The Thessalonican Trachea of
Andronicus III Palaeologus


In 1979 Bendall and Donald published "The Later Palaeologan Coinage" (LPC), the first comprehensive survey of the coinage of the Palaeologan rulers from Andronicus II to Constantine XI, and this ultimately became the basis of the later sections of Sear's popular catalog "Byzantine Coins and their Values".

This article attempts to determine just which of the Thessalonican trachea assigned to Andronicus III in LPC and Sear are in fact issues of that emperor, and to explain why. The discussion uses all the evidence currently available, including some recent finds which are not likely to be published in the academic literature (see the Data pages for details of these finds).

Longuet's "Salonika" Hoard.

The starting point for defining the Thessalonican issues of Andronicus III is Longuet's "Salonika" hoard. Prior to that find, virtually all "Andronicus" types had been attributed to Andronicus II, although without much actual justification, and very few coins were assigned to Andronicus III. The situation changed markedly with the Salonika hoard. This hoard, found in the late 1930's, included about 70 recognisable coins, drawn mostly from 23 different Palaeologan types (plus some debased Frankish deniers of John II Orsini of Epirus). A few of the coins were worn examples of some relatively common issues previously attributed to Andronicus II, but most of them were either quite new, or known from only a few odd examples. Also, as a whole the new types were smaller than the various known "Andronicus II" types, and mostly without legends. It therefore seemed reasonable to assume that Longuet's hoard consisted mainly of Andronicus III types, together with some late Andronicus II issues.

Actually, Longuet initially thought that the hoard included a number of issues of John V, and dated it to c.1360, but this date was later revised downward by Bendall to c.1342, at the beginning of the Zealot revolt in Thessalonica. This is a reasonable assumption, but in fact the dating of this hoard is still rather uncertain - an equally plausible date would be 1345, during the Serbian invasion which left most of Macedonia under Serbian rule, with Byzantine control limited to Thessalonica and its immediate environs. Alternatively, it is possible that this hoard predates 1342 by several years, since, as we shall see, it does not include reasonably common types such as S.2525 which could well be issues of Andronicus III.

Incidentally, the fact that only a few worn examples of the coins previously attributed to Andronicus II appeared in the hoard was significant, since it confirmed the assignment of the bulk of the previously known "Andronicus" types to that emperor, rather than to Andronicus III.

(Full details of the Longuet hoard can be found in Bendall's review in ANSMN (American Numismatic Society Museum Notes) 29, 1984. Note that the type numbers in the ANSMN review differ from those used in LPC, which follow Longuet's original hoard report. In the current article I use Longuet's original type numbers).

On the basis of Longuet's hoard Bendall and Donald defined in LPC what essentially became the "official" set of the Thessalonican types of Andronicus III. To begin with this set consisted of most of the new Longuet types, plus a couple of other non-Longuet types (LPC 234:2 = S.2482, and LPC 238:8 = S.2525), and also a mounted emperor type from the Pella hoard (LPC 242:15 = S.2499). Later, S.2525 was reassigned to John V, while Longuet 10 (LPC 246:1 = S.2500), a type originally assigned to John V in LPC, was reassigned to Andronicus III. (Also, Longuet 1 and Longuet 21 were initially assigned to Andronicus II by Bendall and Donald - more of these later).

To this basic set Bendall later added (apparently on the basis of style, rather than any hard evidence) five more "additional" types (Num. Circ. 1980, p. 45-7, No's 9-13), and thus produced the set of Andronicus III types ultimately listed in Sear as S.2482-2501. Later again, Bendall, in PCPC ("A Private Collection of Palaeologan Coins", 1987), effectively reattributed S.2492 and 2493 to Andronicus II, and apparently dropped S.2494 and 2495 altogether. Finally, in DOC V Grierson reassigned S. 2359 (Longuet 1) from Andronicus II to Andronicus III, as well as S.2561 from Manuel II; also, he added Grierson 1473, which for some reason had been left out of both LPC and Sear, to the "official" set, together with the scarce type DO 1190 (this last type is actually Longuet 22, which was also omitted from LPC).

So, how secure are the Andronicus III attributions? After all, the whole process seems rather vague, with lots of guesswork and subjective judgment. Fortunately, however, we do have some harder facts to go on. To begin with, some of the new types in Longuet's hoard were overstruck on what are, according to Bendall, debased versions of the billon deniers of John II Orsini of Epirus, and since John is styled despot on these coins, but was not granted the title of despot until 1328-29, it follows (we hope) that these overstruck types at least must indeed be issues of Andronicus III.

Thus we have our first reasonably secure assignments to Andronicus III, namely S.2483, S.2486 & S.2487. Assuming that the Orsini deniers were acquired in the wars of 1330-32 with the Bulgarians (these types are often found in Bulgaria - see Note "Bendall's 2000 Hoard"), we can reasonably date these three types to the early to middle 1330's (but more about 2483 and 2486 below). In fact the latter date is probably more likely, since it's also quite possible that the deniers were acquired during Andronicus's recovery of Neopatras from the Catalans in 1333, particularly given that the debased deniers would presumably have dated from the later period of John's reign,

An example of S.2486 on a debased Orsini denier is shown below - part of the legend "De Arta Castr" on the denier can be seen on the reverse.

S.2486 on a denier of John Orsini (1.01g).

Actually, there are difficulties with S.2483. It should be noted that only one of the S.2483's in the Longuet hoard was struck on a denier, which Bendall says - or perhaps assumes - was an issue of John Orsini. On the other hand, I have recently (Jan. 2007) seen two examples of this type which may have been struck on debased Catalan deniers. If so, these deniers, which were issued after 1311, may also have been obtained as a result of the recovery of Neopatras by Andronicus III in 1333, which would of course date S.2483 from the middle 1330's. However, this is all quite conjectural, and it's also possible that the Catalan deniers in question - if that's what they are - were acquired by the Byzantines somehow at an earlier date. If so, then S.2483 could well date from before 1330, and well before S.2486-7.

Bendall's hoard of 2000.

The next step is Bendall's reconstructed hoard of 2000* (Num. Chron. 2001). This find included c. 370 coins, mostly later Thessalonican trachea of Andronicus II, or Andronicus II & Michael IX, together with a few of the "official" Andronicus III types. (These coins were sold through various dealers in Europe in 2000, and also H.J. Berk in the U.S, starting in BBS 116). As we shall see, the differences between this find and Longuet's hoard give us numerous clues as to the correct attribution of the later Thessalonican trachea . (For the statistical details of all the finds considered here see the Article "Sequencing the Thessalonican Trachea" ).

Now Bendall seems to assume that the "2000" find is a typical discrete "snapshot" hoard, i.e, that it was formed "as a piece" at a particular time (as a merchant's working capital, for example), and hence that it gives a snapshot of the coins circulating at that time. Therefore, since it seems to include some of the coins which we have just attributed to Andronicus III, we might have expected him to date it to the mid 1330's, but instead, for reasons that are not at all obvious, Bendall prefers to assign the hoard to the later 1320's, and assumes that the "Andronicus III" types present are either intruders or not really Andronicus III's at all.

Bendall gives no reason for this dating (possibly it is the near absence of the assarion S.2481), but in any case it is clear that whether we date the hoard to the later 1320's or earlier 1330's there are difficulties in reconciling the composition of  the 2000 find with the quite different composition of Longuet's "Thessalonica" hoard of mainly Andronicus III types, which prima facie dates from only about ten years or so after Bendall's hoard. Shifting the latter to the 1320's eases this problem, but doesn't, in my opinion, really solve it, and hence I think we have to consider a somewhat different idea, namely that the 2000 find is not a "snapshot" hoard at all, but a "quasi-accumulation" dating from (or rather terminating in) the 1330's. This need not necessarily mean, incidentally, that it is a typical accumulation, the result of stray losses in some public place - more likely it could represent a hoard put together over time, as the savings of a household, for example. Such an accumulation would produce a broader range of types than a one-off hoard, in line with the actual composition of Bendall's find, and quite different from Longuet's hoard.

* This hoard is now referred to in the literature as Thessalonica (VII) 2001.

The 2004 Finds.

During 2004-6, a large number of Palaeologan coins were sold on Ebay and elsewhere, many of which apparently came from a single source (or number of related sources - see Data page "Some Palaeologan Finds" for details). These finds (the "2004" finds), consisting mostly of coins from the period of Andronicus II, have basically much the same composition as Bendall's hoard, except that they seem to terminate around 1330, or earlier, as they would seem (with perhaps one exception) to contain none of commoner types which can be definitely attributed to Andronicus III. Like Bendall's hoard, it seems reasonable to asume that these finds are more likely quasi-accumulations than discrete "snapshot" hoards.

Dochev at Turnovo.

This last conclusion is reinforced by a third line of evidence - Dochev's excavations at Turnovo, the mediaeval capital of Bulgaria. This work yielded, amongst other things, very large numbers of (mostly) stray Byzantine and Bulgarian coins from the period of the second Bulgarian empire, including hundreds of examples of the issues of Michael VIII, Andronicus II and Andronicus III from both Constantinople and Thessalonica (K. Dochev, "Coins and Coin Usage at Turnovo,  (XII-XIV c.)", 1992).

It is reasonable to assume that the Turnovo coins give us a statistically representative snapshot of the petty coinage of Constantinople and Thessalonica in the earlier Palaeologan period. It is therefore instructive to compare the 2000 and 2004 finds with the Palaeologan coins from Thessalonica at Turnovo, which roughly match the other hoards in total numbers.

Overall, the distribution of types in the recent finds closely matches that of the Thessalonican coins of Andronicus II found at Turnovo, at least for the main period of Andronicus's reign. These finds contain few issues of Michael VIII, and also seem to largely lack types from the earliest years of Andronicus II (although these were not common at Turnovo either), while at the upper end of the time frame they mostly lack the late types S.2484, 2497 and 2501, all of which were found at Turnovo, although again not in large numbers.

Overall then, it seems that the recent finds are essentially a (more or less) representative sampling of the bronze coins issued by Thessalonica in period from c.1290 to c.1330 for the 2004 finds, or perhaps c.1290 to c.1340 in the case of Bendall's hoard.

However, it has to be said that there are possible problems with this idea, since it might seem unlikely that all these different finds, apparently laid down separately as domestic accumulations over quite a long period, would all wind up with essentially the same composition, so that in the end it might yet be easier to assume that they are all simply discrete commercial hoards. (For a more detailed discussion of this question see the Note "Bendall's 2000 Hoard").

Fortunately, whatever the actual case may be, the exact nature of the finds is not really crucial for our purposes. The important point is that there is not much overlap between the coins in the Longuet hoard and those in the other finds, indicating that for the most part they were minted in different (but still overlapping) periods. This in turn allows us to draw some useful conclusions about the assignment of various types appearing in the two finds, which are summarised in the next section. 

The Assignment Rules.

Firstly, for the commoner types at least, we can assume that those Longuet types which did not occur in the 2000, and particularly the 2004, hoards are prima facie Andronicus III's, or possibly late issues of Andronicus II (since the Longuet hoard included more than enough types for Andronicus III's reign - enough to extend, on a one issue per year basis, back to the earlier 1320's at least).

Secondly, if a Longuet type is also found in the 2000, or more particularly the 2004, hoards, then we can surmise that it is most probably an earlier issue of Andronicus III (since the 2004 hoards either terminate, or at least peter out, in the middle 1330's at the latest), or possibly a later Andronicus II type. (Note that many of these "late Andronicus II/early Andronicus III" types are quite probably issues of the joint reign period of 1322-28, but I will not spell this out in what follows).

Thirdly, types which are common in the 2000 and 2004 hoards but absent from Longuet's hoard are almost certainly issues of Andronicus II, since, as we have seen, the Longuet hoard more than covers the reign of Andronicus III.

So let us apply the above rules to the remaining Andronicus III candidates, considering firstly the commoner Longuet types, i.e, those which occurred more than once in the Salonika hoard (since in these cases the hoard evidence has greater statistical weight).

S.2497-99 (and S.2561).

Considering S.2497, we see that there are 10 examples in Longuet's hoard and apparently none in the recent hoards (except perhaps for Bendall's hoard*), so I think we can safely assume that S.2497 is an Andronicus III type. This conclusion is reinforced by the fact that it is a flat type, meaning that it probably dates dates from the 1330's (the later Andronicus III's seem to be basically small flat coins, a trend which seems to have started around 1330). Furthermore, a recent overstrike reported by Bendall shows this type apparently overstruck by an example of S.2487, which suggests that it dates from the earlier 1330's.

Now, since S.2497 is a mounted emperor type, it might seem that we can assume that the presumably related types S.2498 & 2499 are also late Andronicus III's. However, S.2498 (Longuet 8) is apparently unique, and so it seems quite possible that this somewhat messy coin is actually a mule of S.2497 with a St Demetrius type (or perhaps an overstrike - the actual coin is shown in PCPC as #266.1).

Also S.2499, known from only a handful of examples, presents us with a serious problem, which, as we shall see, leads us into distant fields. This rare type was not present in Longuet's hoard, but single examples turned up in both the Pella and Serres hoards of mainly John V and Anna of Savoy types, finds which probably date from the 1360's. Now S.2499 was assigned to Andronicus III in LPC because of its close similarity to S.2497 and also because of the worn state of the Pella coin (= PCPC 267.1 - the Serres example is also worn). However, if S.2499 were an Andronicus III type from Thessalonica, it would have been most unlikely, given its rarity, that it would turn up in the later hoards when none of the other more common Longuet types did (except that, oddly, two examples of S.2500, another scarce Longuet type, did in fact appear in the Serres hoard).

Hence we have a problem - to when (and perhaps where) do we assign S.2499? If, for the moment, we take the hoard evidence at face value and then draw the most obvious conclusion, namely that S.2499 is a post 1342 issue, then we have to assume that this type is most likely an issue of John V's first period of rule alone at Thessalonica, in 1351. However, this assignment leads to several problems.

To begin with, Bendall assigns a quite different type, namely the first John V and Anna "assarion" (S.2516), to 1351. However, this is not necessarily a serious problem, since we could possibly relegate S.2516 to the period 1354-1365(?), during Anna's semi-autonomous rule in Thessalonica, along with the second series of  "assaria" (S.2517-23).

But there are other more serious problems - firstly, S.2499 is clearly stylistically incompatible with the other issues of John V and Anna at Thessalonica (which feature squared B's for example - cf. S.2517 in Sear), and of course it is undeniably stylistically identical with S.2497 (but note the device, perhaps a church or shrine, to the right of the saint on S.2499). Secondly, the S.2499's in the Pella and Serres hoards are both a lot more worn than the 2516's, so even if we move the latter a few years later to the mid 1350's, it still seems difficult to accommodate S.2499 in 1351. Overall, we seem to have no option but to accept the assignment of this type to late Andronicus III, or perhaps the earliest years of John V, so that for the moment at least we are back where we started.

(For interested readers, the Serres hoard was detailed by Protonotarios in the American Journal of Numismatics, Series 2, 1990 - it included 28 examples from the two basic "assarion" series of John V and Anna, two rather battered S.2500's, and worn single examples of S.2362, 2383 and 2499. The Pella hoard included 7 examples from the two "assarion" series, plus the S.2499. The presence of the rare type S.2362 in the Serres hoard suggests that this is a post 1342 issue, so maybe this type at least can be assigned to 1351 - although for some unexplained reason Protonotarios gave it to Andronicus III. Again, note the church or shrine behind the saint on the reverse of this type, which seems to link it to S.2499 - a somewhat similar device also appears on the later John V "Martyrdom of St Demetrius" type, S.2593, although there it is behind the emperor, and looks more like an altar with a reliquary box).

Finally, we also need to consider here the rare "assarion" S.2561, which is basically S.2497 with a lattice cross (actually, a lattice of four small such crosses) to the left of the monogram on the obverse. In LPC Bendall and Donald followed Goodacre in giving this type to Manuel II, but its weight and fabric suggest that it might be an earlier issue, and in DOC V Grierson assigned it to Andronicus III. While this is a not unreasonable idea, this type, like S.2499, did not appear in Longuet's hoard, and there is no hard evidence linking it to Andronicus III rather than John V. Furthermore the style, particularly the squared "B" on the reverse, indicates that this type postdates Andronicus III, so that it seems reasonable to date it to the period of John V and Anna at Thessalonica, or maybe to John's stint as ruler of the Chalkidike region at Ainos in Thrace in 1352-3.

*  According to Berk's (as quoted by Bendall) there were no examples of S.2497 in the 2000 hoard, which is a bit of a problem if this type precedes S.2487. However, S.2497 is not as common as 2487, and also there were in fact two examples of the type in Berk's sales in 2003-4 (BBS134-597 and BBS136-365), which may or may not have been part of the 2000 hoard. In any case we note that this type doesn't occur in the 2004 finds (which probably date from c.1330 at the latest), confirming its dating to the 1330's.

S. 2484, 2500, 2501.

For S.2484, 2500 and 2501 we see that while there were several examples of each of these types in Longuet, none occurred in the recent hoards, so they too are most likely Andronicus III's. Again, these types are either flat or flattish, suggesting that they are late issues. Also, of course, if the second figure on S.2500 and 2501 is John V, then these must be late types, but unfortunately, although S.2500 does in fact have a reverse legend, it is invariably garbled on the known examples so that the reported readings of the legend usually cannot be relied on (although on some of the better more recent examples it does seem reasonable to read the right hand legend as "Iwa...").

Actually we need a couple of caveats here - firstly, as just noted, S.2500 has a (garbled) legend, unlike all the other Longuet types (except Longuet 1, where the legend is part of the design, and Longuet 5, which has a three letter slogan of some sort). This perhaps suggests an earlier date for this type, and in fact it is attributed to Andronicus II & III in DOC V (at least on p.160 - on p.181 it is an issue of John V and Anna!). However, in the case of the earlier dating we might have expected that this type would have shown up in the recent finds, but it didn't, so for the moment I incline to a later dating, although the second figure doesn't seem to be Anna. Also, the type is apparently found overstruck by S. 2501 (cf. ANSMN 29, Pl. 26,2, also shown in PCPC as #275.1), which suggests a close relation between these two types. Finally, two examples of S.2500 appeared  in the Serres hoard of the 1360's, also suggesting a late date for the type, perhaps even after 1342. Overall, the assignment of S.2500 has always been a problem, and it appears that it still is.

Secondly, according to Bendall, one of the S.2484's in the Longuet hoard was overstruck on an example of S.2367, a not very common and probably much earlier type of Andronicus II. Now such an overstrike is not impossible, but it would be rather surprising (since the Palaeologan overstrikes mostly seem to be on recent undertypes), and my guess is that the undertype is more likely S.2490, or possibly 2488. It would therefore be nice to see this overstrike confirmed - overall, it's a great pity that all the overstrikes from the Longuet hoard weren't properly published by Bendall (all we have are the small photos in ANSMN 29). But in any case, the flat fabric of S.2484 confirms this type as an issue of the 1330's.

Note that it's possible that one or two of the types which we are attributing to Andronicus III are actually early John V's, from 1341-2 (from 1342 to 1350 Thessalonica was under the control of a quasi-autonomous local administration, which apparently issued no coinage).

We should also note here that the Palaeologan coins found during Dochev's excavations at Turnovo provide an independent check on our conclusions for the later Andronicus III types. There were roughly as many Thessalonican coins of the reigns of Andronicus II and Andronicus III at Turnovo as in the recent hoards, but the Turnovo coins included not just S .2483, 2486 and 2487's of Andronicus III, but also several S.2484, 2497 and 2501's, which are of course essentially absent from the recent hoards (except perhaps for the possible S.2497's in Bendall's hoard mentioned above). This suggests, as with the recent hoards, that these last types postdate the 2004 hoards at least.

S. 2488, 89, 91.

Next we consider those rare types which appeared only once in the Longuet hoard, and didn't appear in the recent finds, namely S.2488 (Longuet 6), S.2489 (Longuet 5), and S .2491. (The last Sear type includes both Longuet 16 and Longuet 17 - in LPC and Sear these two types are combined into one, but I take them to be separate types here, as in PCPC).

For these types the statistical weight of the evidence is minimal, so all we can say is that these are either Andronicus III's, or late Andronicus II's. Given the turned up lip on the Longuet 5 coin (= PCPC 260.1) we can probably conclude that this type (S.2489) predates S.2486, and hence dates from the very early 1330's, or even earlier. This same applies to S .2488  (cf. the deep lip on Gr.1470 = DOC V 930, and note the obverse wear on the Longuet 6 coin (= PCPC 258.1)).  (See also S.2493 below, which seems to copy S.2489 but with the reverse mirrored).

Longuet 16 is difficult to assess - it is only known, as far as I am aware, from the Longuet coin, and it is not clear that it is a real type. As Bendall suggests, the Longuet 16 coin (= PCPC 273.1) is probably overstruck on a denier, possibly of John Orsini, which would make it an issue of the mid 1330's. But even if this is so, it is still possible that this coin is a hybrid, perhaps a mule of a monogram type with S.2486.

Longuet 17 (= PCPC 264.1) is a little more convincing, being known from at least two examples. The reverse is emperor h/l. under "canopy" (city walls more likely), holding aka . and gl. cr. Its dating is uncertain; the DO example is a brockage, seemingly overstruck on a starred type, perhaps S.2490, although there are other possibilities, and it could possibly be the ultimate undertype in a doubly overstruck example of S.2486 in Longuet's hoard (see discussion of S.2364 below).

(Note that Bendall's report on the 2000 hoard also listed an example of S.2488, supposedly in that portion of the hoard which was sold through Berk's. If so then this type may be late Andronicus II (i.e, during the joint reign), although Andronicus III is still quite possible. However this coin has not to my knowledge shown up in any Berk's sale, and hence was possibly listed in error (Bendall of course had to assume it was an intruder in the hoard in any case). Or perhaps it was sold privately - it would be nice if Berk's could enlighten us.

S. 2485 (Longuet 15 and Grierson 1468).

In Sear, Bendall (presumably following Grierson) has conflated two different coins into the fictitious composite type S.2485 - firstly, the rare standing emperor type Longuet 15 (LPC 236:5, not in DO), where the emperor holds a large patriarchal cross and a sceptre cruciger, and secondly the more common seated emperor type Grierson 1468 (= DO 922), where the emperor holds a patriarchal cross and a large B* - cf. also Dochev p. 253, 28 for this last type. (For an example of Longuet 15, denoted here as S.2485a, see the coin below, or PCPC 269.1, which is actually one of the Longuet coins. For another interesting example, possibly overstruck on Gr. 1468, denoted here as S. 2485b, see the Overstrikes page).

Sear 2485a, Longuet 15 (0.9g).

The obverse of Longuet 15 is a problem - on one example we see clearly a B. Mil. St, but this could be a mule or overstrike, since on most examples what could be the Virgin appears, with what seems to be a large open Lis to the left (and right?); the reverse is very similar to that of the much commoner S.2377, but as well as holding a sceptre rather than an akakia, the emperor usually wears a somewhat odd looking old style loros, with crossed sashes across the shoulders, and there is a manus dei above left and no inscription above right. There is a (small) star below left, as on 2377, but also another below right, although this is off flan on most examples.

There were four examples of Longuet 15 in the Salonika hoard, and supposedly two "S.2485's" in the 2000 hoard, presumably BBS116-717 and BBS117-521. Of these last two coins, the first is the seated emperor type, and the second is probably an S.2486, judging from the way the emperor's left arm starts out close to his side. (BBS128-717, which is possibly from the 2000 hoard, is also listed as an S.2485, but statistically it is more likely S.2377). Thus Longuet 15 apparently does not appear in the 2000 (or 2004) hoards, and hence we can probably safely conclude that this type is an issue of Andronicus III. 

And the seated emperor type? - well, it's a small type (usually - the Grierson coin is abnormally heavy) with no legend, and it has the classic turned up lip of the late Andronicus II types. Also it's in both the 2000 and 2004 finds, but not Longuet's, so it's either early Andronicus III, or more likely, late Andronicus II (i.e, joint reign with Andronicus III?), and is clearly not connected with Longuet 15. (Note that the BBS116-717 coin can be seen in Bendall's Num. Chron. report as coin K, Plate 50. Another example of this type, perhaps part of Bendall's hoard, was offered by Berk's as BBS136-360. Note also that DO 788, listed as S.2377, is possibly an example of Longuet 15).

* In fact, the emperor doesn't actually hold the "B" - his left hand is at his waist.

S.2482, 2490.

Now for S.2482 and 2490. Here the situation is quite different. The were eight S.2490's in the 2000 hoard and only two heavily worn examples in Longuet's; similarly there were no less than seventeen S.2482's in the 2000 hoard and none in Longuet's (except perhaps for Longuet 24, although this could also be an S.2465). As well, there were, at last count, six S .2490's and 36 S.2482's in the 2004 finds, numbers generally comparable with those in the 2000 hoard. Hence these two types would both appear to be Andronicus II issues, or less probably, early Andronicus III's. Now, according to Bendall, one of the Longuet S.2490's is overstruck on a two emperor type, which he assumed was the rare (unique?) type S.2453 (supposedly an issue of Andronicus II & Michael IX at Thessalonica). This suggests that both S.2490 and 2453 (if that's what the undertype is) could both be issues of the joint reign of Andronicus II and Andronicus III. Another more likely candidate for the undertype would of course be the similar but much commoner S.2465, which is also usually taken to be an issue of the joint reign, although it is quite likely earlier - see the last section below.

In fact, as far as S.2482 is concerned, we can can go further and confidently date it to Andronicus II, since the presence of the S.2490's in Longuet's hoard, but not the much commoner S.2482's, seems to show that 2482 predates 2490 by some time. (For what it's worth, we also note that on both these types the emperor is often shown with a pointed beard, as is typical with issues of Andronicus II at Thessalonica).

Finally, it is worth noting that S.2482 is often struck from clearly curved dies, and on relatively heavy flans, which marks it out from the flattish dies typical of the later Andronicus II/early Andronicus III types - see the example pictured below.

Sear 2482 (2.07g).

The significance of this is unclear - most likely, as suggested here, this coin is relatively early, and/or these dies were the product of an above average celator; see the last section of this article, and also the last section of the Note "S.2315 of Michael VIII and S.2486 of Andronicus III".

P.S: We now (Oct. 2010) have an overstrike of S.2377 on S.2482, confirming the above allocation.

(Note that it is an odd fact that until quite recently S.2482 was known, in the west at least, from only two or three examples - presumably one reason for its attribution to Andronicus III in LPC - whereas in the recent finds it's one of the commonest types).

S.2486 and S.2315.

At this point we need to return to S.2486. It has become fashionable to assume that there are in fact two separate versions of this type, one the standard type described in Sear, and another with added stars left and right on the reverse. The latter type has had an interesting metamorphosis - it appears to be nothing more than S.2315, a type originally attributed to Michael VIII by Bendall and Donald, but which is now supposed to be a variant of S.2486.

However I can't accept the current association of the two types - the starred type is clearly earlier than the standard S.2486, since in the Salonika hoard none of the 11 examples of S .2486 (Longuet 14) appear to show any stars. Also, with one or two possible exceptions, all of the "S.2486's" in the recent  "2004" finds of mainly Andronicus II types are starred types,

Overall, this suggests that the starred S.2486 (i.e, S.2315) is more likely an issue of Andronicus II than Andronicus III (or Michael VIII). For a more detailed discussion of this question see the Note "S.2315 of Michael VIII and S.2486 of Andronicus III". (Note also that Dochev assigns this type to Andronicus II).

P.S. We now (Oct. '06) have an example of the starred type which is clearly overstruck by S.2387 - see "Overstrikes and Mules" - confirming its assignment to Andronicus II.

(Note that while most examples of the real, i.e, unstarred, S.2486 are struck with flat dies on mainly flat flans, some are clearly scyphate. It's possible that these "scyphate" S.2486's are issues of Andronicus III, but their presence in the recent finds of Andronicus II types suggests that they may actually be non-starred variants of S.2315, and hence issues of Andronicus II - again, see Note "Bendall's Hoard of 2000". Note that Bendall thinks DO 928 is an example this type, but I don't).

The "Additions", S.2492-96.

Next we consider the five "Additions", S.2492-2496, beginning with S.2494.

Now Bendall does not include S.2494 in PCPC, and in fact, as will be seen below, in recent times he actually forgot all about it. But S.2494 is a real type, although a rather odd and not very common one. One example appeared in CNG 53 (Lot 1593) and another was offered by CNG through Ebay on Feb. 2, 2001. Also, BBS126-483, possibly from Bendall's hoard, seems to be an example of the type (although it was listed as S.2354 of Andronicus II). More recently, two examples were offered by CNG in Sale 63 (Lot 1733) and Sale 66 (Lot 1812), and another example was offered on Ebay in April 2005. (The latter coin, probably not from the "2004" finds, is shown below; the CNG Ebay coin can be seen on the wildwinds.com site under Andronicus III, and the CNG 63 and 66 coins can be seen on the coinarchives.com site, if you are a subscriber, or at CNG itself for free).

Now, the last two CNG coins mentioned were not listed as S.2494's, but rather as examples of a "new" type first described by Bendall  in Num. Circ. 1997 (p.198), which was supposed to be similar to S.2494, but flat, and with a symbol like a broad "K" beneath the monogram (standing perhaps for Constantinople). However, for reasons which will be explained later, it seems that the CNG coins are probably just ordinary S.2494's (as is the "new" type, for that matter).

So where does S.2494 belong? Although most of the examples of the type listed above are well-struck, only one weighs more than 2.0 grams, so at first sight it might seem that this type is more likely an issue of Thessalonica rather than Constantinople. Furthermore, if the BBS coin was indeed part of the 2000 hoard, then this would support a Thessalonican origin for the type. But if so, to what period does it belong? It certainly doesn't seem to be a typical later issue of Thessalonica - for a start, many of the known examples are (reasonably) well struck on neat, uniformly scyphate flans, and hence are quite different to the typical Andronicus III's, with their flattish flans, and also to the later Andronicus II's, with their ragged flans, flattish fields and turned up rims. Furthermore, the Palaeologan monogram on S.2494 differs from that on the Longuet types S.2491 (both types) and 2497 (i.e, it has L over (after) PA, as on most of the Andronicus II monogram types, instead of the L after AP found on the Andronicus III types). Hence if this type is in fact an issue of Thessalonica, then it presumably dates from the early to mid period of Andronicus II.

If so, then perhaps it belongs with other noticeably scyphate Andronicus II types like S.2372 and S.2482. But there is still a difficulty - on S.2494 the emperor is shown with a rounded beard, as on the Constantinople assaria, and as opposed to the forked or pointed beard typical of the Thessalonican trachea of Andronicus II, so that maybe this type is an issue of Constantinople after all. Or maybe it derives from some other mint altogether.

For the moment then, given the conflicting evidence, I think that all we can really say is that the attribution of S.2494 remains uncertain, as regards both emperor and mint (more of this below).

And what about Bendall's new type? Well, this was originally supposed to be a flat coin (Bendall initially took it to be a "tetarteron" of Andronicus III from Constantinople). However, Bendall later published several further examples of the type which were clearly trachea (Num. Circ. 2002, p.189, Nos. 13 and 14(?), and 2005, p.90-1, Nos. 2 and 3), and it seems to me that traces of the K symbol, or something similar (KN perhaps), appear on several of the other examples of S.2494 that we are aware of.

Furthermore, it now seems (pace Bendall) that the original "new" coin of 1997 was actually just a more or less flat example of the scyphate type, which, as the crease line on the coin shows, has been further flattened at some time. (Note that in his first two articles on the "new" type, Bendall didn't actually refer to the original S.2494 type of the "Additions" at all, since, as he later admitted in the 2005 article, he had managed to forget all about it).

So are we dealing with two different varieties of S.2494 here, one with a K, or KN, and one without? It's possible, but it seems more likely that there is just one basic type, with, on some examples at least, varying symbols below the monogram, .

P.S: In his 2005 article, Bendall reported a supposedly related type with a reverse similar to S.2494 but with a cross on a floriate base replacing the monogram on the obverse (p. 90, Coin 11) - however this coin seems to be a known trachy of Ivan Alexander.


Sear 2494 (2.00g)

(For those interested in these intriguing coins, the reverse inscription starts with A/\ (i.e, A N) above right, then runs D/PO/N-K/O (or sim.) down the left side, ending with C or O/C or K/O/C below right, running upwards, although the die-cutters often seem to combine the end of the legend with the emperor's left sleeve cuff, producing a sort of messy shield, as on the coin above. As Bendall has suggested, this unique arrangement of the legend suggests that this type may have been issued by Andronicus III at a mint other than Thessalonica or Constantinople, possibly at Adrianopolis or Didymoteichon in Thrace, where Andronicus resided in the early 1320's. In fact it is tempting to specifically date the coin to 1321-22, when Andronicus ruled in Andrianopolis in his own right after revolting against his grandfather. This could well explain the stylistic and fabric differences between S.2494 and the normal issues of Thessalonica, although it should be remembered that there is no hard evidence to support this idea. In this case the K, or KN, on the obverse - if that's what it is - could perhaps be taken to mean Komnenos, as the Palaeologans were not averse to flaunting their Comnenan connections).

S.2493 (DOC V 845) is apparently known only from two poor examples* , one of which (the D.O. coin) is overstruck apparently on S.2368. This type hasn't appeared in any hoard and judging from its crude style it is possibly an unofficial type. As a reader has pointed out the reverses of this type and S.2489 are mirror images, and this fact, together with the crude style, suggests that S.2493 might be a local copy of S.2489. (Or are the two types just variations of the one type?).

S.2495 (Grierson 1472 = DOC V 934) was originally based on single coin collected by Bertele showing a standing emperor holding a patriarchal cross in his right hand, with a large star at his left shoulder. Like S.2494, this coin is not listed in PCPC, and initially I thought that it was possibly some kind of overstrike, but a reader has reported another example of the type in his possession, with the obverse design possibly being the Virgin. This is a very rare type, with a light flattish flan, and hence is probably Andronicus III. (Note that in Sear the emperor is incorrectly described as holding the patriarchal cross in his left hand).

Note that neither of the last two types were in Longuet's hoard, and therefore they could in principle be post 1342 issues of John V, although stylistically this seems unlikely. However, S.2495 is probably earlier Andronicus III at the latest, since Bertele collected most of his coins in the western Balkans, an area which was lost to the Byzantines in the early 1330's.

The "two cities" type S.2496 is also not very common, but there was an example in Longuet's hoard (as Longuet 23, although Longuet himself didn't fully describe it, and it was originally omitted from LPC), and another (reportedly) in Bendall's hoard, possibly the coin sold as BBS 134-596. Also, several examples appeared in the recent "2004" finds . Hence it is either early Andronicus III, or a late issue of Andronicus II.

Finally we have S.2492, the "two orbs" type. Here we actually have some solid facts to go on - firstly, this is not a Longuet type; secondly, there were three examples in the 2000 hoard, and several more in the similar recent "2004" finds of largely Andronicus II types; thirdly, it is known overstruck on both S.2360 and 2391, and fourthly, and most importantly, it is known overstruck by the two emperor type S.2457 (see Data Page
"Overstrikes & Mules"). Hence I think we can safely conclude that 2492 is most likely a fairly late Andronicus II type.

*  Another example of S.2493 has recently (Feb. 2016) surfaced. The obverse is unclear but seems to be show a B. Mil. St. A fourth (brockage) example appeared in July 2017.

Grierson 1473.

Now for Grierson 1473. As noted above this coin was omitted from LPC and Sear, but it is now included in DOC V as an issue of Andronicus III (DO 936-40). It is very similar to S .2459, but there is a large B (often not clear) to the right of the monogram on the obverse, while on the reverse St Demetrius (on the left) and the emperor hold a Lis on a shaft between them. This type appears in quantity in Bendall's hoard and in the other recent finds, but not in Longuet's hoard, so it's clearly an issue of Andronicus II. Also, it (normally) has a well defined reverse legend, and as we have seen, it is very similar to S.2459, which is securely established as an issue of Andronicus II & Michael IX. On the whole I think it's safe to conclude that this type (Gr.1473) is a mid period issue of Andronicus II.

Longuet 1 and Longuet 2.

So what's left? Well, not surprisingly, we are left with those perennial Palaeologan problems, the crouching emperor types Longuet 1 and Longuet 2. Now I have discussed these types at length elsewhere, so can we say anything new here? I think we can, although perhaps not as much as we would like.

In LPC there are three Thessalonican bronze types showing a kneeling (or rather, crouching) emperor blessed by Christ - namely LPC 204:1 (Longuet 1), attributed in LPC to Andronicus II, 234:1 (Longuet 2), attributed to Andronicus III, and 260:3, attributed to an uncertain emperor in LPC, and later, in PCPC, to John V.

Now these types are all very similar, particularly the last two. For LPC 204:1 (Longuet 1) the obverse is a bust of St Michael, while the reverse copies the hyperperon of Andronicus II, with a full legend above the emperor, and IC/XC (and sometimes part of the legend) behind Christ. LPC 260:3 is similar, except that the obverse is a bust of St Demetrius, while on the reverse the legend has been changed into a cross-in-circle with manus dei "supporting" the emperor, with only an X to the right of Christ (often looking like a sceptre cruciger, which it may well be). LPC 234:1 (Longuet 2) is very similar to 260:3, but without (it was thought) the manus dei above the emperor.

Now, in PCPC, Bendall reduced the three original crouching emperor types in LPC to two, by equating LPC 234:1 with 204:1 (i.e, by equating Longuet 2 with Longuet 1, which meant taking the obverse of Longuet 1 to be St Demetrius rather than St Michael). Unfortunately this seems to be incorrect. The original Longuet 2 coin is not very clear, but the large X behind Christ would appear to mean that this type is the same as LPC 260:3, not 204:1. The two actual types then are Longuet 1 (LPC 204:1 = S.2359 = PCPC 263 = Gr.1428) and Longuet 2 (LPC 234:1 = 260:3 = PCPC 284 = DO1188-9; not in Grierson or Sear). (Curiously, in ANSMN Bendall did in fact equate Longuet 2 with LPC 260:3, but he evidently changed his position later in PCPC).

Note that the obverse of Longuet 1 has been somewhat uncertain in the past (which is what enabled Bendall to equate Longuet 1 with Longuet 2), but it is in fact definitely St Michael, as was clearly shown for example in a 2005 offering of the type by CNG (E-Auction 108-329). Another example of this type showing St Michael reasonably clearly appears below, and cf. also Gr.1428 (and now DO 913-7).

Longuet 1 - S.2359 (1.53g).

But perhaps there is still a problem here - Sabatier ("Monnaies Byzantines") depicts, in line drawings (Pl. LX, No. 5), a coin which does in fact correspond to Bendall's combined type, i.e, it has the reverse of Longuet 1, but St Demetrius on the obverse. However, I know of no actual examples of this type, so I have to assume Sabatier has made an error here. Perhaps he had a poor example of Longuet 2, and filled in the reverse legend from the hyperpyron, or maybe he simply conflated two poor examples of the two types (like other 19th century illustrators in the case of uncertainty Sabatier tended to draw what he thought should be the case, rather than just what he had in front of him).

Leaving aside the Sabatier coin, then, we are left with just the two Longuet types here, and the question now is - to which emperors do we attribute them? The hoard evidence is not conclusive either way. Both the Salonika hoard and the 2000 hoard contained just one example of Longuet 1, and this type has also appeared in the recent "2004" finds, so this type could be an early period Andronicus III or possibly a late Andronicus II. It is reasonable to assume that it is an accession issue of Andronicus III as sole emperor in 1328 (particularly as t he emperor is shown with a rounded beard), but there is no specific evidence to confirm this idea. (It probably doesn't represent the coronation of Andronicus III in 1325, as such an issue would presumably also have shown the senior emperor Andronicus II).

One thing we can say is that Longuet 1 can hardly be a coronation issue of Andronicus II, as originally implied in LPC, since then it would have to be a very common type to appear in the Longuet hoard 60 years later, whereas in fact it is not common at all. In any case it is too small to be an early type, and given its scarcity, the fact that Longuet 1 appears in the Longuet hoard at all suggests that it can't be much earlier than the mid 1320's. In terms of fabric the CNG coin is virtually flat (or flattened), but other examples of the type have somewhat more scyphate flans, and overall Longuet 1 seems to belong with the late Andronicus II/early Andronicus III types, rather than the flatter late Andronicus III's.

Longuet 2 is even more confusing. Basically, Longuet 2 seems to be a simplified or "provincial" version of Longuet 1, and in the "official" scheme, and in DOC V, this type (as LPC 260:3) is assumed to be an inaugural issue of John V from 1341. This attribution is, on the face of it, quite tempting, but it really represents a hangover from the original "three type" theory, and is based more on style than on any hard evidence.

There was one example of Longuet 2 in the Salonika hoard, and according to the list supplied to Bendall by Berk's there was also one in their part of the 2000 hoard, presumably the coin sold as BBS128-719. Now, since Bendall dated the 2000 hoard to the later 1320's he had to treat this latter coin as an intruder, as he assumes the type was an issue of John V (on the face of it this is actually quite reasonable, since the BBS coin is a poor, clipped specimen which may not have been part of the find, like several of the Palaeologan coins in this particular sale). However, here we don't necessarily share either of Bendall's assumptions, so we don't have to treat this coin as an intruder, although of course if it is in fact part of the find then it means that Longuet 2 is probably, although not definitely, earlier than John V.

In terms of fabric Longuet 2 (like Longuet 1) seems to belong with the somewhat scyphate late Andronicus II/earlier Andronicus III types (see the example below), although again it can hardly be too much earlier than the mid 1320's. Note incidentally, that the manus dei extending from the cloud is sometimes drawn as a Lis behind the emperor, as on the coin shown.

Longuet 2 - LPC 260:3 (1.24g)

However, the story is confused by the fact that the original Longuet 2 coin itself appears to be overstruck on an example of S.2486, which would imply that the type dates from the mid 1330's. (Either that, or the undertype is S.2315, rather than 2486. However, this seems most unlikely, as it is now clear that S.2315 is a lot earlier than Longuet 2). Such a dating is perhaps supported by the fact that, in contrast with its possible occurrence in Bendall's hoard, Longuet 2, unlike Longuet 1, has not appeared in the recent 2004 finds, suggesting that it post-dates Longuet 1. As well, we now have overstrikes of Longuet 2 on S.2497, and S.2500 on Longuet 2 - these establish the sequence of the later issues of Andronicus III, but not necessarily their absolute dates (cf. Data Page "Overstrikes and Mules").

Summing up, it has to be admitted that Longuet 1 and 2 remain significant problems. On the basis of rather scanty evidence, about the best we can say is that Longuet 1 seems to be an issue of the mid to late1320's, and may, or may not, be an accession issue of Andronicus III. Longuet 2 probably dates from around 1335 or later; this type could possibly make sense as the accession issue of John V (if you can accept a 9 year old boy as the crouching figure), but this means putting S.2500 and 2501 into the reign of John V.

(It's worth keeping in mind here that there were plenty of crownings, or at least, elevations to the purple, in this period. Apart from being crowned in Constantinople in 1325 and becoming sole emperor in 1328, Andronicus III was also made co-emperor by his father Michael IX in Thessalonica in early 1316. Unfortunately, this eventually led to the civil war between Andronicus III and his grandfather, and in June 1321 the empire was split between Andronicus II in Constantinople and Andronicus III in Adrianopolis, only to be recombined under their joint rule a year later - well before the formal crowning of 1325. However, it would seem more likely that most of these events would have inspired a two -emperor type - the scyphate version of S.2435 perhaps (Num. Circ. '77, p.143, 13), or maybe S.2453 or 2465 - rather than a single emperor type like Longuet 1).

S.2364 (Longuet 21).

There are still two Longuet types to consider, beginning with Longuet 21 (S.2364). As noted at the beginning, Bendall and Donald assigned this type to Andronicus II rather than Andronicus III, presumably because of its similarity to S.2365, and perhaps S.2360. Longuet 21 is extremely rare - in fact the only published examples I know of are the Longuet coin itself and another coin, apparently from the same dies, found at Turnovo (Dochev Pl.18, 2), plus a possible overstruck example to which I will come shortly*.

Now the Longuet coin (which is the one pictured in Sear) is quite well preserved, far better than any of the other Andronicus II types in the hoard, and this fact, taken with its rarity, suggests that it could well be an issue of Andronicus III, or possibly even later. Furthermore, it appears that one of the examples of Longuet 14 (S. 2486) in the Salonika hoard is actually overstruck on an example of Longuet 21 (cf. ANSMN 29, Pl. 26,1), which would again suggest that Longuet 21 is an issue of Andronicus III, possibly a short lived type which was replaced by Longuet 14 for some reason. On the other hand the Longuet 21 undertype (if that's what it is) is itself apparently overstruck on an Andronicus II type (S .2392?), so that this line of evidence is really rather confusing. (Actually, it is not impossible that the ultimate undertype on this last coin is the rare type Longuet 17 - this could make more sense, although the combination of two rare types in one coin might seem very unlikely - cf. Pl. 20,10 with Pl. 23,17J and Pl. 26,1 in ANSMN, which seem to show a "canopy" over the emperor on the undertype)**.

Given that this is a lightweight, flattish type, mid to later Andronicus III seems the most convincing attribution.

Longuet 21 - S.2364 (0.92g)

Note that in his "Byzantine Coins" Grierson equated S.2364 with S.2457, and he evidently carried this opinion over into DOC V. Now it is not unreasonable to surmise that the two types are related, with S.2364 being perhaps simply a variant of 2457, but in fact the hoard and fabric evidence suggests that S.2457 is much earlier than 2364 - cf. also the last section below for the dating of S.2457.

*  A third example (shown above) was sold on Ebay in Oct. 2008 (as S.2457). It weighed 0.92g, and was from the same reverse die as the other two examples (the obverse die seems to be different). Note the lack of stars below on the reverse. A fourth example, from different dies, and with 8 points on the star, was sold on Ebay in April 2012.

** Note that in 2007 Bendall  revised this overstrike by making the undertype S.2457 rather than 2364, but I suspect first thoughts are better here - the top type is clearly Longuet 14, not the earlier scyphate version.

D.O. 1190 (Longuet 22).

Finally, we come to DOC V, No. 1190. This interesting but little known later Palaeologan type has a nimbate figure on the obverse, and, on the reverse, an emperor kneeling (not crouching) below left with St Michael above him, and a nimbate figure (Christ?) to the right, handing a patriarchal cross to the emperor. This type is not listed in LPC, Sear or PCPC, but it is in fact the same type as Longuet 22 (although this was not recognised in DOC V). An excellent example of the type, only the fourth known to me at the time, was offered on the internet (in Oct. '04), and is shown below.

Longuet 22 - DO 1190 (1.36g).

Now, this new coin is important, because it suggests a solution to a problem with this type. In DOC V the type is assigned to John V, as an accession issue, basically on stylistic grounds. However, an example of the type appeared in BBS 117 (Lot 524), presumably as part of Bendall's hoard, suggesting an earlier date for the issue.

Fortunately, however, there is a possible solution. DO 1190 has obvious similarities to Longuet 1 and 2, which seem to be issues of Andronicus III, and also to S.2482, probably an issue of Andronicus II. Also, both the above coin and the Longuet example have the flattish field and turned up lip typical of the late Andronicus II/early Andronicus III types. Hence I think we can conclude that DO 1190 is probably an early Andronicus III or late Andronicus II issue.

Note that the new coin clearly confirms that the obverse figure is an Archangel (standing, apparently, although the design is possibly more complicated than that), rather than St Demetrius. Now this is odd, since it means St Michael is on both sides of the coin - but so be it. Also, to my mind, the right hand figure on the reverse could well be a military saint, but it's hard to be definite.

The above conclusions have been reinforced by another example of this type which was offered  on Ebay in June 2005, apparently as part of the 2004 finds of Andronicus II types (see "Some Palaeologan Finds").

Yet another important example of this type, again from the 2004 finds, was more recently (late 2008) offered through Ebay. This coin was overstruck on an example of S.2368 (see "Overstrikes and Mules"), which is rather surprising considering that S.2368 seems to date from c.1300, much earlier than DO1190 (cf. "Sequencing the Thessalonican Trachea"). Now such an overstrike is not out of the question, but it suggests that either DO1190 may be (somewhat) earlier than I have assumed here, or that S.2368 could be later than I have previously thought, or both.

P.S. Another example of DO1190 overstruck on what is probably S.2485b (Gr. 1468) has now (2012) come to light, confirming the relatively late date of 1190.


Actually, one further type mentioned at the beginning needs consideration, namely the non-Longuet "stamenon" S.2525, a type currently assigned to John V. This is a scarce type (although it appears regularly enough in internet sales) and its precise attribution has always been a problem. From its iconography and (apparent) find locations it is presumably a Byzantine (rather than Bulgarian) type, and an issue of Thessalonica rather than Constantinople, but beyond that little definite can be said.

Originally this type was assigned to Andronicus II by Gerasimov, and then to Andronicus III in LPC. In Sear and PCPC it is given to John V, although Bendall doesn't seem to be able to decide exactly where to put it (in PCPC it is assigned to the late 1360's, for instance, with coins of a quite different character). DOC V follows Bendall and gives it to John V.

The basic problem is that this type hasn't yet shown up in any hoards (the various recent specimens are clearly not part of the "2004" finds), so that we are reduced to arguments about style and fabric, where the evidence is uncertain and sometimes contradictory. Thus the rounded beard (or lack of it) on the emperor and the late style of the figure of the Saint would seem to rule out Andronicus II, and, while the iconography is, superficially at least, perhaps more redolent of John V than of Andronicus III, the fabric of the better examples (e .g, CNG 53, Lots 1951-52) is essentially that of the mid to late period Andronicus III types, with flat dies struck on flat or slightly scyphate flans.

In the end, the basic reason for assigning this type to John V is that it doesn't seem to fit anywhere else. Although not a common type, it is not really rare - it is commoner than many of the Longuet types, for example, and there were three examples at Turnovo, roughly the same as some of the  commoner Andronicus III types. Hence the fact that it did not appear in Longuet's hoard, or in any of the recent finds of Andronicus II types (including  the 2000 hoard), means that prima facie it is difficult to assign it to either Andronicus II or III.

If this type is an issue of John V, then there are three possible periods to which it might belong. Firstly, 1341-42, when John ruled in Thessalonica as a minor, with his mother as regent. Secondly, 1351-2, when he was allowed to rule Thessalonica briefly (initially in his own right and then jointly with his mother), before being moved on to the governership of Chalkidike in Thrace, and lastly perhaps 1365?-69, before the administration of Thessalonica was apparently handed over as an apanage to Manuel II. Now, the fact that on S.2525 the emperor is shown full-size, seemingly bearded and alone suggests that this type doesn't date from the 1341-42, although it must be said that there is no other type which can be convincingly assigned to this period, except perhaps for Longuet 2. Also, as John presumably was granted Thessalonica in 1351 only as apanage, it is unlikely that John VI would have allowed him to issue coins there in his own name in this period (although John VI apparently did allow his son Matthew to issues his own basilicon in similar circumstances a little later on). Furthermore, given that S.2525 is not particularly scarce, the fact that it didn't appear in the Pella or Serres hoards, and also that no post 1342 types from Thessalonica were found at Turnovo*, it would seem likely that S.2525 predates 1351 by some time.

Overall, to me it seems difficult to date this type from later than 1342, but as we have seen, the problem then is, why hasn't it turned up in any of the various finds from this period, particularly the Longuet hoard? For the present, then, the best we can say is that the attribution of this type remains uncertain, but I think we can restrict it, mainly on the basis of style and fabric, to either late Andronicus III or, less probably, given the representation of the ruler as fully grown, very early John V.**

(Note that the design elements of S.2525 seem to parallel the rare late stamenon type S.2592 in some particulars, and possibly this is the reason the two types are grouped together in PCPC as issues of John V from 1365-69. However, the general style of the two types is very different, since on S.2592 the emperor is nimbate, and wears the late skull cap type crown introduced not too long before the revolt of Andronicus IV in 1376, quite different from that on S.2525. Overall, S.2525 is clearly a much earlier type than 2592, which perhaps records, as Bendall suggests, the beginning of Manuel II's administration of Thessalonica in 1369).

* Although there was one example at Turnovo of the odd (imitative?) "assarion" type discussed by Bendall in N. Circ. '96, p.39 showing the Virgin? standing orans between two towers on the obverse and maybe John V and Anna on the reverse. Another example of this type was found in the excavations of Diokitiriou Square in Thessalonica. (Bendall has more recently suggested an attribution of this type to the Asen brothers in Thrace c.1356, which might explain its odd style and neat scyphate fabric, as well as its presence in Bulgaria).

** P.S. Bendall has noted (N. Circ. 2007) a possible overstrike of this type on a denier of John II Orsini. If correct, this might indicate that this type is an issue of Andronicus III, possibly following S.2486 and 2487, which were also overstruck on these deniers. Against this idea is the fact that this type did not show up in Longuet's or Bendall's hoards from this period, although perhaps this suggests that both of these hoards should be dated in the late 1330's, a few years before Andronicus III's death.

The fabric of the later types.

The later issues of Thessalonica can be arranged into three main groups according to fabric - firstly, types struck with flat or flattish dies on scyphate flans, such as S.2488, 2489, 2490 , 2496, Gr.1468 and DO 1190; secondly, slightly scyphate types, such as Longuet 1 & 2, and perhaps S.2500; and lastly, flat or flattish types (which can be conveniently described as stamena, although the term is much older) such as S.2484, 2487, 2497, 2499, 2501, 2525 and 2561 (and also S.2364 and probably 2495). Other types, such as S.2483, 2485 (Longuet 15) and perhaps some 2486, seem to be transitional types between the first and last groups (although some 2483's are quite heavy scyphate types - the one in the "Despot" collection for example weighed 2.82g - and the type may well be earlier). As well, S.2362 is a small flat type.

It is clear that there is a general progression here, from scyphate to flat, but the "slightly scyphate" types seem somewhat anomalous - do these types really constitute a separate group, and if so, where do they belong? On what we have said so far it seems that we have to assume that these types simply represent odd issues scattered throughout the general sequence. It must be said that this doesn't feel entirely satisfactory (can we really separate Longuet 1 and 2, for example, and where does S.2500 really belong?), and we shouldn't be surprised if we have to change our views on these types as further data comes to hand. On the other hand, perhaps we shouldn't expect the progression of coinage to always fit into nice regular schemes.

Other possible Andronicus III types.

One final point needs to be considered. It is clear that the production of coinage under Andronicus III was substantially less than under Andronicus II, since only a few of the former's types are even reasonably common. But as well, there is an obvious lack of common types clearly attributable to the preceding joint reign (and to the early 1320's for that matter). This raises the possibility that some of the coins usually attributed to earlier reigns might really belong to the 1320's.

Two obvious candidates are S. 2457 and 2458. These (basically anonymous) types are currently assigned to Andronicus II and Michael IX, but they both appear frequently in the 2000 hoard, and also in the other similar recent finds, and hence are clearly not early types, so that it's reasonable to consider the possibility that they may be issues of the joint reign of Andronicus II and III. On the other hand, the fact that only a single worn example of these two types (an S.2457) appears in the Longuet hoard militates against this idea, and in any case it is very likely, as noted earlier, that coin production was quite low throughout the 1320's (see Note "Bendall's Reconstructed Hoard of 2000").

Sear 2458 (2.05g).

It should also be noted S.2458 in particular is generally rather more scyphate than most late Andronicus II types (see above picture), a trait which it notably shares with S.2482. In fact it's likely that the better dies for these latter two types were cut by same hand (note the jewelled "sagions" without crossed shoulder straps, for example), which may imply that S .2458 is more closely related to S.2482 than to 2457. Hence it probably makes more sense to reassign S.2482 to Andronicus II, as suggested earlier, rather than to move S.2458 to the 1320's*.

In fact, given the low coin production in the 1320's, the most obvious candidates for this period are the scarcer Longuet types, which otherwise have nowhere to go, plus one or two other scarce types, such as the seated emperor version of S.2485 (Gr.1468) and perhaps the rare type S.2378 (although this may be just a variety of S.2370).

The two ruler type S.2465 is of course generally assumed to be an issue of the joint reign of Andronicus II and III, which on the face of it is a reasonable assumption, given that the full legend, "Andronikoc-Decpotic", does not name Michael (both Bendall and Grierson take the right hand inscription to read  "Andr.." or similar, but I have not seen a convincing example of this legend. (In DO 854 the right hand figure is senior and is probably St Demetrius). However, given that it is not a Longuet type, S.2465 seems rather too common for the 1320's - also, S.2465's are often relatively heavy (c.2.0g), and the full legend seems out of place for this period. Nonetheless, it might just be accommodated at the beginning of the joint reign (which began in 1322, not 1325), along with S.2490, with which, as we have seen, it is possibly associated.

The difficulty with S.2465 is that while it is apparently rather more common in general finds than S.2490, the latter type appears in Longuet's hoard but the former does not. It is possible that Longuet 24 is in fact an example of S.2465 - the obverse certainly fits, and the reverse seems to show two figures - but even so the general lack of S.2465's in Longuet's hoard is still a problem, particularly if we want to associate it with 2490.

In fact, a recently discovered (Sept. '06) overstrike of S.2465 on S.2377 or 2379 suggests that this type may belong in the 1310's, possibly as a coronation issue of Michael IX in 1216, although the fact that Michael is not named in the legend is perhaps a problem for this idea. Such a dating would of course create problems for S.2490, which on the one hand is possibly overstruck on 2465, but which on the other seems to belong in the 1320's. However, it is not impossible that S.2465 and 2490 both date from the mid to later 1310's, and that the survival of two S.2490's to the time of Longuet's hoard is just a statistical fact. (Note that S.2490, like S.2465, is a relatively heavy type, often weighing around 1.9 to 2 .0g - the two Longuet examples are the heaviest and 4th heaviest coins in that hoard).

Alternatively, we could possibly revert to Bendall's original idea that S.2490 was overstruck on S.2453. In fact, S.2453 might fit well into the joint reign of the 1320's - as opposed to S.2465, 2453 is quite rare, and, unlike the other early issues of Andronicus II and Michael IX, it doesn't appear to have much of a legend. However, I have my doubts about S.2453 - it seems to be rather too rare, and hasn't shown up in any of the recent finds - so maybe this is an early issue of Thessalonica after all, or more likely, judging from its crude style, an issue of some unofficial mint, perhaps in Asia Minor.

* P.S: An overstrike of S.2482 on S.2458 was offered on Ebay in March 2005, confirming the relation between these two types - although this does not of itself tell us much about their absolute dating (see Data Page "Overstrikes & Mules").


©Ross Glanfield

December 2004

Latest revisions: 

         7 Jan. '09:   Overstrike of DO 1190 on S.2368 noted. 
         8 June '09:   S.2561 noted (under S.2497-9). 
         2 Aug. '09:   Dating of S.2486 & 2487 to mid 1330's considered.  
       10 Oct. '09:  Dating of S.2497 before S.2486 & 2487 considered.  
         6 July '10: Discussion of S.2561 revised. 
         8 July '10: S.2493 and 2453 unofficial types? 
       28 Oct. '10: S.2377 overstike on 2482 noted.
       15 Feb. '13:  Probable overstrike of DO 1190 on S.2485b noted. 
       19 Mar. '15:  Possible overstrike of S.2485a on S.2485b noted. 
       29 Dec. '15:  Mirror image reverses of S.2489 and S.2493 noted.


[Home] [Articles] [Notes] [Data]