The Bulgarian & Latin Imitative Types

The so-called "Bulgarian Imitative" and "Latin Imitative" types of the 13th century have long been a source of confusion to collectors, and for that matter, professional numismatists. Three basic questions have still not been finally and completely answered, namely - who issued them, and when and where? This article seeks to summarise the current thinking on these problems for the ordinary collector. It is not meant to provide a fully detailed analysis of the various questions, but the reasons behind the key conclusions are given. (It also includes some ideas of my own, which I have generally indicated as such).

The "Bulgarian Imitatives".

The "Bulgarian Imitatives" are debased versions of certain late 12th cent. Byzantine billon trachea which circulated primarily in northern Greece and southern Bulgaria in the early decades of the 13th century. There are three types, normally denoted as Types A, B and C, which copy respectively the last trachy of Manuel I (Sear 1966), and the trachea of Isaac II and Alexius III. The imitative types are distinguished from the originals by generally poorer workmanship, and in the case of the type C in particular, a progressive decrease in size and weight over time. Unlike the originals, they were apparently not silver coated, and had a noticeably lower silver content (c.1.5% or less) than that of official types they copy (4.5-6% for the Manuel and earlier Isaac types, and 2-3.5% for the later issues of Isaac and then of Alexius). Also, they included a small amount of tin (up to c. 2%), which was not, as far as we know, used in the official 12th cent. types, and which was presumably included as a substitute for the missing silver. (The silver content figures quoted here and elsewhere in this article are taken from D.M. Metcalf's article "Silver and Tin in the Byzantine Trachy Coinages, ca. 1160-1261" in Revue belge de Numismatique 1977).

The term "Bulgarian Imitatives" was coined by Michael Hendy in 1969 in his groundbreaking book "Coinage and Money in the Byzantine Empire, 1081-1261". At one time these types were considered to be issues of a provincial Byzantine mint, but Hendy noted two key facts - firstly, the earlier hoards including these types are found particularly in southern Bulgaria, and secondly, these hoards invariably include examples of the official trachy of Alexius III, often with later types but sometimes without. He therefore concluded that these types were issued by the Bulgarian authorities some time after 1195 in southern Bulgaria, possibly from a mint in Verohe, or Veroja (i.e, Stara Zagora - not to be confused with Veroia in northern Greece).

This theory became the accepted view (at least in the west), despite alternative suggestions from other authors. Metcalf, for example, proposed that these types were "austerity" issues introduced initially by the Byzantine authorities as a result of their financial problems in the 1170's, and distributed later in Bulgaria by the Byzantine military during their various campaigns against the Bulgarians and Vlachs in the late 12th century. However, Metcalf's theory has the same difficulty as the old provincial mint idea, that is, he has to explain why the types imitating the trachea of Manuel and Isaac, which he assumes were issued during the reigns of those emperors, don't seem to appear in hoards (anywhere) before 1195.

On the other hand Hendy needs to explain why the Bulgarians, who had probably never issued coins before, who had no expertise in coin production, and whose treasury did not even rely on a cash economy, would suddenly begin issuing coins in huge numbers. Also, why were the mints apparently located in southern Bulgaria, rather than in Turnovo, and, most obviously, why would they have produced copies of Byzantine coins for use in Bulgaria?

Trade reasons could perhaps be advanced as a possible, if unconvincing, answer to both of these last two questions, but even if this were so, there is still a quite different problem with Hendy's theory. Clearly, the Bulgarians would have had to engage someone else to handle the actual coin production and management, most likely the Venetians, with whom they had close trade relations. Now the early "Bulgarian" types, although often poorly struck, are actually quite good copies of the Byzantine originals, done in what is generally a reasonably fine Greek style, and by celators who seemed to understand Greek. It therefore seems doubtful (to me at least) that the Venetians could have struck these types, since they were most likely also responsible for the "Latin Imitative" types (see below), and on most of those coins the crudity of their attempts to produce Greek legends is plain to see.

Clearly, both Hendy's and Metcalf's theories face difficulties, and today other ideas have been proposed. One theory, held by some Bulgarian numismatists, is that the "Bulgarian" types were first produced by the Greek magnates in Thrace in the earliest years of the 13th century. Exactly why is not clear, but perhaps it was because the supply of coinage from the central mint had dried up in the chaotic last years of Alexius III, and the Thracians needed cash to finance their military operations against the crusaders in northern Greece after the fall of Constantinople. Alternatively, given the large number of imitative types issued, it is quite possible that these coins were first struck some time before 1204, perhaps for operations against the Bulgarians. Anyway, whatever initially prompted the issue of the imitative coinage, the reason (on this theory) for the specific types actually minted now becomes clear - since the new coins were not official issues the only way the issuers could ensure their acceptance was to copy the commonest official types already in circulation*.

In any case if the Greeks' military operations were against the crusaders they didn't last too long - initially they may have been carried out in co-operation with the Bulgarians, who also feared the new Latin empire, but the Greeks soon came to an accommodation with the Latin rulers. The Bulgarian Czar Kalojan (Ivan I) then turned against the Greeks, and in 1206 he ravaged northern Thrace, seizing various cities, including, possibly, Adrianopolis. Kalojan was killed in 1207, but the Bulgarians retained control of much of northern Thrace.

After the Bulgarian takeover of Thrace, production of the imitative types seems to have continued. This was presumably now controlled by the Bulgarians, with the coins circulating widely in Greece and also in Bulgaria itself, and versions of these types, on progressively smaller modules, were evidently produced until the mid 1210's at least. These were all still of course Byzantine style coins, but evidently this didn't worry the Bulgarians, perhaps because the people, particularly in the south of the country, were familiar with official Byzantine coinage which had been introduced into the region in the later 12th century. However, just how much of a part all this coinage played in the local economy is unclear - it seems doubtful that there was an extensive cash economy in provincial Bulgaria at this time, so perhaps the later issues were used mostly for payments by the various armies (whether the locals welcomed this or not), just as the earlier official Byzantine coinage had been (more of this later).

It is noticeable, incidentally, that the later and smaller Bulgarian types differ markedly in both style and fabric from the earlier issues, so it is possible that these later issues were produced in mints further north in Bulgaria - traces of a mint issuing this style of coin have apparently been found in Turnovo itself, for instance.

(For the benefit of collectors, Appendix A below details the differences between the Bulgarian imitatives and their originals).

*  On the other hand, another theory on the origin of the "Bulgarian" types, held by some western scholars, is that these types were all produced in Constantinople itself, just before and/or after 1204. But if they were produced before 1204, why would the mint revert to producing copies of earlier types, and after 1204 it is difficult to see how the Bulgarian and Latin imitatives could both have been produced in Constantinople, given their overlapping time periods and differing styles.

The "Latin Imitatives".

The "Latin Imitatives", another term introduced by Hendy, are also debased "billon" trachea similar in style to the 12th Cent Byzantine types. However, unlike the "Bulgarian" imitatives, the Latin types are only generally based on the official originals - in fact the designs are clearly changed to avoid exact imitation of the imperial types - and also they typically, although not always, have crude or blundered legends. In the past, these types were usually assigned to various of the 12th Cent. emperors (see, for example, the well-known Ratto sale catalog of 1931), as many of them include the names of one or other of those rulers in their legends.

However, as with the "Bulgarian" imitatives, the earliest hoards including the Latin imitatives invariably contain official issues of Alexius III as well, and hence the Latin types also have to post date 1195. Since it is clear from the often garbled legends that many of the celators were not fluent in Greek, it seemed reasonable to assume, as Hendy did, that these types are issues of the crusader rulers in Constantinople and Thessalonica, to whom almost no coins had previously been assigned (except, at one time, for some of the anonymous folles). This conclusion is reinforced by the fact that the hoard evidence shows that the earliest Latin types postdate the Bulgarian imitatives.

Like all trachea of the earlier 13th Cent, the Latin types are basically copper with very small amounts of silver - most of which presumably derived from the recycling of predecessor types - plus small amounts of tin. (The earliest Latin issues, the full size Type A's of Constantinople, contained, according to Metcalf, a little over 1% of silver, but this figure quickly dropped to around 0.2% or less for most of the later issues - see below for more details).

Hendy initially assigned 20 of these types to Constantinople (as Types A to T), and three  to Thessalonica (as Types A, B and C). (In Sear the Hendy type is included in the plate references). Later, a number of other similar types (now listed as Types U, V and W of Constantinople) were added to the list of Latin types in DOC IV, as were two additional types, X and Y, of uncertain origin, which may in fact be Latin issues from Asia Minor (see the Note "Some Post Hendy Latin Types" for these various additional types, together with several more rare types published initially by Jordanov).

A feature of the Latin types is that while they come in normal ("large") module format, averaging a little over 3 gms for complete specimens, most of them are also found in small module versions averaging c.1.5 gms, although only the commoner SM versions are listed in Hendy and Sear.

While Hendy's attribution of these types to the Latin period has been accepted by all, the same can't be said of his assumption that they were all issued by the crusaders, and particularly of his specific assignments of the various types to Constantinople and Thessalonica. This is a complex question, so we will have to deal with it in stages.

The earliest Latin types (c.1207-8 - c.1220).

The hoard and overstrike evidence clearly shows that the earliest Latin issues were the Constantinople types A and B, together with the Thessalonican types A, B and C. Labeling the Thessalonican types with an apostrophe, the order of issue is A, A', then apparently B' before B, then C'. Their dates of issue can't be determined with exactitude, but the evidence shows that the Constantinople Type A dates from a few years after the Bulgarian imitatives, i.e, from the later 1200's, at around the same time as the first Nicean issue* of Theodore I Lascaris, while the Constantinople Type B dates from the same time as Theodore's second Nicean issue, probably some time in the mid to later 1210's.

On hoard evidence the Constantinople Type C clearly post-dates the Thessalonican C. It has usually been considered an early type, but it doesn't seem to appear, in either large or small form, in any hoards clearly dating before the mid 1220's, and hence for these reasons, and also on stylistic grounds, for the present this type is included with the later Latin types discussed below.

(For a more detailed analysis of the dating of all the Latin Types see the article "Sequencing the Latin Types").

* Unfortunately the date of this key type is uncertain - it is usually dated to 1208, when Theodore was actually crowned, but it might have been issued some time earlier, since he proclaimed himself emperor in 1206. Given Theodore's precarious hold on power in the early years, a later date for his first type seems more likely, and to me the general run of the hoard evidence also fits best with a later, rather than an earlier, date.

The source of the early Latin types.

The Latin issues were most likely not produced by the Latin emperors themselves, but rather by the Venetians, who controlled the commerce in Latin Greece as well as Bulgaria, and who were probably the only group amongst the crusaders who had the financial expertise necessary for coin production and management, given that the imperial mint of Constantinople had apparently been transferred to Nicea. That the Venetians did actually run the mint is confirmed by the well known agreement of 1219 between Theodore Lascaris in Nicea and the Venetian mayor of Constantinople banning the copying of each other's coinage. In fact, the "Latin" coinage would undoubtedly have been produced by the Venetians primarily to facilitate their own trade in the Balkans - the Latin rulers would simply have been customers of the Venetians' mint, like anybody else.

(This is probably why there are no coins in the names of the Latin emperors themselves - the weakness of the crusaders' claim to legitimacy, and their ongoing financial problems, meant that the Venetians, whose main interest was trade, were concerned only to produce coins acceptable in the region generally, i.e, coins in the familiar Byzantine style, and in the names of Byzantine rulers. Even in southern Greece, where crusader rule was more secure, it would be several decades before Frankish style coinage was introduced, and even more until the local barons were prepared to name themselves on their own coins).

What is certainly not clear is just where the early Latin types were actually minted. While hoard evidence generally supports Hendy's assignment of the Constantinople types, or some of them at least, to that city, his views on the "Thessalonican" types have been open to doubt. Given that Thessalonica was nominally a kingdom, it is not unreasonable to surmise that it might have issued its own coins, but in fact the "Thessalonican" issues are in the same style as the other Latin types, and their assignment to Thessalonica was based mainly on some quite minor stylistic features, which can hardly be considered conclusive.

The find evidence is somewhat confusing on this problem. For example, a hoard found at Yenimahalle, just north of Constantinople in European Turkey, includes all the early types in question here (and very little else) which strongly suggests that the "Thessalonican" types are really issues of Constantinople (Hendy suggests that this hoard is "probably" a combination two hoards, one of Constantinople types, and one of "Thessalonican" -  well, maybe). Also, an example of  Type O was found in the "Istanbul B" hoard, and this type is known overstruck on a Thessalonican Type C, suggesting that both these types are issues of Constantinople - if indeed this hoard came from somewhere near the capital, which is by no means certain. On the other hand the socalled "London 1966" hoard (N. Circ. 2005, p. 375-6), a hoard from the earlier 1220's which very likely originated not too far from Constantinople (quite possibly in Asia Minor), did not contain any "Thessalonican" types, in either large or small module, but by the same token it included only a handful of Constantinople types.

More generally, in hoards from Macedonia and Thrace there is (pace Touratsoglou) no evidence of a significant excess of Thessalonican over Constantinople types. On the other hand, stray finds from Chalkidike include Thessalonikan types (mostly LM Type A), with a few Bulgarian Type C imitatives and no imitative types (large or small) of Constantiple (Maladakis 2011). This is certainly prima facie suggestive of a mint in Thessalonica, but the numbers of coins from Chalkidike are relatively small, and hence their exact significance is uncertain, so that overall the question of the Thessalonican issues still can't be said to be finally settled. However, excavations of Dioikitiriou Square in Thessalonica itself have yielded a large number of coins from the 4th to the 14th centuries, and these may eventually throw some definitive light on the problem (although if history is any guide we can assume that we may have a long wait for them to be published)*.

It is perhaps also worth noting here that the overall frequencies of the Thessalonican types (as revealed by stray finds rather than hoards - see Appendix B) generally match those of the Constantinople type A, while those of Constantinople types B and C are rather lower. On the other hand the Thessalonican C is relatively scarce in hoards, but this presumably reflects the scarcity of hoards from the early 1220's to the 1240's.

*  It is interesting to note that the excavations at Pergamum in Asia Minor yielded Constantinople LM types A, B, G, O and P (one example of each), plus Thessalonican LM Type A (1 example) and LM Type B (2 examples), together with a dozen or so early SM types. Since for non Latin types the issues of Thessalonica are very rare at Pergamum as compared with those of Constantinople, these figures again perhaps suggest that the "Thessalonican" types A and B at least may not be issues of Thessalonica. Contrary to Hendy's dogmatic assertions in DOC IV, the case against the "Thessalonican" types is still very much open.

The early Small Module types.

All the early Latin types are commonly found in small module versions as well as large (cf. Sear 2044 et seq.). There has been, and still is, much dispute as to who issued these small module types, and why.

While the data is sketchy, the lack of small module types in hoards found near Constantinople seems to show that they did not generally circulate in the capital. This might seem to imply that these types were not produced in Constantinople, and in the past they have been attributed to the Bulgarians, or alternatively to various conjectural mints in central or southern Greece, such as Corinth.

However, there are difficulties with these ideas. To begin with, it seems likely that the later and smaller Bulgarian Type C imitative types were issued in parallel with the early small Latin types, which makes it unlikely that the Bulgarians were the issuers of the Latin types - after all, why would they issue the two series alongside each other? Hendy originally suggested that the small Latin types were produced, in a separate mint (by the Venetians presumably), to deliberately undermine the Bulgarian imitative currency, but other explanations for the origin of theses types can be envisaged.

To begin with, it is noticable the small module Latin types are accurate copies of the LM originals (down to the crude legends), and are stylistically indistinguishable from the large versions. As well, we see coins in which one side is struck with a small module die, and the other with a large module die. Finally, in the case of the Type A at least there is actually a continuous spectrum of module sizes between the large and small module versions. All of this can been taken to indicate the SM type A resulted from a progressive reduction in module size of the LM type, and hence that this type was produced in Constantinople, at least to begin with.

Some writers have assumed that the Bulgarian imitative types and the Latin Type A decreased in size in a parallel debasement of the currency, at least for the coinage used in the provinces. However, this idea is not without its own problems. It is often said that the production of the small module Type A began some time after that of the large module originals, but in fact the hoard evidence shows that if any such delay occurred it was not very great (cf. "Sequencing the Latin Types").

As well, the hoard evidence suggests that by the time the large Latin Type A was issued (or shortly after) the Bulgarian imitatives had already been reduced in size, for some reason, and therefore it seems more likely to me that the small Latin Type A's were issued soon after the large type specifically for use in areas where the Bulgarian imitatives were circulating, which was essentially what Hendy had originally suggested.

Anyway, whatever the actual sequence of events, the small module Type A soon became a common coin in Bulgaria and (to a lesser extent) in provincial Greece, but it was evidently not generally acceptable in Constantinople itself, where new LM types, including most likely the "Thessalonican" types, were issued (although the latter are also often found in varying reduced module formats, with Type B in particular being primarily found as a medium module coin). But the small module Type A's had clearly become entrenched in the currency, at least in Bulgaria, and small module counterparts of the other early LM types were produced in turn, probably also in Constantinople, although these, like the small Type A's, were presumably mainly intended for use in trade with Bulgaria.

(As with the small Type A it is often stated that the later SM types were issued a little after their LM prototypes, but in fact in the hoards the small and large types appear, in most cases, essentially simultaneously - again, cf.  "Sequencing the Latin Types").

In the end, whatever the reasoning behind the various small module types was, one thing we can say is that by c.1220 the production of the Bulgarian imitatives had evidently ceased, since by the later 1220's the small Latin types had largely replaced the Bulgarian imitatives in Bulgaria itself (although the latter types continued to circulate in small numbers until the end of the Latin era).

The value of the Small Module types - clipping.

The continuous spectrum of sizes between the small and large module Type A's means that it is reasonably certain that initially at least there weren't intended to be two separate denominations of the Latin types. But if there weren't two different official denominations, how were the small and large types exchanged in practice?  It is noticeable that up to about 1220 the various LM and SM types circulated together (outside Constantinople) without the LM types being clipped. So perhaps in Bulgaria and the Greek provinces at least there was an accepted relation between the two basic sizes - maybe one LM coin equalled two SM's, although the various intermediate sized Latin and Bulgarian types would still have been a problem. (Smaller types returning to the capital in trade would presumably have been bought by the mint at some discounted value, more or less as bullion).

However, one important fact should be kept in mind here, namely that up till the 1210's the bulk of the billon and bronze petty coinage in the provinces had been introduced by armies rather than traders, and since in the countryside people were largely self-sufficient most of this coinage probably found little use as small change (note also that imperial tetartera are scarce in Bulgaria). There was therefore probably a general oversupply of petty coinage in the Balkans, particularly outside the major trading towns, so that the main use for cash in the provinces was probably (as Metcalf suggests) as stores of wealth for ordinary families, in the ubiquitous hoards of the period.

This surplus of money may also have led to a debasement of the effective value of the coins, at least locally, so that until c.1220 the various Bulgarian, Latin and Nicean trachea may have been (it occurs to me) treated more as bullion than cash, so that when coins were actually used in the market, their value, and particularly the relative value of the various large, intermediate and small types, may well have simply been settled by bargaining, on the basis of weight, and hence there was no point in clipping the larger types.

Anyway, whatever may have been the case in the early Latin period, by the early 1220's the rising trade and wealth of Bulgaria, particularly under the new regime of Ivan Asen, would have produced a growing need for a standardised petty coinage in that country, so that eventually, as the number of the older LM coins in circulation dropped, the distinction between the large and small types evidently disappeared. This is evidenced by the appearance of systematic clipping, whereby the few remaining large coins (and any new LM types), were cut down to small module size. This started around the time of the introduction of the LM Type D, and clearly occurred primarily in Bulgaria, since LM coins found in and near Constantinople are generally found unclipped. It is noticable that in this period we see hoards in which the newer LM types are all or mostly clipped, while the older coins are still all unclipped. This confirms the idea that, in this period at least, hoards were mainly accumulative, i.e, they represent domestic savings, as suggested earlier.

The rise in the use of cash and the stability of the Bulgarian state under Ivan Asen are presumably also the reasons that there are relatively few hoards in that country from the early 1220's to c.1240.

The Small Module Nicean Types.

The early SM imitatives also included copies of the first Nicean coinage of Theodore I Lascaris of Nicea (Hendy's small module Type G), issued perhaps a little after the small Latin A. On the basis of some terms in the treaty of 1219 referred to earlier between Nicea and the Venetians (apparently confirming an earlier treaty of 1214), which banned the copying of each other's coins, it has usually been assumed that this imitative type was issued in Constantinople. However, given that this type was the only small module issue to appear in quantity in the "London 1966" hoard mentioned earlier, a hoard probably found in Asian Turkey, it is possible that the SM Type G was actually issued, as also noted earlier, in Nicea, so perhaps the Niceans followed the Latins down the path to smaller modules in the early 1210's.

As well, some less common SM copies of the later types of Lascaris, from both Nicea and Magnesia, are also found - these are generally in good style and appear to have been issued by the Niceans themselves, presumably for use in trade with Bulgaria (via the Danube) and perhaps also provincial Greece.

The mid and later period Latin types (c.1220 to mid 1240's).

The later Latin types raise even more questions than the early issues. To begin with, dating these types is quite difficult. Hoard and other evidence shows that certain types, such as the LM types D and O, followed not too long after the early types, in the early to mid 1220's. (The single examples of Type D reported in the early Asenovgrad and Pazardzhik hoards are presumably intruders, or perhaps misdescribed, as this type doesn't appear in any of the numerous other large hoards of the later 1210's). On the other hand other issues, such as the "Peter and Paul" types S and T, must be relatively late, as the hoards in which they appear date from the 1230's or later, since they include issues of Manuel, or more usually, John Comnenus-Ducas. Beyond this only tentative dating is possible, and in fact Hendy originally grouped the types between D and S simply on the basis of the nominal emperor's name, or, in the case of the anonymous types P, Q and R, the lack of it.

Nonetheless, on the basis of the hoard and other evidence I think that we can at least divide the later types into two broad groups as follows (cf. "Sequencing the Latin Types"):  

    Mid period types (c.1220 to mid 1220's): C, D, E/K, O, P, R.

    Late period types (later 1220's to early 1240's):  F, G, H, J, N, S, T, U, V, W.

The rare types L and M are probably fairly late issues, as the only finds in which they appear are two quite late hoards in north-east Bulgaria - L in the very late (post c.1250?) Dolna Kabda find, and M in Dolna Kabda and also the (mainly) post 1246? find at Krasen (near Russe).

The very rare Hendy Type I is apparently known only from a single unclipped LM example, so it is difficult to date. It has not appeared in a hoard, and the only known example is abnormally heavy at 5.75 gm - so it may not even be a Latin issue at all (see Note "The Latin Type I").

Type Q is also very rare, and, with one possible exception, does not appear in any of the numerous Bulgarian and Greek hoards of the Latin period, as would be expected if this was a Latin issue. This suggests that this (anonymous) type could actually be an issue of Epirus, an idea which is supported by both its rarity and also its similarity to S.2225, another anonymous type now generally regarded as an issue (by Theodore Comnenus-Ducas?) of Arta - see article "The Coins of Michael II of Epirus".

(Note however that two small module examples of Type Q are now said to be have been part of a small parcel of 16 coins in the Athens Numismatic Museum known as the "Thessalonika 1963" find (Hoard 116 at Athens), an odd group that was possibly part of the Veroia find from western Macedonia, an otherwise unremarkable hoard probably laid down in the 1210's. If so, this would hardly be consistent with an Epirote origin of this type. However, the two "Q" coins in question seem to have been listed as Latin Type C's of Constantinople in the original report on Thessalonika '63 (in Arch. Delt. 1964 B' p.12), and, given that this find has not been republished, the identification of the small "Q's" needs to be confirmed).

The rare Jordanov Type I appeared (in SM form) at Dolna Kabda and Krasen, while Jordanov Type II (possibly not a Latin type) appeared in the late Nisovo hoard in northern Bulgaria, and Type IV appeared at Dolna Kabda, suggesting that these are all late types. (For the Jordanov types, and also Hendy's additional types X and Y, see the Note "Some Post Hendy Latin Types").

(It is worth noting here that the later types can also be divided into two groups according to silver content. The mid period types types D, O and P show, according to Metcalf, a silver content of 0.25-0.4%, which is similar to that of the "Thessalonican" types of the 1210's, and roughly double that of the other later types (including E/K). This could be taken as indicating that these three types are the earliest of the later issues, possibly dating even before Constantinople Type C, although we have no silver data for the latter type. However, their silver content also generally exceeds that reported for the Constantinople Type B, which on hoard data would seem to be an earlier issue. It would nice to see some further silver figures for these types).

The end of the Latin Series.

The date of the end of the Latin series is somewhat uncertain. The production of new Latin style types seems to have essentially stopped by the 1240's, since, with the possible exception of a few very rare types (e.g, Jord. VII) all the known Latin types are found in hoards deposited in or before the late 1240's*. (It should be noted that only one or two of the numerous late Latin period hoards seem to postdate c.1250. None of the published hoards include any of the recognised types of Theodore II of Nicea, and other types tentatively attributed to Theodore in some of the hoards have usually turned out to be earlier issues, mostly of Theodore I or John III. On the other hand the coins of Theodore II are rare in Bulgaria anyway, but nonetheless only those hoards with heavily fragmented coins, like Dolna Kabda, can be much later than 1250).

Thus the hoard evidence shows that the Latin series essentially ends by the early 1240's at the latest, and if you assume, as some writers do, that these types were produced (by whoever) in, or for use in, Bulgaria, then it is likely that the series would not have survived the Mongol invasion of 1241-42 and the subsequent collapse of the Bulgarian empire. Also, even before 1240 the deteriorating economic position of the Latin rulers themselves meant that they too would would have been unlikely to have been able to fund new coinage in any quantity (it is said, for example, that in the late 1230's they were reduced to stripping the lead from the roofs in Constantinople to pay their debts).

*  Some of the later Bulgarian hoards probably date from the early 1240's, following the Mongol invasion of 1241. Others have been dated to 1246, or after, because they include one or other of the "Thessalonican" types of John III, most commonly examples of S.2134, Hendy's (original) Type K, a small module type which has generally been assumed to be the first issue of John in Thessalonica. However, finds of this type in Asia Minor (e.g, at Pergamum) suggest that Type K could have been issued at Magnesia, in which case the dating of the type would be uncertain. On the other hand, in the late hoards Type K's often appear in the very small Series III format, so that these types at least were presumably issued at Thessalonica, most probably in the transitional year 1246. For more details of the dating of the late hoards and types see the article  "Sequencing the Latin Types".

The source of the mid and later period types.

Even more controversial than the dating of the later Latin types is the question of who actually issued them, and where. As we shall see later, it has been sugested that the Venetians shifted production of the small Latin A's out of Constantinople some time after c.1220, and if so, did they move the whole mint, taking the production of the large module types with them, or were production facilities left in Constantinople to take care of the needs of the capital?

Metcalf for example has suggested that the types from D onward were minted (somewhere) in Bulgaria during the reign of Ivan II Asen (1218 -1241), rather than in Constantinople (an idea which Hendy describes in his typically charming fashion as "wretched nonsense"). Some Bulgarian numismatists on the other hand think that some at least of the later large types, and also the small module types, may have been issued by the Venetians in Dubrovnik (Ragusa), where they had their main depot for the Balkans, for use in trade with and through Bulgaria. (Ragusa had free trading rights throughout Bulgaria and the Byzantine empire, and Venice had gained control of the city in 1205, as a result of the 4th crusade).

We therefore need to consider the arguments for and against the various proposed sources in detail.

The case for (and against) Bulgaria.

The main argument for a Bulgarian provenance of the later Latin types is that they are found principally in Bulgaria, and only occasionally at or near Constantinople (or anywhere else in Greece, for that matter). This is indeed true, particularly as regards hoards, but it doesn't necessarily imply that the later types were produced in Bulgaria. To begin with, for various commercial and political reasons, Bulgarian finds, and particularly hoards, were usually reported and carefully recorded (at least until 1989 - almost 160 Latin period hoards found in Bulgaria and northern Greece are detailed by Jordanov, for example), while Turkish finds go largely unrecorded, or at least unpublished. Even more importantly, most of the later Latin types found in Bulgaria come from late hoards which were clearly laid down in the 1240's, perhaps to prevent the money being appropriated to finance tributary payments to the Mongols, who had invaded and subjugated Bulgaria in the early 1240's, or as a result of John III's military operations in northern Greece and southern Bulgaria in 1246-49, whereas there was generally no corresponding reason for the hoarding of coins in or near Constantinople. The apparent imbalance in finds is therefore easily explained. (A similar analysis applies, incidentally, to attempts to attribute the trachy of Theodore Mangaphas of Philadelphia, which is also mostly reported from Bulgaria, to a Bulgarian or Thracian ruler).

Thus the hoard evidence alone, while certainly suggestive, does not provide a conclusive reason for us to assume that the later Latin types were minted in Bulgaria, or anywhere else outside Constantinople.

However, there are other reasons for believing that some at least of the later Latin types might not have been struck in Constantinople. To begin with, we see that the Constantinople types from C onwards mostly show clear differences from the early types. In particular, on many of the later types the figures are significantly bigger and bolder than those on the earlier types - on the later types the figures are typically 18 - 21 mm high, while on the early types the (full length) figures generally range from only 15 to 18 mm (with some larger), possibly reflecting the decline in module size in the 1210's. It's certainly possible that this difference might reflect a change in the location of the mint, perhaps resulting from the treaty of 1219 between Theodore Lascaris and the Venetians in Constantinople referred to above, but on the other hand perhaps it just represents a reform in the practice of the mint after the earlier size reductions.

As well, certain of the late types certainly do seem to be out of place in Constantinople, at least at first sight. Peter and Paul are western rather than eastern saints, and almost never appear on Byzantine issues, and so Metcalf suggests that the "Peter and Paul" types S and T were struck by Ivan Asen as part of his short lived flirtation with the Vatican in 1237, during his campaigns against the Latins. This is certainly possible, particularly as these anonymous types don't show a nominal Greek ruler. But it's also conceivable that the Latins themselves could have struck these propaganda pieces, and for the same political reasons. (We also note that John III used the St Peter obverse of the Type S on one of his trachea {S.2133, i.e, the original Type J} and Hendy links this type, which he assumes is a Thessalonican issue, with the Latin coin, so as to date the latter to the later 1240's. However this dating seems unlikely on what has been said so far).

Overall, we see that while there are certainly suggestive reasons for assigning at least some of the later Latin types to Bulgaria, none of them are really conclusive. As regards Dubrovnik, there is some written evidence to suggest that Byzantine style coins were minted there at some stage (c.f. K. Dochev, "Coins and Coin Usage in Turnovo, XII - XIV C.", 1990, p. 30). But it is not clear that these types, known apparently as "Karouvi", or "Karouchi", were actually the Latin imitatives - it seems that they were issued on clipped flans, in which case they may simply have been local Ragusan copies of the clipped Latin types circulating in Bulgaria; and of course, if they actually were issued pre-clipped then they can hardly account for either the large or small module Latin types. Apart from these reports there is (as far as I know) no hard evidence to support the theory that Dubrovnik was the source of the later large module types. In any case, why would the Venetians issue LM types in Dubrovnik (or anywhere else for that matter) for use in Bulgaria, knowing that they would immediately be clipped down to small module size? (Hendy incidentally calls the Dubrovnik source idea "risible", but, typically, does not bother to explain why).

Finally, it would seem unlikely that Ivan Asen, particularly in the 1230's when he was at the height of his power, would allow the production of large module Greek style coins in Bulgaria, especially those showing a Latin, or more accurately a pseudo-Greek, emperor. By this time the LM types were obviously minted mainly for prestige reasons, so if Ivan had wanted to issue such types in Bulgaria he would most likely have done so in his own name, as he in fact did in Macedonia (at Thessalonica or perhaps Ochrid).

The case for Constantinople.

On the whole it has to be said that the idea that the crusaders' main mint ever left Constantinople is most unlikely, and in fact historical documents suggest that it was still located in the city when Michael VIII returned in 1261. We are therefore led back to the idea that the later Latin types, like the earlier, were in fact issues of Constantinople. It must be kept in mind that there is not a great deal of hard evidence positively linking these types to the capital, although what there is is significant. According to Metcalf stray examples of the various types (principally the common types D, O and P) do turn up in the area, if not in great numbers. As well, a hoard from Postallar in south-east Bulgaria, not too far from Constantinople, included types C, P and R. Also, as mentioned earlier, the Troad and London 1966 hoards, both possibly from Asia Minor, reportedly included single type E's, and a type O was found, as we have seen, in a hoard apparently found near the capital. 

In Constantinople itself Latin period finds are rare, but single examples of Constantinople LM Types A, B, C, D, E? and U, and Thessalonican Types A and C, were found at a church at Sarachane, along with a single SM Type A of Constantinople. More importantly, small numbers of the mainly mid period Constantinople Types C, D, F, N, O, P and R, and a Thessalonican Type C, were found at nearby Kalenderhane (in what was originally the church of Maria Kyriotissa, later a mosque), along with a handful of small module types (Const. A, C, and T)*. These coins are particularly important, since of the 19 large module coins found at Kalenderhane, only three were noted as clipped, and of the 6 LM types at Sarachane, only one (the Const. Type E) is obviously clipped, although one or two others seem a little light. Since virtually all the examples that we have of the later large module types (found mostly in Bulgaria) are clipped, the fact that the coins from Constantinople considered here are mostly unclipped confirms fairly conclusively the general idea that the mid-period types at least were minted in Constantinople rather than Bulgaria.

As regards the Types T and U the situation is less definite. Hendy lists the (single) Type T at Kalenderhane with the large module types, but its small size and weight, and the fact that it is not noted as clipped, suggests that it is actually a small module type. As well, the Type U at Sarachane might also be a SM example. Strictlly speaking, therefore, we can draw no definite conclusions as to whether these two types (in LM form) are issues of Constantinople, although this would seem quite likely the case.

(Both the Sarachane and Kalenderhane coins are evidently the result of stray losses, so the clipped coins presumably came as returns from Bulgaria. Note that LM Latin coins currently sold in the bazaar in Istanbul seem to be mostly clipped, as everywhere else, but it seems likely that these coins do not come from Istanbul itself, where coin finds are not common, but from the provincial regions of European Turkey, closer to Thrace and Bulgaria).

* M.F. Hendy in "Kalenderhane in Istanbul - The Excavations", pp. 175-276, Eds. C.L. Striker and D. Kuban, Philipp von Zabern, Mainz, 2007.

The Medium Module Types.

It has been noted by several writers that the later types don't really form a single homogeneous group. Certain types (like J, P and R, and perhaps some of the scarcer late types) are often struck with medium size dies on reduced modules, while others, notably Type G, are primarily medium sized types to begin with.

It is noticeable that the medium module coins are often found with little or no clipping in the later Bulgarian hoards, so that they must (I presume) have been accepted as equal in value to the small module types which dominate those hoards. Therefore it might be tempting to assume that the medium module types were produced in, or for use in, Bulgaria. However, there would seem to be little point in this, given the ongoing large scale production, in Bulgaria or elsewhere, of small module types which clearly were intended for that market (see next section for these last types). The medium module issues were therefore most likely produced in the same place as the LM versions, i.e, in Constantinople, and the reduced weights of these issues could simply reflect a shortage of bullion at various times.

The later Small Module types.

While it may be doubted that the later large module types were produced in Bulgaria, it is certainly tempting to think that after c.1220 some of the small module types were produced there, or perhaps in Dubrovnik, since small module issues (particularly the small Latin A) ultimately make up a large part of the coinage of Bulgaria in the later Latin period.

As we have seen, the early small module types were probably produced (by the Venetians) primarily for use in Bulgaria, most likely, at least to begin with, in Constantinople. But by the 1220's the Latin empire in north-eastern Greece was shrinking rapidly, and since in the end the Venetians were only interested in trade, they may have shifted their small module production to locations better suited to support their commercial operations in the southern Balkans. In fact, such a shift in production might have occurred quite early, perhaps, as suggested above, after the treaty of 1219, since the production of SM versions of the later LM Latin types (i.e, the LM types from D onwards) dropped drastically after c.1220. However, it not clear where the new small module mint (assuming there was one) was located - Dubrovnik is a popular candidate with some writers, although once again there seems to be no specific evidence to support this idea, while Veroja in southern Bulgaria is also possible, as is the Bulgarian controlled territory in northern Thrace (where the Bulgarian imitatives had probably been produced).

Personally, however, I think that while there may well have been a separate mint for the small module types, it was probably sited not too far from the official mint, i.e, it was most likely situated in Constantinople itself, and very possibly in the Venetian quarter, rather than the Latin ruler's palace.

In any case, wherever this mint was, the main small module coin of the 1220's and 1230's was clearly the ubiquitous SM version of the Constantinople Type A, which continued to be minted in huge quantities. Over 50% of reported stray finds of Latin types in Bulgaria are SM Type A's, and in late hoards they still make up 40% of all coins, so that this type was clearly the mainstay of the coinage of Bulgaria until the end of the Latin period. (It's possible that production of the other early small module types also continued for some time, as they also persist into the late hoards).

After 1220 SM versions of  most of the later Latin types (from LM Type D on) were also produced, although mostly only in small quantities, and the presence of these types raises the interesting question of why were they produced at all, given the ongoing large scale production of the small type A's. The fact that the production of the small A's continued for so long can only mean, I think, that the mint which produced them was not involved in the production of the other later Latin types, large or small, and hence restricted itself to the production of the early types with which it was familiar. This is consistent with the idea that the later small A's were specifically produced by the Venetians for use in Bulgaria, while the later large module types and their small module equivalents were produced for the Latin rulers in a different mint (presumably the official mint of Constantinople), with the small types being used by the Latins in local trade with Bulgaria. This scenario is confirmed, incidentally, by the fact that we sometimes see later types struck with a LM die on one side and a SM die on the other (as was also the case with the early period types).

It's worth noting that some of the later Latin types, notably T to W, are more common in small or smallish modules than large (as was the case with the earliest issues), suggesting that these types belong together, at or near the later end of the main Latin sequence. On the other hand the prominence in the later hoards of the small type T, which was evidently struck in considerable numbers, might be taken as evidence that this type at least might have been produced in (or at least for) Bulgaria, perhaps as a replacement for the long lived small type A. (For some further ideas on these types, see the Note "The Statistics of the Latin Type T").

The later Small Module issues of Thessalonica.

From the later1220's a wide variety of small module types from the Greek Empire of Thessalonica began to circulate widely in Bulgaria. These types often copy the larger types from the Thessalonica mint, and are done in basically the same fine style, and hence they are clearly official issues of that mint. They actually begin with copies of Theodore's first coinage, but don't become common until the reign of Manuel in the 1230's, when they are often struck on cut down large module types (presumably coins returning in trade from Bulgaria). Under both Manuel and John Comnenus-Ducas some small module types seem to have been issued on the old Latin SM weight scale (of c.1.5g), but most, particularly the Series III types of John C-D, are struck on very small flans, and on a noticeably lighter weight scale (c.0.8g overall) than the earlier small module types. 

(Note that the term "Series III" is also usually taken to include issues of John C-D under the Latin SM weight scale, and, to further confuse the issue, Bulgarian numismatists often list these last types as Series II issues, although they are clearly much lighter and smaller than Hendy's original idea of Series II).

The Series III types raise a number of problems. Given the large number of types in the series it has usually been assumed (in the west) that the series extended well into the following reigns, but if so it would seem unlikely that they were issued in Thessalonica, since if John Vatatzes had wanted to issue SM coins he would surely have done it his own name and with his own image on the coins. In fact, he seems to have essentially stopped the issue of SM types in Thessalonica soon after he retook the city in 1246, as except for the Type K's, none of Vatatzes' Thessalonican types were issued in small module version in any quantity. Hendy has therefore proposed that the Series III types were issued, not by the Greeks in Thessalonica, but by the Venetians, somewhere else, and that they extend into the 1250's. (For a further discussion of this theory, to which I personally don't subscribe, see the note "The Coins of John Comnenus-Ducas").

It is noticable that the average weight of the Series III types is not much more than half that of the old Latin SM issues; nevertheless it appears that the two classes of coins seem to have circulated together in Bulgaria without serious problems, at least until c.1250, since the Latin SM types are not generally found cut down to very small size before the late Dolna Kabda hoard. (Perhaps the overlapping of the the small Latin types and the Series III types made it impractical to distinguish between them, and as well in Bendall's "Thracian" hoard of the early 1240's we see that the small Latin types have been trimmed to an average of c.1.3g anyway).

SM versions of some of John Vatatzes' Magnesian types are also known, but they are generally scarce - they may have been produced by the Niceans before (and after?) 1246 for trade with Bulgaria and provincial Greece.

The Latin Tetartera.

One result of the debasement and decline of the trachy in the early 13th century was the demise of the tetarteron, which had been the basic petty coin in the Byzantine system, at least in southern Greece. The reduced value of the new small trachea, which now included virtually no added silver, presumably made the tetarteron redundant (and in fact the trachea may have been deliberately reduced to a value comparable with that of the tetarteron), so that in northern Greece at least the original official tetartera must have disappeared from circulation by the 1220's, although Theodore Ducas revived the type briefly in Thessalonica, and they were still issued in limited numbers by the Niceans in Asia Minor, where, apart perhaps from the SM Type G of John I Lascaris, the small trachea did not circulate in any quantity. (Note that the relatively low value Byzantine tetartera were rarely included in the numerous hoards of the late 12th and early 13th centuries in Greece and Bulgaria, which consist almost entirely of trachea. The official 12th century tetartera seem to have circulated mainly in southern Greece, where they are common as stray finds).

The Latins did in fact produce some imitative tetartera, or perhaps half-tetartera, in northern Greece, but only in very small quantities. The dates and place of origin of these types are only vaguely known - Hendy attributed two of them (Types A and B in DOC IV) to Thessalonica, although only on stylistic grounds, and Grierson gave another, whose reverse copies that of the Type O trachy, to Constantinople. Although these types are rare, they are still a matter of some theoretical interest. It might be natural to assume that they most likely belong in the early period, and in fact the Hendy Type B tetarteron is iconically connected to the "Thessalonican" Type C trachy (a relatively late early type), but the Grierson type mimics what is probably a mid period trachy, and there was reportedly an example of the Hendy Type A tetarteron in the late Nisovo hoard (of trachea), implying that this is a relatively late issue (and possibly that it was accepted as equivalent to a small trachy).

Thus, although there was little need after c.1210 for a separate tetarteron, at least in the northern Latin and Nicean territories, they were obviously still produced occasionally, perhaps as a local alternative to the small (and clipped) trachea which circulated elsewhere. If so, then probably they were produced for use only in and around Constantinople, since in later times there was no need for them anywhere else, and we note that they hardly ever turn up in Bulgaria anyway, even as stray finds.

In southern Greece tetartera were always the mainstay of the coinage, and socalled imitative tetartera are in fact reasonably common, appearing as poorer quality ("barbarous") versions of various official 12th century tetartera. However, it is not always clear which types are official issues and which are local copies. Thus the smaller module "official" types, such as S.1931 and 1932 of Alexius I, and S.1979 of Manuel I, are often described in the literature as products of  an "uncertain Greek mint" (or even of mints outside Greece). For an example of a type which has been referred to (by Papadopoulou, for example*) as an imitative tetarteron see the coin pictured below - note the octagonal flan considered typical of the "imitative" types.

Imitative? tetarteron copying S.1931 of Alexius I. (1.16g).

It has been suggested that these "barbarous" copies, or some of them, are actually Latin issues of southern Greece, perhaps from a mint in Corinth. However, it seems unlikely that the particular tetartera considered by Papadopolou, like the one shown above, are 13th century issues, given that they copy a type of Alexius I from the early 12th century.

On the other hand the Argos 1988 hoard**, which dates from c.1260 or later, contained a variety of small imitative tetartera which clearly date from the early to mid 13th century, and which can therefore reasonably be classed as Latin issues. These coins are also based generally on earlier types, notably issues of Manuel I, and like Papadopolou's types are also generally octogonal, although they are smaller and a lot cruder in execution than the earlier "imitative" issues. These are obviously local issues, but their place in the monetary system of the Peloponnese is not clear, since their scarcity and low value means they can hardly have constituted a significant part of the coinage. (These types should not be confused with small octagonal "trachies" apparently produced as emergency coinage in Bulgaria in the 1250's by clipping and bashing old recycled types, including old tetartera).

*  P. Papadopoulou, Tetartera d'imitation du XIIIe siecle: a propos du tresor de Durres (Albanie), Revue numismatique, 2005, p. 145-162, Pl. XIII-XIV (available on the Persee website at www.persee.fr).

** J. Baker, Two Thirteenth-Century Hoards and some Site Finds from Argos, Num. Chron. 2007, p. 311-33.

The end of the Latin period coinage.

In 1246 John III re-established control in Thessalonica and northern Thrace, reducing Latin Constantinople to a tiny rump. As we have just seen, he soon took steps to reform the currency in Thessalonica by largely ending the production of the small module coins there, and it is clear from early Palaeologan hoards that in the parts of Greece under his control John quickly replaced the degenerate coinage of the Latin period with his own issues. (One result of this is that post 1261 Greek hoards are of little interest to the study of the Latin coinage).

As we have also noted, the Latin series proper effectively terminates in the early 1240's, but it is possible that a few new Latin types were issued after 1241, presumably in Constantinople - but if so, it is clear that they were issued only in very small quantities.

In Bulgaria political and economic conditions deteriorated rapidly in the 1240's as a result of the Mongol invasions of 1241-2, and few new coins seem to have been added to the circulating currency in the 1240's. By the 1250's the lack of new official coins led to a severe crisis in the petty currency, as debasement of the old coins, by clipping and cutting them up into increasingly smaller pieces, became endemic. In the very late Dolna Kabda hoard for example the small module Latin types are mostly halved or clipped down to Series III size, or less, and even the very small Series III types themselves are often clipped. (To this period no doubt also belong various small local emergency issues, such as clipped examples of old Byzantine tetartera, sometimes bashed into a roughly scyphate shape).

Clearly the authorities in Bulgaria lost control of the coinage in the 1250's, and we can't even guess at the effective value of these tiny fragments, or their relation to real value coins like the hyperperon - presumably they were eventually treated simply as bullion. In the end the currency problem in Bulgaria was relieved, firstly, by the use of the coins of Michael VIII, although to begin with these were often clipped down just as the earlier types had been, and then, in northern Bulgaria at least, by the introduction of a regular official Bulgarian coinage in the shape of the large module trachea of Mitso Asen and Constantine Tikh Asen (which are generally, although by no means always, found unclipped).

It therefore appears likely that the deteriorating conditions in Bulgaria ended the Venetians' interest in the country, and hence their production of coins for it, in the early 1240's. 

In those areas of Greece which ultimately remained under crusader control (Central Greece and the Peloponnese) the later (post c.1220) Latin types are rarely found, and the main coinage for the earlier part of the 13th century was apparently a mixture of old imperial era tetartera with some cut down trachies and a few locally made imitative tetartera. Eventually, this Byzantine style coinage was replaced, in the middle decades of the 13th century, by imported, and, eventually locally made, Frankish style deniers.

The silver content of the Trachy in the Crusader period.

At this point it is worth reviewing the decline of the silver content in the trachy from the late 12th century onwards, using the data given by Metcalf in Revue belge de Numismatique 1977 . From these figures is noticeable that, provided we date the Bulgarian imitatives to c.1204 and the years immdiately after, then we see a steady decline in the silver content in the trachy everywhere, from 3-3.5% in Isaac II's later issues to 2-3% in those of Alexius III, to 1-2%  in the earlier Bulgarian imitatives, to around 1.2% in the large Latin Type A of Constantinople and the (presumably) earliest issues of Theodore I's first Nicean type, to 0.3-0.4% in the "Thessalonican" issues of the 1210's, and then down to 0.2% or less in most types from c.1220. (The anomalously high silver content in the Latin types D, O and P has been noted above).

This suggests, to me at any rate, that the Bulgarian imitatives were not, as has been generally assumed in the past, deliberately issued at a lower standard, in terms of both size and silver content, than the original "official" types (including, in particular, the supposedly contemporary trachea of Alexius III), but rather that they, and most of the subsequent coinages , were simply issued at the prevailing standard of the day. On this view, the decline in the silver content of the trachy simply reflected an increasing lack of silver bullion, and in particular that the silver in the post 1204 types was largely a steadily declining residue from the recycling of the earlier issues, with tin being added to try to make up for the lack of silver (so that issues after c.1220 are better described as bronze than billon). The tin is an important marker of the imitatives, since as far as we know it was not included in the official issues of the 12th cent. - however, it is not inconceivable that the later issues of Alexius III, which are often very similar in workmanship to the "Bulgarian" versions, included tin, and if so they may be very difficult to distinguish from the imitatives by any objective test.

Nominal attempts were occasionally made to restore the real value of the "billon" coins - for example, the silver content of the large module coins of the Greek Empire of Thessalonica was raised to c.0.5% under Manuel for a while, although the figure for John Ducas's small Series III types, or at least some of them, was 0.2% or less. Some of John Vatatzes' Thessalonican issues also had a slightly higher silver content of 0.3-0.4%, but after that we have, as far as I know, little data on the composition of the trachy.

It should also be noted that the silver coating of the official 12th cent. Byzantine trachea is not apparently found on any post 1204 issues.

The "Latin" hyperpyra

In the immediate aftermath of the Crusader takeover gold coinage was not produced in Greece, but production was resumed by John III Vatatzes in Nicaea in the 1220's. It has long been been suggested that some of the hyperpyra attributed to John Vatatzes (Pegolotti's "Perperi Latini") may have been produced by the Latins (i.e, the Venetians) in Constantinople.

Recent studies by Oberlander-Tarnoveanu have lent some support to this idea. As described by this author, the general style of the "imitative" types is loose; Christ's nimbus is usually drawn with a continuous instead of a dotted line (although it can sometimes be finely dotted, which sounds suspicious), and the arms of the cross behind Christ's head are generally drawn with splayed, rather than parallel, lines. The most common sigla on these types seem to be one dot, or, less often, three dots. (The one dot siglum is also common on Nicaean issues of the hyperpyron).

However it's noticeable that Oberlander-Tarnoveanu's "imitative" types don't generally correspond with Pegolotti's, which are described as having sigla of 3 or 4 dots, and Lianta for one is not convinced that these types are indeed Latin issues (Num. Chron. 2006). (Whether Pegolotti's candidates are any more convincing is of course yet another question, as there are problems with his allocations as well).

For an example of a "Latin" hyperpyron on Oberlander-Tarnoveanu's definition see e.g. CNGe325-765:


Conclusion and caveat.

Overall, while the above summary seems to provide a reasonably coherent picture of the petty coinage of the Latin era, it is noticeable that it involves more than a few "probably's" and "presumably's", and that certain key problems have only been partly resolved. While the publication of the Kalenderhane material now (2007) makes it clear that the mid period Latin types were produced in Constantinople, it still can't be definitely stated that all the late types were produced there as well, although this would certainly seem likely. Many of our current ideas on this period are therefore still only provisional, and hence we should not be surprised if some of these ideas are overturned by future discoveries.

Ross Glanfield

August 2006.

Appendix A:   Distinguishing the Bulgarian Imitatives.

Distinguishing the Bulgarian Imitatives from the official originals is not always easy, and even the experts don't always agree on the question. Sometimes the two series can be distinguished by specific details of the design, but in many cases the designs are essentially identical, so that they can often only be separated by their general appearance. As a general rule the imitatives are of lighter weight than the official types, averaging about 2.8-2.9 gm for Types A and B (about 0.8 gm less than the originals), while for Type C the weight starts out about the same as for the other types, but then declines steadily, so that the later issues are on the same weight scale as the earlier small Latin trachea, i.e, around 1.6 gm on average. However there is still a considerable overlap of the weights of the official and imitative series, so that often we are forced back to differences in style. Often the imitatives are poorly struck on thin flans, which are often split, although issues on thicker, crudely cut flans are also found. However, well struck imitatives are not uncommon, and the actual design itself is often quite well done, in a reasonably good style.

Further details of the various types are given below.

Type A. (as Manuel I's 4th trachy).

    In the commonest version of  Type A the obverse copies that of Hendy's 3rd variety (Var. C) of Manuel's 4th trachy (S.1966), i.e, it has stars each side of Christ, above the arms of the throne, while the reverse copies that of the original variety A(b), with three dots (jewels) on the chest panel and 6 or more dots on the collar piece. This combination is not found in the original issues. However, according to Joppich some imitations have only 3 dots on the collar, and not all have obverse stars. Coins with one or more square jewels, or large circled jewels ("crystals" in Joppich), on the emperor's clothing are generally assumed to be official types.

Type B. (as Isaac's trachy).

    Here the imitative type normally copies Hendy's 2nd variety of Isaac's official trachy, with a waistband with just five dots, and various numbers of dots on the collar and chestpiece. Also, at least in the opinion of Joppich, some examples of the 1st official variety, with St Andrew's cross style diagonals in the waistband as well as the dots, are copies. It seems there are no  imitative versions of Hendy's 3rd official variety, where there is a star or dot below the emperor's right arm, or of the 4th, which has large jewels on his clothing.

    However, since the details of some of the imitatives are identical with those on the official types it is often very difficult to say whether some examples, particularly smaller coins done in neat style, are imitatives, as opposed to (presumably later) issues of the official types; and in fact in my opinion the matter probably can't really be decided, in many cases, without determining the silver, or more crucially the tin, content of the coins.

Type C. (as Alexius III's trachy).

    On the original type the waistbands of the two reverse figures are different, and the right hand figure is nimbate. On the copies the two figures generally have the same waistband, and the right hand figure is often not nimbate, particularly on the later and smaller issues. On the official types we find three different waistbands on the left-hand figure, and just one on the right-hand figure, as shown below - various other waistband designs are found on the imitative types.

    Official type waistbands:

                   Left-hand figure.                    Right-hand figure.

    On the original types large jewels are also sometimes found elsewhere on the emperor's clothing, but this never seems to occur with the imitatives. On the other hand, it should be noted that the circled jewel above the reverse figures on most original types is also found on some imitative types. (This device often appears as a single dot on the imitatives, and also, perhaps, on some original types). As well, the obverse legend "+KERO-HqEI" is often missing on the smaller imitatives, but it doesn't appear on some of the official issues either. 

    So far so good, but unfortunately there is a complication, since it is generally thought that some of the issues with different waistbands are in fact imitatives (notably Jordanov's variety A, probably the earliest common variety of the imitative type, which copies Hendy's official variety B, with the 2nd of the left hand waistbands shown above, and the standard 6 dot right hand waistband). Once again, only the silver and tin content can really tell the later official types from the early imitatives, and even this may not really decide the issue, since, as suggested earlier, it's possible that the later official types simply shade smoothly into the imitatives, in terms of both size and composition. It is clear that a comprehensive study of the composition of examples of these various types from a sequence of hoards is needed if these issues are ever going to be properly sorted out.

One implication of all this is that the identification of the imitatives, particularly Types B and C, is to some extent a matter of personal opinion, so that in hoard reports the figures for these types, and also for the corresponding original types, should be treated with some reservation, especially in cases where one particular type (most often the Imitative Type B) is supposedly missing or under-represented.

For more details of the differences between the official and imitative types see Ivan Jordanov's "Coins and Coin Circulation in Bulgaria in the Middle Ages 1081-1261" (Sofia 1984) and Stefan Joppich's "Die Billon-Skyphaten der Komnenen und Angeloi 1092-1204" (Petzlaff 2005). Note that Joppich assigns more varieties to the imitatives than Hendy or Jordanov, and given the odd statistics of many of the reported hoards (as noted in the previous paragraph), he may well be closer to the complex reality of these types.

Appendix B:    The statistics of the Latin Issues.

Given that the large module Type A of Constantinople was produced when the Latin empire was at its peak, one might be excused for thinking that this type was one of the commonest of the LM Latin issues, and this idea would seem to be confirmed by the fact that Type A is easily the commonest large Latin type in the numerous hoards of the early 13th C. However, this initial impression is not borne out, at least at first sight, by the evidence of stray finds from the period, which should, in principle at least, give us a better idea than the hoards of the relative frequencies of the various Latin types overall. Averaging the figures for all the stray finds documented by Jordanov, the large A's constitute only about 5% of all the LM coins from the issues assigned to Constantinople, whereas the mid period types D, O, P and R make up 16, 7, 15 and 11% respectively*. Also, the figure for the late(?) Type N in Jordanov's stray finds overall is 19%, a surprisingly large value for what is seemingly a post 1230 type, although it may be biassed by the relatively large contribution from the finds at Isakcha on the lower Danube (i.e, Isaccea in Rumania - note that Type N is also common in some, but not all, of the late hoards of the Danube valley, notably Krasen, Nisovo and Dolna Kabda).

The figure for the Type A's here looks surprisingly low, but it does not include the "medium" sized coins reported for Turnovo in Jordanov, which could add as much as 7% to the total for the Type A's. Some of these medium sized coins are presumably transitional issues between the early large and small module types, and some of these at least would probably have circulated in the capital, and hence should be included in the Type A figures, while others are perhaps larger issues of the later small module A's, which were produced for use in Bulgaria , and were possibly not even struck in Constantinople. Thus the totals for the large Type A's should evidently be increased, although it is hard to quantify by how much.

This general picture is confirmed by Dochev's later figures for Turnovo alone, which, from a much larger sample, show an even lower figure for the large Type A's of  less than 3% of the Constantinople coins (with no specifically "medium" types, which here may be mostly included in the SM figures), as against 24, 7, 15 and 10% for D, O, P and R, while the figure for Type N is 11%, rather lower than before, but still a fairly high value. For all other types the frequencies are not more than about 4%, except for Type F at around 6%. For the three "Thessalonican" types the frequencies at Turnovo are all generally similar to that of the Constantinople A, although here again it's possible that some medium module issues have not been included in the LM figures. (Note that Dochev also doesn't always clearly separate the large, medium and small versions of the later types, but I have tried to allow for this).

What do all these figures tell us? Well, if we assume that all the large module Latin types were produced at Constantinople, then we might have expected that as the fortunes of Constantinople declined, then the production of coins would also have decreased. Ultimately, this seems to have been the case, at least after c.1230, since except for Type N (which may in fact be a mid period type), the production of the late LM types seems to have been quite limited.

On the other hand, the relatively low percentages of large Constantinople Type A's (and the other early types) in the stray finds are harder to explain. Since the stray finds invoked here are all in Bulgaria or Rumania, this might be taken as support for the idea that while Type A was struck in Constantinople, all the later types from D onward were produced in Bulgaria, as Metcalf suggested.

However, significant as the figures seem to be, we certainly musn't rush to judgement here. Firstly, as stated earlier the figures for the type A's may be significantly underestimated because they don't include medium module issues (as is possibly also true for the "Thessalonican" types). Secondly, if we going to draw any conclusions from the stray finds about the possible sources of the various types, then we have to make two fairly specific basic assumptions, namely that the various finds give us valid statistical samples of the coinage circulating in given areas of Bulgaria over the whole of the Latin period, and also that the circulating coinage in those various regions reflects accurately the relative levels of production of the various types in the first place, again over the whole Latin period. Both of these assumptions are obviously open to doubt, particularly given the drastic changes in the military and economic circumstances in the region over the Latin period, which must have repeatedly disrupted trade, as noted earlier. Nonetheless, the non-uniformity of the frequencies of the various types in the stray finds is certainly very striking (particularly the high frequencies of most of the mid period types, which is reflected in the market place), and is worth keeping in mind when considering the question of their origin.

*  The figures for Type D in this section are likely inflated to some extent by the inclusion of Type E/K's, which often can't be distinguished from Type D when clipped.

Latest revisions:

     1 Aug. '07:   Discussion of Types I and Q revised. 
     2 Aug. '07:   P.S. on the Latin types found at Pergamum added to section "The earliest Latin types". 
    16 Aug. '07: Latin Type I as an issue of Andronicus I Gidon of Trebizond.  
    24 Aug. '07: Treatment of Series III types revised.  
    28 Nov. '07: Type Q tentatively assigned to Epirus.
      1 Dec. '07:  "The Small Module Nicean Types" revised.
    17 Dec. '07: "The early Small Module types" revised, yet again. ("When the facts don't fit my theory, I change my theory" - J.M. Keynes).
    20 Dec. '07: Clipping discussion revised.
    16 Jan. '08: Type Q discussion revised again.  
      7 Feb. '08:  Kalenderhane coins taken into account. 
    13 Feb. '08: Some corrections to the Kalenderhane discussion.  
      5 Dec. '08:  Discussion of "Bulgarian" types revised. 
     17 July '09:  Southern Latin tetartera theory noted.
      5 Apr. '10:  Argos 1988 hoard noted. 
      4 Nov. '15:  "The "Latin" Hyperpyra" revised.
    14 Mar. '16:   Chalkidike stray finds noted.

[Home] [Articles] [Notes] [Data]