When was silver nickel first made?



1 83 JOSEF RIEDERER ELKE BRIESE METAL ANALYSIS OF ROMAN USED OBJECTS From a find of metal objects recovered from the Tiber, above all nails, needles, fittings, fishhooks, fragments of figures and coins, numerous metal objects of daily use were made available to the Doerner Institute in Munich for metallurgical investigations 1), while the analysis of 130 sesterces and the investigation of sewing needles, for example, are described elsewhere. Knowledge of the alloys that were used to manufacture objects of daily use can give a clear picture of the metal technology of a certain time, since with these one did not have to take into account the complicated form structure, as in the case of sculptures, but only the function and usability, so that you could set up the choice of alloy components and the type of Metallbearbei processing accordingly. Since the majority of the objects examined were still in a perfectly usable, partly unused, condition, it can be assumed that they date from the middle of the 3rd century AD, like the most recent coins found with them. This time is particularly interesting for the history of metallurgy, as a few centuries earlier, shortly before the beginning of the imperial era, brass was deliberately used for the first time. Initially, only coins, namely sesterces and dupondia, were made from this new alloy. Shortly afterwards, the first small decorative and everyday objects appeared, which are certainly made from coins. The first statuettes were cast from brass later, but cannot be dated due to the small number of confirmed analyzes. It is noteworthy that the zinc content of the sesterces decreases more and more from emperor to emperor2). The sesterces of Augustus had an average of 22% zinc, those of Vespasian 18%, Hadrian 12%, Marc Aurel 8% and Commodus only 4%. One of the causes of this decrease is the evaporation of the zinc as older coins are melted down. However, it is still unclear why the later emperors only used scrap metal and not new zinc ore for their coins. Such questions about the metal budget of the Roman state can only be answered by a sufficiently large number of analyzes, to which the results reported here are intended to contribute. The authors are indebted to Mr. F. Waldner, Munich, for the permission to examine these objects. 2) J. Riederer, Metallanalysen Roman Sester ^ en. Year Numismatics Monetary History 4.1974 (in press). E. R. Caley, Orichalcum and related ancient alloys (1964). D. W. McDonall, The quality of Nero s orichalcum. Tail. Mün ^ bl. 16, 1966, 1 o 1 ff.

2 84 The metal analyzes were carried out at the Doerner Institute with the help of atomic absorption spectral analysis according to a method developed by E. Briese for the smallest sample quantities. Only 5-20 mg of substance - about the amount of a large pin head - were required for the analysis, which could be removed from larger objects with the help of a fine drill or a diamond bur without any significant impairment. The elements copper, tin, zinc, lead, iron, silver, nickel, cobalt and antimony were determined quantitatively. Analyzes were carried out on the following 40 pieces (table): 1-14 handles 15, 16 solid fittings 1719 sheet metal fittings 2023 nails 24 thin chain 25, 26 forks 27 buckle strigilis and similar metal sheets various fragments 29 of these compositions listed in Table 1 fit into the Scheme of the alloys standardized according to DIN 1708, 1709, 17660, 17662, 1705 and 1716, 7 analyzes are close to standardized alloys, and only 4 analyzes cannot be classified in this system. The finding that 90% of the compositions found correspond to industrial standards of our time, refutes the widespread view that an allegedly antique metal object can be recognized as a forgery if it has a composition that corresponds to a DIN standard. The industrial standards do not define specific compositions, but name all technically usable alloys, which can fluctuate within wide limits, with standard designations. In the multicomponent tin bronze MSnBz4Pb standardized according to DIN, the tin content can vary from 3-5%, zinc from 3-5%, lead from 3-5% in addition to the copper. This is followed by the group MSnBzö with 5-7% tin and 5-7% zinc, so that in this area all possible compositions can be classified in one group. The 36 objects that correspond to the DIN standards in their main components can be classified into twelve different alloy groups (Table 2). This compilation shows that in Roman times all alloys of copper that could be produced were used with tin, lead and zinc. The publication of metal analyzes therefore does not - as is often feared - give the forger any information about which material to use, since all possible compositions are already in the


4 86 Table alloy group% copper% tin copper tin bronze tin bronze tin bronze SnBz 2 Sn Bz 6 SnBz J-9L, GSnBz 10 GSnPbBz 10 GSnPbBz 15 GSnPbBz 20 multi-component tin bronze MSnBz 4 Pb brass Ms 72 Ms 80 brass Ms. 85 brass i, zinc % Lead analysis number I ([39]) I, G IO (28) antiquity were used. The metal analysis can therefore only rarely contribute to the decision of the authenticity of Roman objects. Four of the objects examined, a handle (i), two nails (21, 23) and a strigilis (31) are made of pure copper, three objects, namely two further nails (20, 22) and a decorative plate (34) are made of copper with low tin content. All seven objects are made by forging, for which pure copper is particularly suitable, as it is initially soft when processed and increases in hardness considerably through cold deformation, so that nails could be forged without deforming during later use. The composition of the handle (1) gives an impression of how pure Roman copper could be. Only iron (0.24%), silver (0.06%) and antimony (0.05%) are contained in significant traces as unavoidable impurities from the ore. The low tin content of a nail and a strigilis, as well as the tin content of 1-2% in objects 20, 22 and 34, indicate that scrap metal was also melted here. As the tin content increases, the tin-containing copper turns into tin bronze, the tin content of which fluctuates between 6.3-9%. The lack of tin contents between 2 and 6% indicates that the lower tin contents in the copper are unwanted additions that can be traced back to the use of scrap metal, while tin contents above 6% are intended as additions to improve the material properties. Six objects, namely the handles 2, 3, 6, 8, 9 and 11 consist of a pure tin bronze, for the production of which pure copper was fused with pure tin. The low lead and zinc contents show that no scrap metal was used here.

5 87 With two other handles (5, 13), one sheet metal (29) and two strigiles (30, 33), the lead content of 0.9-1.5% slightly exceeds the limit permissible for the norm of tin bronze. What is the advantage of adding 6-9% tin to the copper, if this alloy was used to make most of the handles, especially the nicer ones? All handles in this group were made from a metal ingot by forging and, with the exception of handle 2, which was formed at elevated temperature, were still machined on the lathe to make the final discs and buttons. Pure copper, even if it is hardened by cold working, cannot be turned on the lathe because the metal tends to smear. Therefore the Roman handle maker had to add at least 6% tin to the copper in order to be able to manufacture the artistically turned handle ends on the lathe. The bell 36 and the finger 39 were safely cast, which are close to tin bronzes due to their tin content of 8.20% and 4.20% respectively, but differ therefrom by increased lead contents of 2.0 and 2.6%, respectively had to be added to improve the casting properties. It is difficult to imagine that a caster tried to cast a life-size figure, to which the finger 39 belonged, from a very high-melting tin bronze which is unsuitable for casting. If the tin content increases further, one comes to the area of ​​the cast bronze with more than 10% tin, which of the objects analyzed here only includes the approximately 6 cm long small shield (37), which was certainly no different from casting. In Rome, alloys in which the tin was partially or completely replaced by lead were particularly popular for metal castings, because lead was cheaper than tin and, moreover, the melting point of these lead bronzes was considerably lower than that of pure tin bronzes. The handles 10, 12 and 14 belong to this group, which contain, in addition to 7.5% tin, 8.3, 17.5 and 18.5% lead. Their strength must have been far below that of the thinner forged handles. The sheet 28, apparently a mass product, is made of this easily processable alloy by forging. In addition to copper, the fragment 35 also contains 34% lead. Such alloys can hardly be used in practice because the cast becomes brittle and the heavy lead collects in the deeper parts of the mold after the cast before it cools. Low-tin and high-lead is the metal of chain 24 with 2.40% tin and 9.60% lead, which may have been made from a piece of scrap. The multi-substance bronze of which the fork 26 consists, which contains 3.10% tin, 7.00% lead and 7.50% zinc, leads over to the types of brass. The handle 4 is made of pure brass with an unusually high zinc content of 25%, which otherwise only occurs in the first sesterces at the time of Augustus. Either this handle comes from the time of Augustus, when one was still able to zinc-rich Manufacture of brass, or in later times, zinc-rich brass was still made for decorative objects, while sesterces were already made by melting down

6 88 scrap metal, or only coins of Augustus were used to make this handle. The handle was forged and machined on the lathe, for which this alloy was just as suitable as the tin bronze, which was, however, much cheaper. Four other pieces, the sheet metal fittings 17 and 19, sheet metal 32 and the fragment 38 are made of zinc-rich brass with 1923% zinc without significant additions of lead and tin. The zinc content of the massive fitting 15, the sheet metal fitting 18, the decorative buckle 27 and the ring 40 is slightly lower, at 14-17%. In all cases, these are decorative pieces that may have been made from sesterces or dupondia as they are lead and tin only in small proportions. In this case, the sesterces used would have to come from the time before Trajan, since the zinc content had already dropped significantly in Hadrian. All pieces of this group, with the exception of the broken piece 38 and the ring 40, were made by forging. Brass of this composition is well suited for forging, hammering and other deformation techniques. Three pieces of alloys rich in brass, which can be referred to as zinc bronzes, cannot be classified in the system of DIN standards, since they are not used in today's metalworking practice. These are the handle 7, the fitting 16 and the fork 25. The handle 7 consists of an alloy which, in addition to the main component copper, contains 2.40% tin, 8.20% lead and 13.40% zinc. Such an alloy is likely the result of the melting of brass and bronze. The fitting 16 is composed similarly, but contains less lead (1.90% tin, 2.20% lead and 17.00% zinc), which makes it similar to sesterces of Hadrian. Here too, scrap metal was probably added to the brass. The fork 25 is cast from a mixture of copper with 4.20% tin, 9.90% lead and 9.50% zinc. Fortunately, the fork broke off the handle due to this unusable alloy. The high lead content of the fork would certainly have led to lead poisoning of the owner. The analyzes of these everyday objects show that the Roman metalworkers knew and used the properties of the various copper alloys. They forged the coarse handle 1 from copper, for delicate handles that they still worked on the lathe they chose the tin bronze, and they cast massive handles from cast bronze rich in tin. They hammered sturdy nails out of copper rods, but just as they were able to hammer and shape the copper into thin sheets. Brass was also used for casting and, with the same skill, for forging and driving work. The appropriate use of the various alloys and the mastery of the most diverse metalworking techniques prove the high level of metalworking in Roman times.