Civil war, numismatic slight and confusion at the mint (Constantine/Licinius)

[Heliodromus]

The first civil war between Constantine I and Licinius I consisted of two pitched battles, the Battle of Cibilae in 316 AD (near Siscia), and the Battle of Mardia in 317 AD (in-between Heraclea and Serdica).

Having lost the first battle, Licinius decides to do two things.

1) Appoint Valerius Valens (his dux limitis) as co-emperor (causing vexation to future coin collectors), raising a new/expanded army

2) Change some of his coin legends to insult Constantine!

It's not quite clear which of these was his top priority, since Valens only appears on his coins after the numismatic slight has been delivered! 🙂

The form of numismatic insult was to change the obverse legend on his coins, where applicable, from IOVI CONSERVATORI AVGG (plural) to IOVI CONSERVATORI AVG (singular). One wonders if he sent samples to Constantine with a suitable note, to make sure the message was received! Perhaps he was just sincerely hoping for Jupiter to stop protecting Constantine?

There were three of Licinius' mints that made this same legend change: Thessalonica, Heraclea (both close to the Battle of Mardia location, where he had retreated to), and (somewhat oddly) distant Alexandria.

The purpose of the singular "AVG" legend was of course to exclude Constantine, but the reason I'm writing this post is the coin below, a new acquisition, where the mint messed up and paired this AVG reverse with an obverse of Constantine.

There are 2 Alexandrian issue marks, but it seems at least 4 emissions, dating to this war period. The two issue marks are K-wreath/A/X, and K-wreath/X/A, where A is the officina. It's not obvious what the common K-X component is referring to. It might seem that these are just two different executions (A/X vs X/A) of the same issue mark, but coins of Valens only occur in the 2nd X/A mark, as do coins of the new caesars who appear after post-war reconciliation, so it seems clear these are distinct issue marks occurring in that order.

The 4 emissions, in order, appear to be:

1) A/X AVGG L1 + C1
2) A/X AVG  L1 (+ C1 mule)
3) X/A AVG  L1 + Valens (+ C1 mule) <-- my coin
4a) X/A AVGG L1 + Valens
4b) X/A AVGG L1 + C1, plus X/A CAESS Crispus + Constantine II + Licinus II

The first reflection of war is emission 2), where Licinius changes AVGG to AVG, but Valens has not yet appeared. I have also seen two Constantine mules (with AVG) in this emission.

The second stage is emission 3) where Valens now appears for the first time, unlisted with this AVG reverse, and where my Constantine mule belongs.

The final stage, reverting to the plural AVGG, is emission 4), including both Valens and the post-war newly appointed caesars (with reverse CAESS). It's not clear whether this was really two emissions (4a, then 4b), with AVGG initially meaning Licinius+Valens, or perhaps this was post-war with Valens and Constantine included in same emission (which would have been a short-lived state of affairs since Valens was quickly executed).

The inclusion of the new caesars in this final emission is useful since they were appointed in march 317 AD, and this helps date the civil war to 316-317, when earlier it had been suggested as 314 AD.

Here's my Licinius II (which I've shown before), RIC 21, from this final emission, interesting because it includes his full name (only briefly used) of VAL[ERIVS] CONSTANTINVS(!) LICINIVS.

Please post your coins from these issues, or anything related!

 

Edited November 14, 2024 by Heliodromus

[rasiel]

That's an amazing Licinius II you have there ❤️

Only thing I can contribute in this area is my own much less interesting Licinius II (which at least happens to be from Alexandria). I can also contribute something a little more special from the second civil war.

Rasiel

 

[seth77]

So Licinius wanted to exclude Constantine from Jupiter's protection but someone at Alexandria didn't quite catch that so instead he was like heck let Jupiter protect ONLY Constantine. Bamboozled.

[Orange Julius]

Great post and a very nice new coin, very interesting. I’d known that my coins below were part of the Valens issue but the legend variation will be something I’ll look out for as my coins do not show it.

Below are three coins from the issue you reference, all with AVGG reverse legends, one of Constantine and two of Licinius. 

Additionally, following (for fun) are three more coins of Licinius, from Alexandria, from a year or so earlier, 314 and 315. 

Edited November 14, 2024 by Orange Julius

[Heliodromus]

6 hours ago, Orange Julius said:

Additionally, following (for fun) are three more coins of Licinius, from Alexandria, from a year or so earlier, 314 and 315. 

I always find RIC 7, with LICINNIVS legend, interesting, since this wasn't how Licinius spelt his own name - it's a misspelling used by Maximinus II, which Licinius inadvertently failed to immediately correct when he assumed control of the mint after Maximinus' death.

Preceding RIC 7 there's also Licinius' first Alexandrian issue, again showing transitional confusion/indecision, where he continues Maximinus' GENIO AVGVSTI with head of Serapis type, but switches the legend to GENIO POPVLI ROMANI, again with the LICINNIVS misspelling.

Edited November 14, 2024 by Heliodromus

[mcwyler]

I got this Licinius II because I thought the obverse looked like nothing so much as a kid dressed up as a soldier. Which, I suppose, is what he was really. Didn't stop him suffering the fate of all Constantine's enemies.

[Heliodromus]

20 minutes ago, mcwyler said:

I got this Licinius II because I thought the obverse looked like nothing so much as a kid dressed up as a soldier.

Yes, I think he was born c.315 AD (not sure the source of that), so at the beginning of that type in 321 AD he was only 6 years old, and only 2 years old when he was appointed as caesar in 317 AD, and first started appearing on coins!

 

[Victor_Clark]

The birth of Licinius II, according to information given in Zosimus 2.20.2, is figured to be c. August 315

[galeriusmaximinus]

Very nice Constantine I of Alexandria mint and extremely rare. There seem to be Constantine I folles for several of the Licinius issues with “AVG”. Not sure if they are all to be considered mules.

Here are my Constantine I folles of Alexandria mint with “AVG” and with “AVGG”.

[galeriusmaximinus]

There are folles of Licinius II of several mints which refer to the longer name “Valerius Constantinus Licinius”. This may have been be the initial agreement for the name placing Licinius II also in the Constantinian family (by his mother). This was then changed / omitted in later issues in the east.

Here are some of my examples with the interesting name variation from Alexandria, Heraclea, and Nicomedia mint.

[Heliodromus]

5 hours ago, galeriusmaximinus said:

There are folles of Licinius II of several mints which refer to the longer name “Valerius Constantinus Licinius”.

Great coins - especially that Heraclea campgate. It's interesting that your coin and RIC 24 both have the larger bust as well as CONST legend, while the regular/common coins have small (tiny) bust paired with the LICIN LIICINIVS legend.

[galeriusmaximinus]

Coming back to the original discussion of the fairly rare "AVG" reverse legend endings. Here are my 3 types for Licinius I from Alexandria mint with the “AVG” reverse legend ending.

The evidence raises some questions. If Licinius decided to stop referring to multiple emperors at some time, why do we have in all three issues both variations of the reverse legend ending - “AVGG” (common) and “AVG” (very rare)?  Why doesn’t he stop minting for Constantine in these issues, and why do we also have Constantine reflected with both variations - “AVGG” (common) and “AVG” (very rare)?

If we consider the issues to be dated correctly by RIC as consecutive issues this does not make much sense. We would expect (as Heliodromus suggested) for Licinius to consciously drop the reference for multiple Augusti at some point in time and only continue with “AVG”. But he does not do so but has both legend variations in all issues.

If we would consider that he does stop the multiple ruler reference due to civil war with Constantine, why does he continue to mint for Constantine? Why should he use a Constantine obverse and pair it (theory of mules suggested by Heliodromus) with a rare Licinius reverse?  

Should we rather consider a parallel issuance of these three issues with an initial phase for both rulers with the “AVGG” legend variation, then Licinius minting only the “AVG” variation for a very small extremely rare issue for himself - but still creating equally rare mules for Constantine?

All this seems rather unlikely.

Maybe we are facing in Alexandria just a very rare “mistake” of the die cutter running out of space and not finishing the reverse legend, i.e. omitting the second “G” for “AVGG”. This mistake being seemingly considered as not so important and the die was still used – or it was not recognized at the time. The rarity and appearance in all three issues could thus be explained.

We do have a different practice in Heraclea mint. There are clearly two separate (and common) issues which are differentiated by the “AVGG” (RIC VII 11, 12 for Constantine and Licinius) and the “AVG” (RIC VII 13 for Licinius, only). I am not aware of any "mules" and the size of the two issues for Licinius is far more balanced in Heraclea mint.

[Heliodromus]

1 hour ago, galeriusmaximinus said:

The evidence raises some questions. If Licinius decided to stop referring to multiple emperors at some time, why do we have in all three issues both variations of the reverse legend ending - “AVGG” (common) and “AVG” (very rare)?  Why doesn’t he stop minting for Constantine in these issues, and why do we also have Constantine reflected with both variations - “AVGG” (common) and “AVG” (very rare)?

I'm happy to have this debate, but am about to go away for the weekend, so may have to wait until after that to reply in full.

Just as a minor correction, the RIC 11 description is an error, and should be Licinius not Constantine, matching your nice coin (incl. officina). RIC's source was Dattari from RIN 1906, p.44, where he lists the coin as Licinus. He also illustrated it (plate I # 18), but the only copy of RIN 1906 I've been able to find online so far is missing that plate! It'd be interesting to see if Dattari's coin is the from the same rev die as yours.

Edit: I see that on Not In RIC Lech has your coin, but also another from different dies (but also off. H).

https://www.forumancientcoins.com/notInRic/7ale-11.html

Edited November 15, 2024 by Heliodromus

[galeriusmaximinus]

Great research – you are correct that the Constantine referred to in RIC seems to be rather a Licinius I.

We then have at least 3 examples of Constantine with the “AVG” legend, 3 examples of Valens with the “AVG” legend, and at least 10 examples of Licinius with the “AVG” legend. The “AVGG” legend seems to be known for all 3 rulers and much more common (apart for Valens).

I still do not believe that the 3 Constantines should be mules. Then the 3 Valens would also need to be mules. Having 6 mules found to the also very rare Licinius issues (for which we do not know so many more examples) seems rather odd.

 

[Heliodromus]

There were certainly a number of legend errors being made at Alexandria at this time, perhaps reflecting Latin-illiterate engravers, so mint errors of some sort seems a reasonable part of any explanation of the AVG reverses, but the question is exactly what error(s) !

For example, here's a couple of misspelled versions of Licinius from Not In RIC - LICIINVS & LICINNIS.

The same Dattari article from RIN 1906 that includes RIC 11 (which he regards as intentional, but RIC does not), also lists a couple of misspellings of Constantine's name - CONSTANSTINVS & COSTANTINVS.

RIC 11 (AVG reverse in wreath/A/N issue) is problematic since - as you note - it implies a switching back and forth between AVGG and AVG reverses (or deliberate parallel production) which seems highly unlikely. If we were to disegard RIC 11 as intentional then the problem goes away since then we'd have a single uninterrupted period of AVG usage straddling the A/X and X/A issues:

A/X AVGG
A/X AVG
X/A AVG  (incl. Valens)
X/A AVGG (incl. Valens)

I'm only aware of two or three specimens of RIC 11 - the two on Not In RIC (incl. yours), and the one Dattari had seen reported by RIC. All are from officina "H".

There are a few considerations which may support the interpretation of RIC 11 as unintentional.

1) In an issue utilizing 8 officinas (A-H), RIC 11 appears to only be known from 1 officina, although there are (at least) two different rev dies, so perhaps an engraver profilerating his own mistake. If this was intentional, then why are all 8 officinas's not being used ?

2) If RIC's dating of this issue, 315 AD, is correct, then it is pre-war.

3) Both known rev dies for RIC 11 have the CONSE-RVATORI legend break, uncommonly used with the AVGG legend (CONSER-VATORI typical) since the tail half of the legend becomes excessively cramped (see examples below). Perhaps the engraver ran out of room, and failed to understand the importance of AVGG vs AVG ?

Switching topic to the two K-X (A/X, X/A) issues, and whether the AVG there was intentional or not, I think the default assumption would have to be that it was, given that the presence of Valens tells us the date is correct and the mint had received fresh instructions, and given that we see AVG also used at this time at both Heraclea and Thessalonica (my two specimens below).

The theory that Alexandrian engravers continued to *repeatedly* shoot themselves in the foot and run out of space for an intended AVGG vs AVG legend (i.e. AVG was unintentional) is presumably based on the correlation of CONSER-VATORI/CONSE-RVATORI breaks with AVGG/AVG respectively, but I would argue that legend engravers were proficient at planning space utilization and these legend breaks were not haphazard execution, that they belatedly had to react to, but rather pre-planned based on the number of letters in the legend they were engraving.

IOVI CONSER-VATORI AVGG 10-10
IOVI CONSE-RVATORI AVGG 9-11
IOVI CONSER-VATORI AVG 10-9
IOVI CONSE-RVATORI AVG 9-10

Clearly for AVGG the CONSER-VATORI break, with it's 10-10 split, is preferable, and is what the vast majority of coins in these issues have. On the rare occasion of an engraver sloppily planning letter spacing, the normal thing to do was to cram in the tail of the legend however necessary, as can be seen on my Constantine below (really an "unforced error" - look at all the space for an "R" between the "E" and victory!).

For the AVG legend there doesn't seem much to choose between a 10-9 or 9-10 letter split, although 9-10 not encroaching on victory seems more visually pleasing. In any case, the engravers had no problem fitting 10 characters after the break for the usual AVGG 10-10 split, so there's no reason to suppose that using a CONSE-RVATORI break and 9-10 split would have routinely induced them to truncate the legend.

Below is a RIC 16 AVG specimen of mine with an unusual CONSER-VATORI break that provides additonal support for the AVG legend being intentional.  The engraver here goes with the CONSER- break, perhaps out of force of habit (AVGG), but then has an abundance of space for the second half of the legend, and still ends with AVG, not AVGG.

On the other hand, one oddity of the AVG legends is that all specimens that I've seen (for any of Licinius, Valens or Constantine) are from officina "B". Interestingly the mint reduced from 8 to 2(!!) officina's for these K-X issues, but we'd expect "A" and "B" to be in sync. Certainly there exist coins for Valens with the AVGG legend from both officina's "A" and "B". Perhaps for much of the war Alexandria was only using one officina (B), and then resumed with two later?? I suppose we could alternatively choose to interpret this as evidence of AVG being an unintentional officina-B only error, but with the arguments against, and number of dies involved, this seems to me unlikely.

Given that there was (here, at Alexandria) no officina allocation between Licinius, Constantine and Valens, either before or after the war, the conditions would have been ripe for mules.  While there are multiple muled die pairings, all it took for this to occur was a single failure - for the Constantine obv dies not to have been removed from the operational area of the mint where a clueless mint worker might pick them for use.

To be honest, neither explanation - mules, or unintentionally truncated AVG legends - seems totally satisfactory, although I think both seem better than the alternative of AVG legends being deliberate and deliberately used for Constantine. It seems to me that mules, requiring only sloppy mint organization, is a more likely explanation than a series of error dies where the engravers, happily able to put 10 letters after the break the majority of the time (even squeezing in 11 if they had to), multiple times ran out of space and chose to truncate rather than cram the legend in.

 

Edited November 23, 2024 by Heliodromus

[Heliodromus]

1 hour ago, galeriusmaximinus said:

I still do not believe that the 3 Constantines should be mules. Then the 3 Valens would also need to be mules.

We see the AVG legend first in the A/X issue where Valens is not included, then continued into the X/A one. Logically when Valens was added they should have reverted back to AVGG immediately (as they eventually did after the war, in this same issue mark), and removed/destroyed the AVG dies, but I don't find it too surprising that they didn't, especially in light of the other mint failings we see and are assuming in either case.

It'd not be very surprising if the mint order was just "add Valens" rather than a more thoroughly thought out "add Valens, and revert to AVGG since the previous order to use AVG no longer makes sense".

Edited November 22, 2024 by Heliodromus

[galeriusmaximinus]

Here are the Heraclea mint issues RIC 12 and RIC 13 which reflect the shift from AVGG to AVG in 5 officinae.

There are two variations of the diadem, 2 bands vs 3 bands. The 3-band variation appears on around 60% of the pieces vs around 40% with the 2-band variation. Noting that Heraclea mint used predominantly 2-band diadems until the occupation of Maximinus II and then shifted more to using 3-band diadems.

For the different officinae, Constantine was allocated in RIC 11 the use of B and E. However, Licinius used all 5 officinae, i.e. also B and E likely in parallel to Constantine. The officinae B and E use predominantly the 3-band variation while officinae A, Gamma and Delta predominantly still use the 2-band variation.

[Heliodromus]

Yes, interesting in of itself, although I don't think it is of relevance to Alexandria. Officina usage (e.g. allocation or not) seems to have been highly mint specific, as was prevalence of errors. Some mints were less error prone than others, and this also depended on period of operation.

Victor Failmezger had this useful graphic in his book that summarizes officina usage across mints for the 313-337 time period.

It's interesting to note how the number of officinas tended to be somewhat consistent across mints within each Diocese (a reflection of localized mint management perhaps), but Alexandria is quite an outlier. The other oriental mints used a large number of officinas, but under Licinius Alexandria quickly reduced to 2, seemingly due to Antioch producing the bulk of the coinage.

[Victor_Clark]

1 hour ago, Heliodromus said:

Victor Failmezger had this useful graphic in his book that summarizes officina usage across mints for the 313-337 time period

 

You have the wrong Victor...it was made by Victor Clark.

https://constantinethegreatcoins.com/activity/

[Heliodromus]

Oops - sorry Victor!

It's a great graphic - very compact presentation of a lot of information.

I hope you don't mind me posting it? Since I wrongly thought it was Failmezger, out of print, I figured it was ok!

 

[Victor_Clark]

1 hour ago, Heliodromus said:

I hope you don't mind me posting it? 

 

Of course, feel free. It is from RIC VII. I just colorized the black and white chart to make it easier to see what was happening.

 

numbers represent how many officina were open
green = mint was open
black = mint was closed
orange= mint was only open part of the year
grey= uncertain if mint was open or closed
solid blue line- roughly the time Licinius lost the mint