PREFERENCE OF THE ANCIENT GERMANS FOR OLD MONEY AND THE SERRATION OF ROMAN COINS By B. L. ULLMAN In my review of Gudeman's edition of the Germania of Tacitus (Classical Weekly, XV, 35) I discussed briefly the meaning of the sentence in ch. 5: Pecuniam probant veterem et diu notam, serratos bigatosque: "the Germans favor old and long familiar money, (such as) the mill-edged and chariot types. I am led to a fuller discussion by an interesting letter from Mr. T. O. Mabbott of New York. Two points are involved. The first led to my original remark: the reason why the Germans preferred old, well known money. I suggested the rather obvious reason, which I have never seen stated, that the Germans learned to beware of new coins because the Roman traders sometimes worked off spurious coins on them. Mr. Mabbott accepts this and quotes a very interesting parallel: "The Austrians for over 100 years minted thalers with the head of Maria Theresa for use in Abyssinia, where the natives would have no other coins-and insisted that the number of 'pearls' in certain accessory details of the design be always the same. The money of the Emperor Menelik II was circulated with difficulty at first, I believe, because it was new.' Mommsen (Roem. Muenzwesen, 771-772) thinks that the Germans preferred the early coins because they could more easily be distinguished from the degraded Neronian currency than could the coins of the early Empire. This may be true but it clearly was not the view of Tacitus. Norden,1 though citing the parallel of the Maria Theresa thalers, gives an explanation that does not explain. He says that the Germans preferred bigati and serrati because they were easily recognized. We may grant that this is true of the serrati but the bigati would be no more easily recognizable than any other type. 1 The other point involves the last two words in Tacitus' sen 1 Die Germanische Urgeschichte in Tacitus Germania, 1920, 280 ff. tence. As the chariot coins and the mill-edged coins antedate 50 B. C., Gudeman thinks that they could not possibly have circulated to any extent in Tacitus' time, 150 or more years after the last one was struck, and that Tacitus therefore was using an early source. As shown below such coins are very rare in postNeronian hoards found in Germany. It might also be added that if bigati were actually preferred by the Germans the only possible explanation would seem to be that the Romans began trading with the Germans when the bigati were the commonest types. I suggested that "perhaps Tacitus is using his own more vivid expression, serratos bigatosque," and I cited as an analogy the phrase vinum consulare, which is not to be taken literally. Mr. Mabbott is inclined to agree with Gudeman that Tacitus used an early source for this expression. I do not believe that one can settle this matter, but there are a few points of interest connected with it. Bigati were popular for over a century from about 216 B. C.2 Serrati were most common for about 30 years after 92 B. C., though a few isolated instances are found before that time. Most bigati were not serrated; nor do all serrated coins have chariots on them. It is clear then (as is indeed generally assumed) that Tacitus had in mind two different sets of coins, one set serrated and the other bigated. Mr. Mabbott believes that "the purpose of serrate edges was to prevent counterfeiting the indentations making a core of baser metal easily visible and hence the Germans, who undoubtedly were victimized at times by the sharp traders who carried a few counterfeits for their benefit, preferred them." We shall take up the matter of serration below, but granting that its purpose was to prevent counterfeiting it does not follow that the Germans, if they actually did prefer this particular type of coin, preferred it because it prevented counterfeiting rather than because it was old and familiar to them. There is in fact a strong argument against such a view. Disregarding the fact that Taci 2 Incidentally Gudeman has 126 through a misprint. Grueber, Coins of the Roman Republic in the British Museum, I, 66, says that the earliest bigati date from about 196 B. C. If this is true, Livy 23.15.15 (216 B. C.) is anachronistic. Babelon's view that the bigati date from about 217 B. C. strikes me as more reasonable (Description des Monnaies de la République Romaine, I, xxi). ... tus mentions these coins for their antiquity and not for their success in preventing counterfeiting, Mr. Mabbott's explanation does not account for the bigated coins. The whole phrase must be explained. Mr. Mabbott thinks that serratos bigatosque could not have been Tacitus' own phrase for old coins, because coins even older than the bigati were still to be seen on rare occasions and because relatively few Roman coins were ever serrate. On the first point I may say that the older the coin type the less likely that Tacitus would be acquainted with it. Furthermore he is giving an example of "old" coins (veterem) and not "oldest" coins. On the second point it should be observed that the only Roman serrate coins antedated Tacitus' time by 150 years or more and that Tacitus would therefore inevitably associate such coins with the Republic. It may further be observed that Tacitus would naturally use the bigati as an example of the older Republican coinage because of his familiarity with Livy, who used the phrase argentum bigatum as a synonym for argentum signatum, coined silver.3 I realize of course that some may hold that the similarity between Tacitus and Livy shows that the lost books of Livy are the early source claimed for the entire sentence under discussion. But consideration of another sentence in the Germania will show how unnecessary this assumption is. To quote from my review of Gudeman: In chapter 3 Tacitus says quae neque confirmare argumentis neque refellere in animo est. As is well known, this expression is found in Livy, praef. 6, ea nec adfirmare nec refellere in animo est, and again in 5.21.9, neque adfirmare neque refellere operae pretium est. Gudeman grants that Tacitus borrowed from Livy but argues that it is psychologically improbable that Tacitus should have remembered this particular phrase from his earlier reading of 'Livius ingens.' His explanation is that Livy may have used the phrase again in his description of Germany and that Tacitus found it there. But I cannot see the psychological improbability which troubles Livy's 3 So too Kubitschek in Pauly-Wissowa, Real-Encyclopaedie, III, 467. usage may be clearly made out by comparing 34.10.4, argenti infecti (uncoined silver) et signati bigatorum with 34.10.7, argenti bigati et bigatorum; and 34.46.2, argenti infecti 08censis, auri, with 26.47.7, argenti infecti signatique. In the first passage signati is explained by bigatorum, in the second bigatorum alone is used, in the third bigati alone is used to contrast with infecti, just as signati is used in the fourth passage. Gudeman. Most of us recall single striking phrases from books we read years ago. And besides is it not entirely likely that Tacitus often reread the famous preface of Livy? In the same way it is quite as probable that Tacitus recalled the word bigati from his general reading of Livy as that he found the phrase we are discussing in that portion of Livy, now lost, which dealt with Germany. A thorough examination of the hoards and single coins found north of the Alps might throw light on the matter. The facts thus far published are somewhat confusing and subject to varying interpretations. Norden asserts that serrati and bigati are common in hoards laid down before Nero but that they practically disappear in later hoards. From this he draws the conclusion that Tacitus must have used a pre-Neronian source, and that source must be Pliny. This is a possible but unnecessary assumption. He points out that in the Natural History Pliny mentions bigati and that he also uses the rare word serratus, though not of coins, which fact in itself makes the argument questionable. As we have seen, Livy mentions bigati even more frequently. But Norden's last argument in favor of Pliny as the source of Tacitus is really an argument against that source: he notes that Pliny uses denarios probare, which is similar to Tacitus' pecuniam probant, but he fails to note that the two authors use the verb in entirely different senses, for in Pliny it means "to test." But let us return to the facts on which Norden bases his theory. We not only have no exact figures, as Norden admits, but no statement of any kind as to bigati in German hoards. For the serrati he refers to an article by K. Regling and to a private communication by the same scholar. In the latter is mentioned a hoard found in Liebeshausen, Czecho-Slovakia, consisting of 200 republican denarii, most of which are serrati, though no exact figures are given. It is to be noted that this is a very early hoard, consisting entirely of republican coins. There is no evidence to support Norden's conjecture that it was laid down as late as 18 A. D. Pending further evidence it is just as safe to assume a much earlier date, nearer the time when serrati were common in many parts of the ancient world. 4 He quotes from a separate pamphlet but I had previously discovered the article in the Zeitschrift für Numismatik, 29 (1912), 189 ff. In Regling's article there is a summary of the larger finds in free Germany. His argument is inconsistent: on the one hand he states that the occurrence of serrati and bigati in the preNeronian hoards confirms Tacitus' remark, on the other he explains the large number of denarii coined by Antony and the scarcity of early imperial coins as due to the fact that the former, like the Neronian coins, were debased and consequently forced the better coins into the melting pot. If the latter is true there should be few serrati, as in fact there are, relatively. Another explanation may enter in: the coins of Antony are after all early coins and may explain Tacitus' remark. In the case of the preNeronian hoards Regling gives the following figures for serrati and coins of Antony respectively: Feins 1, 7; Onna 8, 35; Barenau 7, 23; Franzburg 3, 3; Denecamp 8, 0; Niederlangen 41, 0, etc. He does not give figures for other republican coins, but they must be fairly large, especially in the case of Niederlangen. For the post-Neronian hoards Regling gives these figures: Middels Osterloog 1, 0; Fröndenberg 0, 6; Niemegk 9, 19 (and 22 other republican coins). It will be seen that where the coins of Antony occur at all they are far more numerous than the serrati. Regling has included only the hoards in free Germany. For our purpose those in the Roman provinces of Germany are quite as pertinent. One described by Mommsen distinctly favors my view. It was found at Székely. It includes 16 republican denarii and 186 belonging to the time of the last triumvirate. The latter are early enough to justify the first part of Tacitus' statement, whereas the former are not numerous enough (even if all of them were either serrati or bigati) to account for the last part of Tacitus' sentence. It may also be pointed out that in all the post-Neronian hoards listed by Regling, coins dating from 81-96, the period just before Tacitus wrote the Germania, are relatively very scarce. The matter of serration is a perplexing one and needs investigation. The latest authorities agree that the purpose of serration was not to prevent counterfeiting, for bronze coins and even plated denarii were serrate." 5 This is probably a very early hoard. 6 Loc. cit. 7 Grueber, op. cit., I, 159; G. F. Hill, Historical Roman Coins, 1909, |