А COIN OF BOHEMOND I OF ANTIOCH OVERSTRUCK BY TANCRED

After the city of Antioch was captured by the crusaders in 1098 Bohemond of Tarentum declared himself Prince of Antioch. In the summer of 1099 he was captured by the Danishmendids and remained a prisoner for four years. During his absence Antioch was ruled by his nephew Tancred. In 1104 Bohemond left the Holy Land and never returned. Until his death in 1111 he titled himself Prince of Antioch and was succeeded by his infant son, also called Bohemond. Tancred (1104–12) and his successor Roger (1112–19) also used the title  ̳Prince‘ and obviously considered themselves more than just simply  ̳regents‘ for Bohemond I or II. The coinage attributed to Bohemond I, prince of Antioch (1098–1111) consists of a single type in copper in Byzantine style. It depicts a bust of St Peter on the obverse and a floreate cross, with the letters B H M T in the angles, on the reverse. Ever since it was first attributed to Bohemond I by de Saulcy in 1847 it has been generally accepted that this type is a coin of Bohemond I and not Bohemond II (1119–30). The copper coins of the princes of Antioch were usually overstruck on preceding types and until now there has been no sign of a coin of Bohemond‘s successors, the  ̳regents‘ Tancred, Roger and Bohemond II, overstruck on a coin of Bohemond I. The article publishes a clear example of a type 3 coin of Tancred overstruck on a coin of Bohemond I, so the identification is now secure. The coins of Bohemond I are scarce but recently a relatively large number have appeared in trade. This provides an opportunity to analyse the coinage in more detail. There are two types in somewhat different styles and it is not clear whether they should be regarded as substantive types or just the work of different die cutters. The article considers the possibility that one could belong to the first period of Bohemond‘s presence 1 The author is grateful to Susan Tyler-Smith for her help in the preparation of this article. А COIN OF BOHEMOND I OF ANTIOCH OVERSTRUCK BY TANCRED The Ukrainian Numismatic Annual. Issue 6. Kropyvnytskyi – Kyiv – Pereiaslav , 2022. 207 in Antioch and the other to his second. If this is the case then it is possible that some of Tancred‘s early coins date from the time of his first  ̳regency‘ since the precedent for coinage had been established. It seems more likely, however, that both types belong to Bohemond‘s second stay and that Tancred did not begin to coin until after Bohemond left Antioch for good. As a postscript a new  ̳variety‘ of the coinage of Bohemond of scyphate fabric is published.

in Antioch and the other to his second. If this is the case then it is possible that some of Tancred's early coins date from the time of his first ‗regency' since the precedent for coinage had been established. It seems more likely, however, that both types belong to Bohemond's second stay and that Tancred did not begin to coin until after Bohemond left Antioch for good.
As a postscript a new ‗variety' of the coinage of Bohemond of scyphate fabric is published.

Bohemond I
Bohemond of Tarento (c.1054Tarento (c. -1111 was the leader of the Italian Norman contingent on the First Crusade. He became Prince of Antioch when the city was captured in the summer of 1098 but his authority was disputed by other crusaders until November. In August 1099, he was captured by the Danishmendids and held for ransom. His nephew Tancred assumed the regency. Bohemond was released in August 1103 after the ransom had been paid by the king of Jerusalem and the patriarch of Antioch. In September 1103 Bohemond I returned to Europe ostensibly to obtain reinforcements. Instead, he organised an attack on Byzantine territory in Dalmatia which ended in disaster. He was forced to sign a humiliating treaty with the Byzantine Emperor Alexius I in 1108. His reputation was irreparably damaged and this made it impossible for him to ever return to Antioch. Bohemond died in 1111 and Tancred became the ‗regent' for Bohemond's infant son until 1112 when Tancred died. His successor as ‗regent', Roger, fell in battle 1119. Bohemond's son finally arrived in Antioch as the new prince (Bohemond II) in 1126.
The status of the two ‗regents', Tancred and Roger, has been the subject of some debate. William of Tyre regarded them as regents but it is clear from documents and seals that they regarded themselves as legitimate princes, though Roger was the only one to style himself ‗Prince of Antioch' on his coins, and then only on one type. What is less clear is Tancred's position during the lifetime of Bohemond I. Tancred was a brilliant soldier but his ambition was notorious and he made no attempt to raise the ransom money for Bohemond. This was done by Baldwin of Jerusalem because he was worried by Tancred's growing power. Until his marriage to Cecilia, a daughter of the King of France (arranged by Bohemond), in 1107, Tancred was technically a landless governor. 2

Bohemond I's coinage
Copper coins in Byzantine style were issued by both Bohemonds and the two ‗regents'. The sequence is clear enough from the pattern of overstriking. 3 Only one type is now attributed to Bohemond I. It is a Byzantine style copper and shows a bust of St Peter on the obverse and, on the reverse, a cross with the letters B H M T in the angles. It is customary to refer to it as a follis.
Before the First Crusade, while still in Italy, Bohemond used a lead seal with the bust of St Peter and an invocation, customary on Byzantine seals, on the reverse: ‗Lord help your servant Bohemond'. The design on Bohemond's seal was closely copied on the first issue of Tancred.
The coin of Bohemond I was one of the first crusader coins to be published. A piece in the collection of Christian Jurgensen Thomsen in Copenhagen was published by Bishop Frederick Münter. 4 In spite of what later authors have said he transcribed and illustrated the coin accurately but could not decipher the third letter of B H M T. He did not attempt to attribute the coin. There were none in the collection formed by E. Cousinery who mistakenly published a coin of Bohemond II as Bohemond I, an attribution corrected by Félicien de Saulcy. De Saulcy was also the first to identify the Thomsen coin as belonging to Bohemond I though he, and the subsequent authors of the Thomsen catalogue, mis-read the reverse letters as B H H T. 5 The reading was corrected by Gustav Schlumberger. He only knew of two other examples one of which, from the Paul Lambros collection, he illustrated. 6 The coin has until recently remained quite scarce. There was not even a specimen in the Antioch excavations though single examples were found at al-Mina, Corinth and Athens. 7 There was one in the Dimitri Dolivo collection and Michael Metcalf, Martin Rheinheimer and John Porteous published examples. 8 Curiously, in a series where the chronology depends on overstriking, the coin type attributed to Bohemond I was never found overstruck by coins attributed to his successors.
Nonetheless, no one questioned the attribution to Bohemond I except, who sounded a note of caution: ‗The main arguments for giving so early a date to this rather scarce coin are its primitive style and the arrangement of the brief inscription about the cross … There is also the negative argument that this coin is not found overstruck on Tancred's coins. However, it would be reassuring if a specimen could be found overstruck by one of Tancred's issues'. In 1991 Rheinheimer confirmed that overstrikes on the coins of Bohemond I were unrecorded. 9 Reviewing the situation in 1995 Metcalf had no doubts of the correct sequence but commented: ‗Coins of Bohemond on clean flans and of good weight are plausibly given to Bohemond I rather than Bohemond II'. 10

A coin of Tancred overstruck on one of Bohemond I
The main purpose of this article is to publish just such an overstrike (Fig. 1). It is a type 3 of Tancred which shows a standing figure of St Peter on the obverse and a cross with the letters D S F T (domine salvum fac Tancredum) in the angles, on the reverse. The floreate base of the cross is clearly visible as an undertype at 5 o'clock on the obverse (Fig. 1b). The bust of St Peter with traces of the inscription is clearly visible under the reverse cross and indeed obliterates the D and F of the inscription.    That the overtype should be type 3 is not necessarily a surprise. Type 1 of Tancred is rarely found overstruck on another coin except very occasionally on Seljuk coppers. The same is true of type 2 though overstrikes on type 1 are not unknown. By contrast it is difficult to find a type 3 that is not overstruck, usually on Tancred types 1 or 2.
Nonetheless it seemed worthwhile taking the opportunity to also look in more detail at the issues of Bohemond I. There is still the question, posed by Metcalf, whether the coinage of Bohemond I should be dated to his first or second period of physical presence in Antioch and when the first issue of Tancred which, it should be remembered, copies a seal of Bohemond, should be dated. 11 The fact that the overstrike is on the third issue of Tancred leaves this question open. In addition the coins of Bohemond I are no longer as scarce as they were. 12 The two styles/types of Bohemond I folles  The coin illustrated by Schlumberger (Fig. 2) was rather different from other known examples (Fig. 3). On the obverse the drapery of the bust was different and the staff of the cross obliterated part of the nimbus. On the reverse the arms of the cross were thinner and the ornamentation at the ends was simpler. No one seems to have commented on this and it may have been assumed that the drawing in Schlumberger was inaccurate. That would have been unfair as Dardel's illustrations in Numismatique de l'Orient Latin are uniformly excellent. It was only with the sale of the Slocum collection that the situation became clear (Fig. 4). This contained six coins of Bohemond I, two in the style of Schlumberger's coin (lots 54-55), and four in the more familiar style (lots 56-59). In the following I have arbitrarily designated the first two as type 1 and the last four as type 2.  (Figs 5a and 8) and there is also a better illustration available of Slocum lot 54 (Fig. 5b). On type 1 the impression of a rather gaunt face is misleading due to wear and weak striking. The bust is in reality bearded but perhaps, as already noted, rather older and thinner than the bust on type 2. The simplest way to tell worn specimens apart is the position of the cross against the nimbus already referred to. Dardel's depiction of the reverse is accurate. There is a difference in the ornamentation of the reverse cross, which is certainly thinner (Fig. 6). Type 1 remains the rarer of the two types. Of the 30 odd examples I have noticed from recent auctions only six are type 1.
Type 1 Type 2 Fig. 6. Ornaments on arms of reverse cross.
In practice there is some overlap between the types. The reverse cross on Slocum 59 is quite thin. Examples exist of type 1 with a thick reverse cross though the ends seem to be the simplified type (Fig. 7a). Coins of type 2 sometimes have a thin cross though the ornaments at the end of the arms are, where visible, typical for type 2. The coin illustrated as Fig. 7b clearly shows the Seljuk undertype. The only real justification for calling them different types as opposed to mere stylistic variants is the arrangement of the obverse legends. On type 1 the curious monogram form of the last four letters of ΠЕТΡΟϹ is to the right. On type 2 the obverse legends are often illegible but where visible the ‗monogram' is to the left of the bust (Fig.  7). On type 2 the saint's right hand held in blessing is clearly visible across the breast. I have only seen one example of type 1 where this is present. The fingers are pointed up towards the chin rather than across the breast (Fig. 9).
Type 1 Type 2 An attempt was made to check the dies but this proved difficult owing to the incomplete striking and patchy surfaces of the coins. There are certainly two obverse and two reverse dies in type 1 (compare Figs 5 and 7a). There seem to be several reverse dies for type 2. For example, in Fig. 4 lots 57 and 58 are certainly different reverse dies but it is not possible to be sure about the others. I have to say that I have so far failed to find two examples of type 2 where the obverse die was clearly different. One can certainly conclude that both series were struck from very few dies.

Fig. 8. Obverse legends
Is there any significance in the two types? It is possible than one dates from Bohemond's first residence in Antioch and the other from his second. If this was the case the question arises: what happened during Bohemond's four year captivity between 1099 and 1103? The coins are too rare to have been issued over a four-year period so either the mint ceased production or Tancred took the opportunity to coin in his own name. Everything we know about Tancred suggests this would have been the case, but his coinage is so different in fabric and so abundant that it hardly fits the proposed interval in Bohemond's coinage. Consequently, I think it is more likely that both types belong to the period after Bohemond's release and before his departure for Europe a year later. Either Tancred was too preoccupied, or else uncertain of position, or both, so he did not initiate the Antioch coinage. Once Bohemond, inspired by the example of the counts of Edessa, had set the precedent, albeit on a small scale, Tancred followed. He may also have felt more secure once Bohemond had quit Antioch though he was not to know that Bohemond would never return. The volume of Tancred's coinage, especially his type 1 is such that it is highly likely that he started immediately Bohemond left. The two ‗types' of Bohemond I may, of course, simply indicate another die cutter who chose to arrange the obverse legends differently.

Postscriptnew type/variety?
The coin illustrated as Fig. 10 recently appeared in trade. It seems to be a coin of Bohemond 1 but the lettering is quite different. The piece is double struck and of scyphate fabric. All that can be said for the present is that it is unlike any other published coin of Bohemond I.