Revising Crawford’s Roman Republican Coinage from RRDP: The case of RRC 367

By The American Numismatic Society

by Alice Sharpless

This blog post is a preliminary version of an article to be submitted for peer review. Comments and additional relevant specimens are welcomed by the author.

Coinage of the Roman Republic Online (CRRO) is a database that was created by the ANS to be, in effect, an online version of Michael Crawford’s 1974  Roman Republican Coinage (RRC) which remains the primary typology for Roman Republican coinage. Since the publication of RRC there have been significant revisions to our knowledge of Republican coinage provided by new hoards and ongoing scholarship (for an updated summary of these new discoveries see Yarrow 2021). This year, the Roman Republican Die Project (RRDP) began to publish material based on Richard Schaefer’s archive of Roman Republican Dies. Schaefer’s materials provide numerous specimens from auction catalogues and museum collections which add significantly to our knowledge of particular Republican issues. In addition to the die studies carried out by Schaefer, these specimens, which are accessible through the database SITNAM and linked to CRRO, provide, in some cases, revisions to Crawford’s typologies.

One particularly significant issue for which RRDP has allowed revisions to Crawfords typologies is RRC 367, a joint issue of L. Cornelius Sulla and. L. Manlius Torquatus from 83–82 BCE. This was one of the issues produced for Sulla’s army on their return from the East. For more on the significance of this and the other Sullan issues see Lucia Carbone’s two-part blog post on the financing of Sulla’s reconquest of Italy.

Crawford divided RRC 367 into five types—two aurei and three denarii (see below). The primary distinction of all the types is the obverse legend, its spelling and placement of the letters in relationship to the head of Roma. The first, 367/1 (denarius), only has L·M behind the head running downward, and ANLI PROQ (no T) running upward before the head. It is also the only type for which Crawford said the reverse legend may read L·SVLLA·IMPE or L·SVLLA·IMP (not L·SVLLA·IM). For the other four types, Crawford allowed L·SVLLA·IMP or L·SVLLA·IM as possibilities but excluded L·SVLLA·IMPE. The primary distinction of these four types is whether or not the obverse legend has a T (turned on its side) and the denomination. RRC 367/2 (aureus) and 367/3 (denarius) include the T in the obverse legend, while 367/4 (aureus) and 367/5 (denarius) do not. The name always appears behind the head running upward and the title before the head running down words.

Although Crawford did not distinguish the reverses of 367/2-5, Schaefer’s die study reveals that there is, in fact, a difference between the reverse dies of these types. The reverse design of all types shows a triumphator, crowned by flying Victory, in a quadriga moving right. But Schaefer’s materials show that on 367/2 (aureus) and 367/3 (denarius) the horses of the quadriga are arranged so that the lead horse is closest to the viewer (Fig. 1) while on 367/4 (aureus) and 367/5 (denarius) the lead horse is furthest from the viewer (Fig. 2).

Figure 1. RRC 367/3. British Museum 1860,0328.77 (Donated by Count John Francis William de Salis).
Figure 2. RRC 367/5. ANS 1941.131.172.

This arrangement of the horses does not vary within a particular type except in two ways. There are a very limited number of denarii that are hybrids that have the obverse legend of 367/3 but the reverse of 367/5 (lead horse furthest). Schaefer found seven examples of this hybrid, two of which are die linked to 367/5 through reverse die 367/5 EW (Figs. 3–4).

Figure 3. RRC 367/5; F. Panvini Rosati, La moneta di Roma repubblicana. Storia e civiltà di un popolo: catologo (Bologna: Museo Civico Archeologico, 1966), no. 273.
Figure 4. Hybrid of RRC 367/3 and 367/5 with reverse die link to Bologna specimen of RRC 367/5 (Figure3); Ernst Justus Haeberlin Collection = Adolph E. Cahn and Adolph Hess Auction, 17 Jul 1933, lot 1628.

There is also a denarius hybrid of 367/3 and 367/5, with the obverse legend of 367/5 (no T) and the reverse quadriga arrangement of 367/3 (lead horse closest) (Fig. 5).

Figure 5. RRC 367 Control mark type with XXV before. Classical Numismatic Group, eAuction 272, 25 Jan 2012, lot 323.

This variation has a further highly significant difference. The reverse dies bear control numerals either before or behind the quadriga (Fig. 6).

Figure 6. RRC 367 Control mark type with X behind. Phil Davis Collection (image by Andrew McCabe).

Based on the reverse, Schaefer included most of these control-marked coins with RRC 367/3 despite the difference in the obverse legend. While Schaefer did recognize the control marks on these “hybrid” coins in most cases, there are some specimens where the controls are off flan or otherwise not well preserved. Based on Schaefer’s materials, the author has carried out a die study that shows that all of these “hybrid” types have control-marked reverse dies. Schaefer’s materials include 26 examples of this control-marked type, while my die study includes an additional 12 specimens. The control marks on these coins are roughly added and therefore not well preserved so it is not possible to identify the numeral in all cases, but the die study shows that all of the coins with the 367/5 obverse/367/3 reverse typology have reverse control marks. There is only one specimen where there is no trace of the control mark preserved, but the specimen has an obverse die link with a control-marked coin (Figs. 7–9).

Figure 7. RRC 367 Control mark type. No control mark is preserved on this specimen but it is obverse die-linked with a control-marked coin. Gorny & Mosch GmbH, Auction 196, 7–9 Mar 2011, lot 2432.
Figure 8. RRC 367 Control mark type. Only a portion of the control mark is preserved (IX, behind); the control can be seen on other specimens (Figure 9). Heritage Numismatic Auctions, World & Ancient Coins Select Auction 232149, lot 62217.
Figure 9. RRC 367 Control mark type. Control IX, behind; reverse die-linked with Figure7. Noble Numismatics Pty Ltd., Auction 70, 9 Jul 2002, lot 3427.

Crawford identified only four control marks on this issue. He associated two of them with RRC 367/3 (control VI [Phillipe 311=Gorny 228, 375] and IX [Berlin]) and two with RRC 367/5 (control XV and XX [Vatican 2346]) (Figs. 9–10).

Figure 10. Reverse control marks observed by Crawford. Right to left: IV, before, Gorny & Mosch GmbH, Auction 228, 9 Mar 2015, lot 375 (note that this die actually has XXX, before, as well). XV, before, British Museum 2002,0102.3155; XX, before, British Museum 2002,0102.3159; for IX, behind, see Fig. 8.

Because the control numerals are small and not carved into the die in the same way as the legend, Crawford suggested they were not meant to be present on the final coin and were instead intended to be removed when the die went into use. He therefore suggested that the whole issue may have had these control marks and only a few have survived that were not removed before use. The RRDP data shows, however, that the control marks are associated specifically with these 367/5 obverse and 367/3 reverse coins. It is therefore unlikely that they were used throughout the whole issue and instead appear to distinguish a particular type within the issue. Why it would be marked out in this way remains a mystery for now.

The most up-to-date estimates for the issue are shown in Tables 1–2. The approximate issue size estimates differ slightly based on the obverse and reverse dies counts, but we can estimate a range between 165-660,000 (assuming 20,000 coins struck on average per obverse die and 15,000 per reverse die, a much-debated topic!). This is comparable to the approximate size for 367/1 (40–270,000) and 367/3 (730,000–1.5 million), while 367/5 was significantly larger (5.94–9.41 million) (see Carbone’s blog post on Sulla Part I, Table 4). These new numbers provide an approximate issue size for denarii of 6.88–11.84 million.

Obverse DiesCoinsd1Die estimatedPlus 95Minus 95CoverageApprox. issue size
1538523331687%320–660,000
Table 1. Production estimates based on obverse die counts (Esty 2011)
Reverse DiesCoinsd1Die estimatedPlus 95Minus 95CoverageApprox. issue size
1238215201195%165–300,000
Table 2. Production estimates based on reverse die counts (Esty 2011)

The presence of this control-marked subtype within Sulla’s production allows it to draw interesting parallels to the contemporary production of the mint of Rome (see Carbone’s blog post on Sulla Part I).

Currently, there is no process to update CRRO to include revisions to Crawford. We plan, however, to explore ways that will allow us to add updates to Crawford’s typologies, dating, etc., based on findings from RRDP as well as other studies. We envision that CRRO will remain a searchable index of Crawford’s typologies using linked open data to connect specimens in collections worldwide (thanks to Nomisma), but will in the future provide access to subsequent and ongoing research on Roman Republican coinage.

Revised RRC 367 Typologies

Below are revised typologies for RRC 367 differentiating between reverse as well as obverse dies and including the control-marked type.

Denarius; Unknown; 82 B.C.; Silver; 0.004 kg (0.0088 lb.); 77.NE.64.5; No Copyright – United States (http://rightsstatements.org/vocab/NoC-US/1.0/)

 obverse: L M downwards, behind, ANLI  PRO Q upwards, before

 reverse: L SVLLA IMPE or IMP; lead horse closest

obverse: L MANLI T upwards before, PRO downwards, behind (T sideways)

reverse: L SVLLA IMP or IM; lead horse closest

obverse: L MANLI T upwards before, PRO downwards, behind (T sideways)

reverse: L SVLLA IMP or IM; lead horse closest

Figure 13. RRC 367/4. ANS 1967.153.195.

obverse: L MANLI upwards before, PRO downwards, behind

reverse: L SVLLA IMP or IM; lead horse furthest

obverse: L MANLI upwards before, PRO downwards, behind

reverse: L SVLLA IMP or IM; lead horse furthest

  • Denarius [new] (Figs. 5–10)

obverse: L MANLI upwards before, PRO downwards, behind (cf. RRC 367/5)

reverse: L SVLLA IMP or IM; lead horse closest, control number before or behind (cf. RRC 367/3)