIII/A VIII/A VIII/A VIII/B VIII/B VIII/B VIII/B VIII/B VIII/C VIII/C IX/A IX/A IX/A IX/A IX/A X/A X/A X/A X/A X/A X/A X/A X/A X/A X/A X/A X/A X/A X/B X/B X/C XI/A Cu Zn Sn 72.21 73.34 84.14 84.1 75.64 84.49 6.86 3.02 83.11 31.46 83.85 80.68 85.25 84.5 82.91 82.89 79.27 70.06 75.82 19.33 95.88 84.09 47.12 76.48 81.82 83.37 73.06 82.7 82.79 88.02 75.69 84.26 72.92 74.1 72.49 90.03 8.33 82.78 71.64 77.95 80.77 75.14 79.25 75.73 85.88 71.79 85.09 75.32 90.81 78.42 77.92 80.11 82.6 82.22 78.99 76.97 81.71 4.7 22.86 11.6 7.36 16.96 5.1 0.75 0.13 4.31 0.69 13.67 17.71 12.06 10.2 13.7 14.32 19.53 8.41 19.32 0.71 0.93 12.23 0.43 20.73 5.12 12.8 25.73 2.15 0.55 1.35 10.06 13.15 0.87 18.89 1.32 0.63 0.96 0.58 10.34 6.18 2.75 1.17 0.38 0.57 11.64 25.45 3.95 3.26 0.85 15.03 17.13 0.8 8.96 1.39 15.34 19.42 14.94 14.32 0.99 0.71 2.96 3.23 6.87 4.81 Pb 1.15 1.94 3.13 2.84 1.41 3.43 0.18 6.40 0.56 1.23 1.25 0.97 10.88 1.72 1.87 1.06 0.68 10.36 1.99 10.87 0.22 1.59 0.5 6.18 0.89 1.05 1.58 2.51 1.38 0.6 5.85 1.69 7.72 5.73 1.58 5.34 0.5 2.81 1.25 16.87 0.87 11.93 0.99 7.42 9.68 6.63 5.89 7.18 8.83 7.29 8.16 6.67 7.77 9.4 15.39 10.83 15.55 7.42 1.55 5.39 9.57 7.11 0.99 12.52 5.17 7.63 1.52 0.65 12 pieces contain no tin at all, and 16 contain less than 3% of this element; 18 finger rings contain 5–10% Sn in the alloy, and 5 examples have more than 10%. The highest Sn content is 14.32%. In 8 cases, lead was not identified in the alloy, while the lead content of 26 further items was below 3%. The low Pb in these objects was interpreted as the result of the recycling of bronze. Eleven pieces contain 0.43 4.38 10.22 0.26 3.04 2.61 6.18 1.85 8.13 1.71 0.38 0.44 Fe 0.67 0.7 0.48 0.39 0.65 0.81 1.88 0.33 0.45 0.38 1.12 0.21 0.49 0.47 0.74 0.74 0.38 0.98 0.79 0.33 0.46 0.52 0.32 0.28 1.18 0.44 0.28 0.75 0.38 0.12 0.54 0.67 0.29 0.78 0.29 0.69 0.84 0.26 0.83 0.61 0.81 1.01 0.28 0.3 0.54 0.52 1.48 0.34 0.41 0.38 0.17 0.41 0.22 0.68 0.74 0.41 As 0.33 Ag 0.97 Alloy Group 6. 5. 5. 8. 8. 6. 75.94 95.18 8. 64.57 5. 5. 5. 5. 5. 5. 5. 8. 5. 75.31 1. 5. 47.67 0.07 2.49 88.55 0.23 1.95 5. 6. 5. 5. 2. 3. 2. 7. 5. 4. 5. 4. 2. 4. 8. 8. 8. 4. 4. 4. 5. 5. 8. 8. 2. 7. 7. 4. 6. 4. 5. 5. 5. 3–10% Pb, and 6 have more than 10% Pb. The maximum lead content among the analyzed pieces is 16.87%. The amount of copper used in the alloys ranges between 70.6 and 95.8%. 9 pieces contain 70–75% of Cu, 15 pieces contain 75–80%, while most of the analyzed objects (21) contain 80–85% Cu. Six of the finger rings were made of alloys containing 85–91% of Cu, and only one example was 996 A. R. FACSÁDY AND A. VEREBES Downloaded By: [Facsády, Annamária R.] At: 10:16 27 July 2009 made of almost pure Cu. In relation to the proportion of the alloys, 8 groups can be differentiated. (Pieces with a high percentage or with almost 50% of silver were not taken into consideration.) 1) Copper. The copper content is above 95%. Only one example belongs to this group, a finger ring of type VII. An almost pure copper was selected because it is very malleable and can be used to produce the thin thread employed for this ring (Fig. 2). 2) Copper–tin. The copper content is between 80 and 90%, but there is a small amount of zinc and tin in each alloy (most probably the result of recycling) (Fig. 3). Four pieces belong to this group from types VIII, IX, and X. 3) Copper–lead. The metal is nearly 83% Cu, with a small amount of zinc, and no tin; 1.76% Al detected by the analysis comes from the environment and the long burial. Only one piece belongs to this group, but this item, a 4 mm thick, undecorated band, made of a cheap alloy, is most probably a functional element of a larger object and cannot be considered as jewelry (Fig. 4). 4) Copper–tin–lead. Eight objects contain 70–80% Cu, 6– 12% Sn, and 6–17% Pb. They belong to the types IX and X. Also in this case the alloy is a cheap leaded bronze, not particularly suitable for ornaments (Fig. 5). 5) Copper–zinc. The majority of the analyzed pieces (21) belong to this group and to the types I, IV, V, VI, VIII, IX, and X. They contain 71–80% Cu and 11–26% of zinc (Fig. 6). These alloys are made of real brass and their color is similar to gold. As previous analyzes of Roman decorative objects have shown, brass was preferentially employed for ornaments [5]. 6) Copper–zinc–tin. The four finger rings belonging to this group contain 70–80% Cu, 4–9% Zn, and 5–14% Sn. These pieces of jewelry belong to the types I, III, VIII, and IX (Fig. 7). The metal used for these specimens was brass diluted with bronze, an alloy with good technical properties [5]. 7) Copper–zinc–lead. The 3 specimens belonging to this group contain 75–78% Cu, 10–17% Zn, and 3% Pb. They belong to the types VIII and X. This alloy could be easily cast, has a rather golden color, and is suitable for pieces of jewelry (Fig. 8). 8) Copper–zinc–tin–lead. The alloy with 70–85% Cu contains all three alloying elements, is most probably the consequence of intensive recycling, and was employed Figure 5.—Ring type IX. (Inv. No. 40082). Figure 6.—Ring type I. (Inv. No. 54102). Figure 2.—Ring type VII. (Inv. No. 55274). Figure 7.—Ring type VI. (Inv. No. 40007). Figure 3.—Ring type VIII. (Inv. No. 56.185.207). Figure 4.—Inv. No. 54411. Figure 8.—Ring type III. (Inv. No.40017). ANALYSIS OF ROMAN BRONZE FINGER RINGS FROM AQUINCUM Figure 9.—Ring type X. (Inv. No. 62.2.4). Downloaded By: [Facsády, Annamária R.] At: 10:16 27 July 2009 for 9 pieces of jewelry (Fig. 9). They belong to the finger ring types II, IV, VI, and X. Relationships between shape, period, and alloy Type I finger rings, which are considered “typical Roman” rings, are dated to the 2nd and 3rd centuries A.D. The ring head which holds the stone does not protrude from the shoulder line. The shoulders are undecorated and become evenly thicker towards the head. Eight pieces represent three subtypes of the shape. Of these, 3 are made of copper-based alloys. Two pieces belong to group 5, and one to group 6. In the case of type II, the head is higher than the shoulders, which do not thicken. The parallels to the 7 examples from Aquincum are made of both metal and jet and are dated to the 3rd century A.D. Two subtypes can be recognized among the seven specimen. Two pieces are made of alloy group 8, while the composition of the fragmentary piece, Inv. No. 56.19.42, slightly differs from the others as it contains arsenic and silver too. In the specimens of type III, the head is decorated and slightly protruding and the shoulders are evidenced. Finger rings of type III illustrate well the change of design with the altered and broken unity of body and head, which started at the beginning of the 3rd century. These rings can be dated between the late 2nd and the early 3rd century. They were generally fashionable towards the end of the Severian period and normally made of gold or silver, with the combination of different decorative techniques. Only one piece is made of a copper alloy belonging to group 6. In the case of type IV the shank of the ring is a nearly regular circle of even thickness. Its appearance can vary depending on the proportions, and it is made of wire or of a narrow (2–3 mm wide) band. The head is evidenced and decorated in various ways also on the side of the head. The piece Inv. No. 56.23.8 from a late Roman grave shows characteristics representative of the imitation and survival of design from goldsmith’s technology. The “granulation” on the side of the head was made by casting. Variants of the 3rd century design are also common in the 4th century. Similar items were discovered in a 4th century layer during the excavation of Novae, and in a grave in Intercisa. Their long employment is demonstrated by pieces dated to the second half of the 4th century, from the Somogyszil and Ságvári cemeteries. The four late Roman pieces were made of an alloy of copper and zinc (group 5), and one piece was made of a quaternary alloy (group 8). Type V is similar to type IV, but here the head and the ring shoulder play a primary role which is further stressed 997 by decoration. On the piece Inv. No. 60004 made of brass (group 5) the small grains imitating granulation remind of type IV/a. The majority of the rings belonging to the Guiraud 4/d type show a glass paste inlay, but this 3rd century type can also have an enamel decoration. In type VI the head forms a continuous curve with the circular or slightly oval shank of the ring, and it is wider than the lower part of the ring. The decoration is limited, as the motifs are engraved or punched. It belongs to the late Roman jewelry. Three pieces are made of group 5 alloy, and one shows the composition of group 8. Finger rings of type VII are made of a thin wire, and there are also bracelets and earrings produced in a similar way. They are not rare among the grave-goods of late Roman cemeteries. In Aquincum, a silver and a copper (Inv. No. 55274) example of this type has been discovered. The wire twisted on itself on the opposite end serves as decoration, but also works as a spring and allows the adjustament of the size. Only one piece belongs to this type, and it is made of almost unalloyed copper (group 1). Type VIII includes penannular finger rings. The ends of the wire are bent side by side or on top of each other. In the case of one subtype, they terminate in animal heads. With the exception of one copper, one silver and one iron piece, all are made copper-based alloys belonging to groups 2, 6, and 7. One specimen of this shape, made of a copper-based alloy (group 3) is not a finger ring, but a functional element of some larger object. Band finger rings classified as type IX are exclusively made of copper alloys. Their cross-section shows different shapes (circle and D-shaped). They are either chiselled or undecorated. Of this type, there are 23 examples and only two are made of silver, while the rest are made of copper alloys. One piece belongs to alloy group 2, two pieces belong to group 4, and two pieces to group 5. Closed finger rings of type X have circular or D-shape cross-section and are normally undecorated. This type was made of the following alloys: one piece was made of the alloy of group 2, five of group 4, two group 5, two of group 7, and five of group 8. It is not easy to date type IX and X, i.e., finger rings made of band or wire of different cross-section. Most of them come from graves and none of the other finds in the same burial is significant for the determination of its date. In the case of type XI finger rings, the ornament is attached to the loop with a short tang. The pieces known in Aquincum are made of copper-based alloys of group 6. The groups set up according to the analysis of Roman finger rings cannot be dated to a particular period. One of the reasons might be that the material was always recycled. However, the relatively large group of brasses suggests that the choice of material depended from the wealth of the single customers. The large number of alloys used demonstrates that different technologies and skills lived side by side in the investigated period. In addition to the bronze groups that have a low content of alloying elements (Cu–Zn, Cu–Pb, and Cu–Sn groups), there are also pieces of the Cu–Zn–Pb group containing nearly 3% Pb, which belong to the Cu–Zn group because of their relatively high zinc content. In Roman times lead was a common impurity in brasses and the artisans were probably not aware of its Downloaded By: [Facsády, Annamária R.] At: 10:16 27 July 2009 998 presence if the content was as low as this. Within the main group the objects of the Cu–Pb and Cu–Sn groups can be differentiated from the typical brass (Cu–Zn) finds on the basis of their very low zinc content. In Roman bronzes tin and lead are mixed to copper together [7], and this is reflected by one group of finger rings (Cu–Sn–Pb, 10% of the analyzed rings). The group of finds can be further narrowed on the basis of their tin and lead content. In this way, the following subgroups can be identified: five pieces have high lead content, one finger ring has high tin content, and two finds contain nearly the same amount of tin and lead. A high lead content is advantageous for the casting of complex shapes. We have to note that the technological skill meets the increased demand for cheap finger rings, and other types of jewelry, but it does not produce new shapes. All finger rings containing a high percentage of lead are band and wire rings, i.e., characteristic funerary gifts of late Roman graves. This means that the object was with all probability produced for funerary use only. Concerning the Cu–Zn–Sn and Cu–Zn–Sn–Pb groups, further trends in the change of the Zn content, already mentioned in the international scholarly literature, can be observed. In the finds from Aquincum, the proportion between Cu and Zn is around 80 to 20%; however, in the case of pieces of the Cu–Zn–Sn group there is some divergence: the amount of the alloying metals Sn and Zn together is between 20–10%. This means that cheaper alloys, produced by recycling scrap metal and adding copper, were employed. Also the pieces of copper-based jewelry from Augsburg show a similar composition and were interpreted as produced from recycled scrap metal. In the objects of the Cu–Zn–Sn–Pb group, the copper content remains high, its rate does not go below 70%, but the relative proportion of the alloying metals varies. This fact also clearly indicates recycling. Conclusions Finger rings of the early imperial period, 2nd century A.D., belong to groups 5 and 6, and are made of Cu–Zn and Cu–Zn–Pb alloys, respectively. The composition of pieces dated to the 3rd century A.D. is more varied. Jewels of different shapes were made of Cu (group 1), Cu–Sn– Pb (group 4), Cu–Zn (group 5), Cu–Zn–Sn (group 6), and the quaternary alloy Cu–Zn–Sn–Pb (group 8). Finger rings dated to the 4th century were also made of alloys belonging to the groups 2, 4, 5, 6, 7, and 8. However, in the 4th century A.D. finger rings were exclusively made of a Cu–Zn–Pb alloy (group 7). It is therefore rather difficult to identify a close relationship between alloys, time, and shapes. Alloys of different composition were used for wire or band finger rings, because of the different working procedure. Jewelry of the most varied forms was made of group 5 Cu–Zn alloys. The analysis has demonstrated that, in imperial times, brass was the preferred alloy employed for jewelry. The reason for this choice was the color of the alloy, which among nonprecious metals had a color which was most A. R. FACSÁDY AND A. VEREBES similar to gold. This fact played an important role in the appearance of the jewelry. The finger rings found in Aquincum are made of the typical Roman alloys commonly employed for the jewelry of the time and known from other researches published in the international literature. It should be noted that the group of rings made of a Cu–Sn–Pb alloy seems to be a local characteristic and was employed especially for cast finger rings. Finger rings were found to be made of a great variety of alloys. In many cases, this might be due to their small size and the small quantity of metal needed, so whatever alloy was at hand was employed for cheap jewelry. Nevertheless, as we have seen, well-defined groups can be observed among the finds. As their manufacture was rather easy, it is likely that they were produced in nonspecialised workshops. The characteristics of their material composition depended on the local casting traditions, the availability of the material, and the demands of the customers. Acknowledgments The analysis was carried out by Dr. Attila Tóth L. in the Central Research Institute for Physics. The samples were prepared for the analysis, and the results evaluated by Vecsey Ádám (Aquincum Museum). I wish to thank them here for their work. References 1. Condamin, J.; Boucher, S. Recherches techniques sur des bronzes de Gaule Romaine 4. Gallia 1973, 41, 157–178. 2. Giumlia-Mair, A. Roman metallurgy: Workshops, alloys, techniques and open questions. In Ancient Metallurgy Between Oriental Alps and Pannonian Plain, Giumlia-Mair, A. Ed., Workshop, Trieste, October 29–30, 1998, Associazione Nazionale per Aquileia, Trieste 2000; pp. 107–120. 3. Facsády, A.R. Az aquincumi ékszerek. Adatok a római testékszerek forma és anyag szerinti korszakolásához. Ph.D. értekezés. Kézirat. (Jewelry from Aquincum. On the periodization of Roman body jewellery by shape and material. Ph.D. dissertation. Manuscript). Eötvös Lóránd University, Institute of Archaeological Sciences, Budapest, 2003, pp. 12–19, Fig. 1. 4. Szabó, K. Officine del bronzo in Pannonia. In Ancient Metallurgy Between Oriental Alps and Pannonian Plain, Giumlia-Mair, A. Ed., Workshop, Trieste, October 29–30, 1998 Associazione Nazionale per Aquileia, Casa Bertoli, Trieste 2000; pp. 154–164. 5. Giumlia-Mair, A. Technical studies and chemical analysis of the Roman copper-based finds from Emona. Berliner Beiträge der Archäometrie 2001, 10, 5–42. 6. Költö, L. Avar kori bronztárgyak röntgenemissziós analizise (X-ray emission analysis of bronze objects from the Avar Age). Somogy Megyei Múzeumok Közleményei 1982, 5, 21. 7. Riederer, J. The use of standardised copper alloys in Roman metal technology. In I Bronzi Antichi: Produzione e tecnologia, Giumlia-Mair, A. Ed., Atti del XV Congresso Internazionale sui Bronzi Antichi, Grado-Aquileia, 22–26 maggio 2001, Édition Monique Mergoil, Montagnac 2002, Monographie instrumentum 21, 288–290.