Introduction.

The hyperpyra of John V and John VI are rare, and as their design is essentially the same as those of earlier rulers their existence as separate types was only recognised in the 1960's.

The first definitive list of these types, including seven different sigla, was published by Simon Bendall in his survey of the Palaeologan hyperpyra in Revue Numismatique in 1984, and also in his "A Private Collection of Palaeologan Coins" (PCPC) in 1988. This list, with some additions, was adopted by Grierson in DOC V, increasing the list of (alleged) sigla to twelve.

In the late 1990's and early 2000's a number of John V and VI hyperpyra appeared on the market, apparently from a find which may have included a significant number of these types, amongst others. This added quite a few more possible sigla to the list of accepted varieties, and in the Numismatic Circular of 2004 Bendall published an expanded list of 23 sigla for these types. For the benefit of collectors who may not have access to this publication, and in order to enable discussion of certain problems arising from it, I have summarised the various sigla in the Table below (hoping I have not infringed copyright), together with two possible new sigla which have been promoted by some dealers in more recent years. Note that the table includes references to quite a number of examples which have appeared (or reappeared) since the publication of Bendall's 2004 article.

One obvious question that arises is how reliable are the attributions of the various John V and VI types in the first place, particularly given the fact that their legends, even when they are visible, are invariably garbled to a considerable degree. This is a not insignificant problem, but, while individual legends are generally not immediately readable, taken as a whole they reveal a legend, for each reverse figure, which can reasonably be read as "IW EN XW TW", and which is usually taken to mean "John, by the grace of Christ (Despot)" (cf. Appendix for more details). Furthermore, we find that the types currently attributed to John V and VI on the basis of these legends turn out, with only one real exception, to have different sigla from those of earlier reigns, which does suggest that they constitute a properly defined separate group of types, and hence that we can have a reasonable degree of confidence in the attributions as a whole.*

Nonetheless, it still seems likely that a number of the current attributions are, for various reasons, of doubtful validity. Some of the reported sigla may not exist, and it also seems likely that some the scarcer sigla are minor variations, or merely poorly produced versions, of other types. In the latter case, then, while these particular sigla still need to be listed, they can't be taken to represent separate issues of the hyperpyra.

In view of these difficulties I have tried assess the validity of the evidence supporting the various assignments in the Table below in order to produce a more reliable defining list of the sigla of John V and VI, and also to determine which of the sigla represent separate issues and which are merely the product of variant dies. As well, I have included an appendix with some hints on the identification of the John V and VI types, to help collectors judge for themselves the validity of offerings of these types.

*   This idea applies to earlier reigns as well. The general lack of duplication of specific sigla between the different reigns certainly tends to support the idea that the sigla are meant to define different issues not just within reigns but also across reigns. However, the precise significance of the combination of symbols in each siglum remains unclear - they don't seem to be regular enough to indicate dates of issue, or overall sequence numbers, so most likely they are used primarily as references for quality control purposes. Given that, for the gold coins at least, the sigla are mostly letters, it may be that they represent the names of the mint officials charged with the responsibility for a given issue, although against this is the fact that with the issues of Michael VIII the same sigla are often found on both the types with the gospels and the types with a scroll, which are known to belong to two separate periods. (For the silver coins, where quality control was generally not a problem, the sigla are mostly impersonal symbols, or the generic Palaeologan "B-B").

In any case, whatever the actual logic behind the sigla, we can probably assume that, for the hyperpyra at least, a specific combination of letters is used to define a basic issue, or at least (as is often confirmed by die linkages) a group of related issues, while the simpler symbols - dots and stars - are used for sub-issues within a group. (Occasionally these last symbols also seem to be used to distinguish similar sigla of different reigns).

Discussion.

The table below lists Bendall's 23 siglum types, from A to W, together with two possible new sigla which have appeared in recent years. The old siglum references in PCPC and DOC V are also given. The last column lists examples, together with my own comments on the various individual sigla*.

While many of the sigla seem to be well established, it will be seen that we have to have doubts about more than a few of them. To begin with we have Types A, I and Q. Now Grierson added these to Bendall's original list on the basis, he said, of advice from Bendall, but Bendall says he has no record of these types appearing anywhere, and I am not aware of any examples either. For the present I therefore feel inclined to conclude that Grierson has misunderstood Bendall and inadvertently added these extra types at some stage.

Type E does not exist as described. The coin Bendall relied on is actually shown upside down in the original sale catalog, and is presumably Type R or S.

The only published example of Type J is not at all clear, but for the moment the type is accepted as an issue of John V and VI.

As indicated earlier it's likely that not all the sigla listed by Bendall can be classed as separate issues. Certain of the scarcer sigla seem likely to be merely unintended variants of other basic types, due, as Bendall himself suggests, to poor die cutting. Thus Type K is probably a poorly cut C, while O may be variant of P, and so on. If so, then until we see examples of these sigla from more than one die they can't really be classed as separate types, representing separate issues.

Types L, M and N were not personally examined by Bendall and there is some doubt (in my mind at least) as to the accuracy of both the legends and the reported sigla (although the recently offered Type L/M example GM 181-2794 seems to have a fairly convincing part legend). Hence even if these are issues of John V and VI they may not all be distinct types, and in fact, given that a poorly cut Pi can look like a Gamma, it may even be that they are all one type, with siglum Pi - G'ma (or maybe Pi - G'maK(lig.) - the right hand symbol is still unclear). In any case, the present state of knowledge of these particular types is rather unsatisfactory.

The similarity of siglum P to No. 24 of Andronicus II and III raises a question as to the validity of this type. Clearly its validity (and that of No. 24) depends critically on the details of the designs and the validity the reading of the reverse legends; however, given that this type is part of Bendall's personal collection, we can presumably accept it as an issue of John V and VI.

Type T is a curious case - Bendall simply gives "Sear 2526" as his exemplar here, but the coin actually shown in Sear is surely Type H. 

Bendall gives no examples of Type V, and presumably relies on the reports of others. Examples that I have seen of this siglum show no evidence that they are issues of John V and VI, and in fact seem to be a known variant of the very similar No. 15 of Andronicus II and III.

Stylistically Type W seems out of place, and recalls some issues of earlier reigns, but similar sigla are found on some basilika of John VI, so that Bendall's assignment of this type can be accepted here. Nonetheless, it would be nice to know what the significance of the triangles of dots on this and many earlier types really is.

As for the two post Bendall types, these have little to recommend them as issues of John V and VI other than some debatable legends.

*  Note that examples of most of the types considered here can be seen on Coin Archives (search on "Hyper and 2526", and try Hiper as well), while some can be found on acsearch.com (search on "Hyper_* 2526"),  or in the CNG and Stack's archives. A couple of the more common types can also be found on Coin Archives misdescribed as Sear 2396 of Andronicus II and Michael IX.

(P.S. Unfortunately, as from July 2009 free access to Coin Archives is limited to the previous six months only, but you can still mount a full search for free on Ancient Coin Search at www.acsearch.info, although not as many vendors are available there as on Coin Archives).

Conclusion and disclaimer.

The current situation with the issues of John V and VI is not very satisfactory. Some of the older types are well established, but others are not, and some even appear to be imaginary. As regards Bendall's new types, the best we can say about them is that while we probably have some genuine new issues of John V and VI, their exact number and description are uncertain. There are two problems with these proposed new types: firstly, the legends on the examples I have seen (in photos) of these types are usually very incomplete, and in themselves rather less than convincing, although the fact that these coins apparently came from a late hoard does generally support their allocation to John V and VI. Secondly, even if they are actually issues of John V and VI, in most cases the sigla are not defined with any certainty, which makes it difficult to separate the various types into distinct issues, as opposed to mere variants.

Bearing all this mind, the situation can be summarised as follows:

Firstly, we have four well known basic types, namely C, D, H and R, which are relatively common, and known from a number of different dies. Probably related to these are a number of similar but scarcer types (B, F, G, K, and S), some of which are possibly just variant (or simply poorly engraved) versions of the basic types, and so may not represent separate issues. Except for the presumably accidental type K, all of these types were recognised in DOC V (although Type S - Grierson's No. 6 - was misdescribed).

Secondly, there are a number of rarer types, namely J, L, N, P, U, V, W, and perhaps the last two in the table, some of which seem to be known from only two or three examples (in some cases from one or two dies), or even just one coin. With the exception of Type U, and perhaps P and W, most of these can only be described as not very well defined, or not well established as issues of John V and VI, so that they should be regarded as provisional issues of the reign at best. Again we have some possible variants (M, O and T), and Types L, M and N may all be the same type (in which case we have six published examples from just three sets of dies). Of all these types, only O and U are listed in DOC V.

Finally, there are at least three types (A, I and Q) for which we seem to have no actual examples, plus Type E which appears to be a straight-out mistake.

Thus we see that, while Bendall lists 23 possible sigla overall, the number of confirmed sigla is rather less than that, and the probable number of separate issues less again. In fact, if we subtract the more questionable types (shown in italics in the Table) from Bendall's list, we are left with just 13 basic types (now 14, with Hv), and even this is perhaps still too generous a figure, given the uncertainty still surrounding some issues.

The statistics of these issues are interesting. One of Bendall's favourite theories about the Palaeologan hyperpyra in general is that they were issued on a regular basis, every three months or so (perhaps to meet the official payroll), a conclusion he reached by dividing the number sigla for each reign by the number of years of the reign, which yielded roughly the same figure in each case*. However, at least from the time of Andronicus II onwards it is clear that certain sigla, or groups of related sigla, are much commoner than the rest, so that it seems that the bulk of the coins for each reign derive from only a small number of large issues, while the rest derive from a large number of quite small issues, often involving only a handful of dies (or less, with quite few issues using, or even sharing, only one reverse die). This pattern is repeated with the issues of John V and VI, where most of the coins derive from only four basic types. This is not necessarily inconsistent with Bendall's hypothesis of regular issues, at least for the less common sigla, but it does make it clear that the picture may not have been as simple as he originally thought.

Having said all this I must admit that, not being a dealer, or a collector with unlimited funds, I am not a hands on expert in the John V and VI types, and claim no special expertise in the field. All I have done here is to review the published evidence as a collector who wants to be convinced of the reality (or otherwise) of the various issues, particularly the new types that have appeared in recent times. No doubt there is more evidence available and I would be glad to hear from anyone with information that can throw light on any of the problems noted here.

*  It should be noted that these calculations involve somewhat arbitrary assumptions concerning the periods over which some types were issued, whether to include reversed figure types (some of which are probably accidental variants), and which types (if any) should be assigned to Thessalonica, and also that Bendall originally took each siglum to indicate a separate issue, when it is clear that many sigla are related, and hence may have been issued together. Originally Bendall thought that new types may have been issued on a quarterly basis, but the steady increase in the number of discovered sigla for the various reigns makes this idea hard to maintain, although grouping related sigla together as single issues could perhaps save it.

Table:  Sigla of the Hyperpyra of John V and John VI.

*  Doubtful or accidental types are shown in italics.

Notes:

a:  As noted earlier, we seem to have no recorded examples of Types A, I and Q, and it may be that they do not exist. Type Q in particular seems to be a typo, as in DOC V it replaces Bendall's original Type 6 (which is now restored as type S). If any of these sigla do exist, it may be only as varieties of other basic issues.

a1:  Note that Sigla 21 and/or 22 of Andronicus II and III may be the basis of "Type A" here (Siglum 21 is probably just a poor version of Siglum 22 anyway). The sigla on the interesting coin CNGe72-121 are formally the same as Type A, but stylistically this coin (which even has the extra columns of dots of John V and VI types) seems to be a mirror image version of the "Avtokratores Romaion" type No. 153 of Andronicus II and Michael IX.

 b:  Known from examples from a number of different dies. The die statistics that we have are rather minimal, but suggest that these particular types derive from perhaps a dozen or so reverse dies each. Some of the other types listed here seem to be known from only one set of dies (or from a single example).

b1:  Same obverse die as CNG 63-1739.

c:  Type E is misdescribed by Bendall. The obverse of the CNG 53-1954 coin is actually shown upside down in the catalog, and is presumably an example of Type R or S.

d:  Note that in DOC V Grierson misreads Bendall's description of Type F - the E is actually on the bottom of the obverse with the Phi, not at the top - although it has to be said that on the only purported example of this type (F) that I know of (the Despot coin listed here) the symbols are not at all clear. Coin 1193 in DOC V was described as Type F in Bendall's 1984 article, but Bendall and DOC V now make this coin a Type H (as F, but with stars above).

e:  Neither the legend nor the stars above on the obverse are clear on the BBS 137 coin as pictured, and the latter are crucial, as without them this type could possibly be just a variant of the various common B-B (reversed) types of Andronicus II and Michael IX (some of which have the extra columns of dots of the John V and VI issues). However, this particular coin was seen by Bendall, so for the moment I have to accept it.

f:  Reserved.

g:  On the published examples of this type the right hand symbol is not very clear (on all except one it is obscured by a common die flaw), and it may well be that this type is simply a mirror image version of Type M/N, or that all these types are one and the same (see next note).

h:  Bendall gives no examples to support this type, although GM 181-2794 (offered in Sept. '09 with misread sigla) seems to be one, and since this coin is from the same dies as two of the Type L's (GM122-2394 etc.), it seems that this and the previous type are actually the same (with sigla Pi - G'ma, or maybe Pi - G'maK). 

i:  Known apparently only from BBS 136-27 (= BBS 133-34 = BBS 125-52), where the sigla (and particularly the dots above) are not very clear in the available photograph, and need to be confirmed. This coin seems to shares dies with the clipped Type M? coin CNGe221-550, so that M and N are possibly the same type. In any case the right hand siglum seems to be more than just a Gamma.

j:  Bendall gives only Lanz 106-815 as an example of this type. On this coin, which is cut in a rather crude style, the right hand letter is not very clear, and the left hand letter could be a reversed N; thus this coin is possibly a poorly cut Type P, or even a Type R. As well, the legend on this coin is hardly convincing.

k:  This type has the same siglum as No. 24 of Andronicus II and III. Presumably the two types differ in detail - either that, or at least one is misattributed. The siglum is also found on a reduced weight (i.e, later period) basilikon of Andronicus III and also on a hyperpyron of John V and Anna.

l:  Note that Type Q is Grierson's (misdescribed) Type 6 in DOC V, while Type S is Bendall's original Type 6.

m:   This siglum is the same as No. 163 of Andronicus II and Michael IX. However, the latter was quoted by Bendall from Gerasimov, who very likely was looking at a John V and VI coin but didn't realise it at the time (1960).

n:  Bendall gives "Sear 2526" as his exemplar here, but the Sear coin is actually Type H. It is possible that this type is a variant of Type S, but note that the N in this case is, according to Bendall's drawing, not reversed.

o:  Reserved.

p:  Bendall gives no examples of this type, which in fact has the same siglum as a relatively common known variant of No.15 of Andronicus II and III. On this variant the N is reversed, as on "Type V", but there are no obvious extra columns of dots, and both rulers are bearded; for an example cf. LHS Numismatics Sale 97, Lot 252. Note also that no other accepted siglum of a John V and VI type includes an "X". Until a clear example of "Type V" with the defining features of a John V and VI issue (e.g, a beardless John V) is produced I will consider the type as unlikely.

q:  Type W is based apparently on a single example seen at a coin fair. Stylistically it recalls earlier reigns, but similar sigla are found on some basilika of John VI. This coin was examined personally by Bendall and his assignment is accepted here.

r:   There is some doubt about this type. The siglum is found on coins showing two bearded emperors, which seem to be examples of the type originally noted in Num. Circ. in 1961 by Veglery and Zacos, and listed in Bendall and DOC V as Siglum 12 of Andronicus II and III. But it also appears on coins showing a bearded emperor with a beardless companion in the junior position. On these latter types traces of  what might be "CNXu" in the legends of the NAC 51 coin suggest that these are issues of John V and VI, but the N is not reversed. The extra columns of dots typical of John V and John VI appear on the two or three reverse dies that I know of, but they are not as well defined as on other issues of these rulers. Possibly these types are issues of Andronicus II and Michael IX, while Bendall thinks the the two figures may be Andronicus III and John V. (For another example of this type see CNGe36-64044(?), which can be found as the first entry in Wildwinds.com under S.2461).

s.  Another doubtful type. The symbols have no affinity with other sigla of John V and VI, and there are definitely no extra columns of dots. It's true that the legends on the Lanz coin could not unreasonably be read as those of John V and VI, but my guess is that they are actually a garbled version of "Avtokratores Romaion", as on the mirror image type No. 128 of Andronicus II and Michael IX. (For the Rauch coin search Coin Archives with "hyper and 2396 and 2,64"), or acsearch with  "hyperpyron 2396 "links th" "rechts B" ").

Appendix:  Recognising the John V and VI types.

The basic design of the hyperpyra of John V and VI is identical to that of the four tower issues of Andronicus II and Michael IX, and of Andronicus II and Andronicus III. Thus only the legends allow us to distinguish the various issues in the first instance, and hence establish the sigla for the different reigns. Unfortunately, the legends on the John V and VI types are invariably rather garbled, making the identification of these types an uncertain business at best, so that we should always be sceptical of new types, particularly when they are represented by only a single purported example.

The usual full legend (for each ruler) reads, when complete, something like "Iuu(l/lXuuIu" (note the reversed N), i.e, "IW EN XW TW", which is thought to stand for "IWANNIS EN XPISTW TW ThEW", meaning  "John, by the grace of Christ (Despot)". Often the "TW" is not included, and the legends are only partly visible, but if you can see at least "N(rev'd) X" you are off to good start (but note that parts of the legends on some earlier types can be misread as the John V and VI legend - for a good example of the legend see CNG 84-1645).

Other markers can help to confirm the hyperpyra of John V and VI. On all their confirmed types there are extra columns of dots (not always clear) between Christ and the two rulers, often, but not always, with only three or four dots in the column. It needs to be remembered however that the extra columns of dots also appear on several issues of earlier reigns, so that they are necessary but not sufficient markers of John V and VI.

Note also that all types with Star - Star above on the obverse are probably issues of John V and VI - sigla Nos. 183 & 184 of Andronicus II and Michael IX are miscopied by Grierson in DOC V and actually have the stars below (i.e, they are probably just variants of the common types 185-6), and No. 206 is possibly a messy version of some type of John V and VI. (As noted elsewhere, No. 163 is  presumably also John V and VI).

Normally at least one ruler, presumably the older John VI, appears with a full beard, while the other is beardless, or has only a short, rounded beard. In some types John VI appears on the right, on others on the left. There may, as Bendall suggests, be some historical significance in these permutations, but I am not in a position comment usefully on such ideas.

Ross Glanfield

April 2009.

Latest revisions:

25 Aug. '09:  Example of Type M (or L?) from Gorny & Mosch noted.
15 Oct. '09:  Example of Type M? from CNG noted.
  1 Apr. '10:  Variant type H added.
15 Aug. '10:  Comments on K/L'da - A type revised.