Bearded Head Quarter Staters: Fact or Fiction version 2

Bearded Head Quarter Staters: Fact or Fiction? Dr. David Robinson, January 2021 Verison2: Updated April 2023 This is an update version of the original paper. Abstract: Concerns about the authenticity of Bearded Head quarter staters have lingered, unpublished and unchallenged, for over quarter of a century. This paper examines all of the concerns that have been raised about these coins and shows that they are no longer valid. The iconography, metallurgy, metrology, and find spots are all consistent with genuine coins, and the evidence points to them being minted alongside the British C "Yarmouth" stater. The coins are renamed Horse Geometric quarter staters in recognition of the more likely interpretation of the obverse image. Version 2: April 2023 I published the original paper in January 2021 and a number of things have changed since then that warrant a slight update to the paper. I’m taking the opportunity to fix some small mistakes as well. Updates 1. The coins have now been accepted as genuine, and Chris Rudd Ltd. have started selling their original coins 2. A coin from a third reverse die was found in January 2023 3. The find spot for the coin found on “the Ridgeway” has become available, as have the find spots for some of the older coins Errors 1. I originally said the first coin appeared in the trade in Blancon's 1993-1994 fixed price list. This turned out to be incorrect. The first one (CCI 97.0804) was sold in 1991 (Münzen & Medaillen list 459 number 137 Nov/Dec. 1991) 2. The “Reverse Iconography” section said the reverse imagery was derived from Gallo-Belgic D (GB-D). That should have said Gallo-Belgic Ca2 (GB-Ca2). The footnote also incorrectly said GB-D. 1 3. The “Obverse Iconography” section said “ABC 770 (Cogwheel Smiler) and DK 319 (Cheriton Annulets) have a modified "abstract-beast/men-in-a-boat" design from GB-D.” That should have said GB-Ca2. Introduction "Bearded Head" quarter staters first appeared in the trade in 1991 and then again in 19931 and were attributed by Chris Rudd to either the Durotriges or the Atrebates when he advertised one for sale in 19952. Nine were recorded in the Celtic Coin Index (CCI) between 1995-19983, all marked with "authenticity not certain" and with notes such as "False provenance of Westhampnett bypass", "Cracks in edge of flan. Fake or genuine?", and "Genuine? From the usual dies". It is rumoured that the original finder(s) of the coins lied about the find spot, possibly to conceal an undeclared hoard. Being a newly discovered type with novel imagery, this put the authenticity of the coins in doubt. If the coins are fake then we would hope to find some evidence of that on the coins themselves, but even as late as 2018 a meeting between Chris Rudd Ltd., Dr. Philip de Jersey, and Dr. John Sills, failed to reach a conclusion. Stated another way, in the quarter of a century after the first coin was sold, the premier experts in the field, along with other respected dealers4 who have sold these coins, have been unable to identify anything that condemns any of the coins. Despite a detailed analysis of the coins individually and as a group, it appears that there is nothing definitively wrong with them. They have been found in recent years by independent metal detectorists who have no connection to the original finders. When this is considered alongside the continuing inability to prove the coins to be fake, then the argument that these coins are simply the victims of overcaution has to be given serious attention. This paper reviews the available evidence in the hope that this situation can be resolved. For reasons that will become clear later, these coins will be called "Horse Geometric" quarter staters from this point on. The first section discusses Horse Geometric quarter staters with the working assumption that they are genuine. It analyses their iconography, weight, metallurgy, and typology, and investigates if they can be assigned as the companion quarter to a known stater. The second section looks at each of the concerns that have been stated about the authenticity of these coins, and shows that none can continue to stand as valid concerns. The catalogue of coins used in this report is given in an appendix. 1 CCI 97.0804 - Münzen & Medaillen list 459 number 137 Nov/Dec. 1991; CCI 97.0805 - Blancon PSL 13 no. 886 2 CCI 95.1195 - CR List 19. 3 95.1195 (CCI Link), 96.1797, 96.1798 (CCI Link), 96.3353 (CCI Link), 97.0804 (CCI Link), 97.0805 (CCI Link), 97.0806 (CCI Link), 97.0807 (CCI Link), 98.0724 (CCI Link) 4 Gilles Blancon of Hannover (1993), DNW (2008), and Mike Vosper (1998, 2020). 2 About the Coins This section introduces Horse Geometric quarter staters and assumes for simplicity that they are authentic. The concerns over authenticity will be addressed in the next section. Reverse Iconography The image on the reverse of the coin consists of two vertical "fish" shaped objects which are slightly offset from each other horizontally. Two "hand" style objects and two "axe" style objects appear in pairs in the upper right and lower left quadrants. Pellets connected by arcs make up the rest of the image. Figure 1: The reverse image. The photo is a composite from various sources The full image is around 10.7 mm wide and 11.6 mm tall, and most of the design fits on most of the flans, giving the impression that the coins are well centred. The imagery on the reverse seems at first glance to be unique amongst Celtic quarter staters, but it is actually quite faithful, in the round, to the "geometric" quarter stater designs that derive from Gallo-Belgic Ca2 (GB-Ca2). These are characterised by two offset vertical5 lines with Y shaped objects in the lower left and upper right quadrants, with an object in the lower right with lines radiating from it. You can see the similarities to other coins if you look at a broad-brush rendering of the images. The trick is to match overall shapes, not details: 5 Technically this description should be rotated 90 degrees anti-clockwise to match the original GB-Ca2 imagery. In that, the line is horizontal on two levels, with the trophy/tree like object above and one Y shaped object below. 3 ABC 2205 "Duro Boat Gold" British Ad2 Class 3 "Geometric". Also known as ABC 553 "Phallic Geometric" Horse Geometric Figure 2: The reverse image in context One interesting feature is that the "fish" shaped objects which make up what's commonly known as the "thunderbolt" appear to be the neck used in the Westerham style horses. For example, here's an Af1 Lepe stater where the horse's neck is joined to one of the pellets by a die flaw. Figure 3: The "fish" motif on both the Ad1 Tarring stater and the Horse Geometric quarter stater. CCI 98.0417 and CCI 96.1797 4 Obverse Iconography The obverse is difficult to explain on this type. It was called the Bearded Head quarter by Chris Rudd when he advertised one for sale. The head of Apollo traditionally appeared on the obverse of early Celtic coins, and a cluster of pellets occurs where a beard might be. The appearance of a head looks to be an optical illusion on his particular coin. The crescent shaped object in front of the head was mainly missing, and it did have head-like qualities due to wear. It has been difficult to visualise a head on other coins. Figure 4: The obverse interpreted as a head on Chris Rudd's original coin (the one used to name the type) There's very little on the obverse to suggest that it is a head. The three straight lines at the bottom would be the "cloak" from the Gallo-Belgic C (GB-C) staters and early British staters, and the crescents would be the those normally found at the end of the hair bar. Figure 5: The obverse interpreted as a head. The basics However, the hair bar, the wreath, the hair locks, and the hair curls are missing. 5 The cloak The hair bar crescents and hair bar (missing) The hair locks (missing) The wreath (missing) The hair curls (missing) Figure 6: The obverse interpreted as a head. Missing elements It's hard to interpret this as a further abstraction of the traditional Apollo's head design because so many features are missing. Figure 7: The obverse interpreted as a head. Details It's not clear where the inspiration for a bearded head would have come from. A bearded head appeared on first generation Philippus coins, for example this 3rd century BC hemistater from the Ambiani, but those coins predate these quarters by at least 150 years. 6 Figure 8: Ambiani Hemistater. Copyright CNG When Chris Rudd advertised one for sale in 1995 he wondered if it was derived from the Zeus Dodonaios stater of Alexander of Epeiros (342-330 BC), but that is similarly distant in time, and would be a novel departure from all other Celtic coins minted in Britain. It's hard to form a hypothesis to why this feature would be resurrected for one coin type after all this time. We can't rule out that the beard was a local innovation, but we also can't assume it is a beard. There's an array of pellets in front of the "head", but they could be background objects unrelated to the head itself, much like the pellet triad above the head appears to be. Westerham style staters have a horse on the reverse with an array of pellets above (for example, the British B Chute stater), but the idea that the pellets are part of the horse is never raised. There's no reason to assume the pellets on this quarter are a beard at all, especially as that was not part of the traditional iconography for Apollo's head. We also can't rule out the possibility that it's not a head at all, bearded or otherwise. If we ignore the "cloak" motif, there's very little reason to consider it a head. The standard motifs of the hair bar, the wreath, the hair curls, and the hair locks, are all missing. Another interpretation of the obverse exists if the coin is rotated 90 degrees anti-clockwise. In this orientation an abstract horse can be seen with one of its forelegs raised. This would tie in with the "abstract beast" description sometimes used to describe the design seen on the GB-C, GB-D and later quarter staters (which originally derive from Apollo's wreath on GB-Aa2 quarter staters). Figure 9: The obverse interpreted as a horse 7 The horse is segmented with a small, sleek, rear and a large muscular front. The tail is rendered as three lines radiating from a common point, reminiscent of the triple tailed horses common on unscribed coins. The two rear legs are hooked crescents. One of the front legs appears in a group with the rear legs and is formed by an inverted "foot" shape which ends with a crescent, giving the effect of a hook or sickle. The second front leg is raised and is formed by a distorted "M" shaped object. The head and neck can be hard to spot, but on the best coins the shape is distinct and ends with a clear muzzle. An indistinct "S" shaped object appears in front of the horse's muzzle, and a crescent object with four rays appears in front of, and above, the horse's head. The ten pellets that appear above the horse are reminiscent of the pellet clusters above the horse on Westerham style staters, such as the British Ad1 "Tarring", British Af1 "Lepe", and British B "Chute" staters. A horse with thirteen pellets is a common reverse image, but is unheard of as an obverse image. The only close parallel might be GB-Ca2, GB-D and later quarters, where the obverse is said by some to be an "abstract beast". The interpretation of the obverse as a bearded head or an abstract horse doesn't fit into any standard iconography. However, other quarter staters from the same area have a range of "nonstandard" obverse images. ABC 542 (Aldingbourne Annulets) has a row of three ringed pellets, ABC 548 (Tangmere Eel Spear) has three bars in a double topped "T" formation, ABC 551 (Selsey Diadem) has a head, ABC 623 (Wittering Rings) has two small rings and some banding, ABC 626 (Two Crescents) has two opposing crescents on either side of a wreath style bar, ABC 629 (Willet Triple Line) has two pellet rings to each side of a triple line made from pellets, ABC 632 (Selsey Tramlines) has two parallel corded lines with three ringed pellets within, and ring triads outside. ABC 635 (Guildford Tramlines) and 638 (Ewell Tramlines) show similar designs. ABC 770 (Cogwheel Smiler) and DK 319 (Cheriton Annulets) have a modified "abstract-beast/men-in-a-boat" design from GBCa2. DK 180 (Eye Star) quarters reuse the wreath-plus-eye design from the "à l'astre" series. This style of design was a radical deviation from the norm when it first appeared. ABC 779 (Three Wheeler) has two large spoked wheels in a pellet field, while ABC 782 (Huxtable's Eagles) features a fourarmed spiral, with pellet infill, topped with horse and eagle heads. ABC 785 (Bird's Head Cruciform) has four leaves around a sunburst and birds head motifs. We can go on citing examples 6 of quarter staters which used something "non-standard" on the obverse. Experimentation and deviation from traditional norms was common. In this light, a bearded head or an abstract horse shouldn't seem surprising. It's also worth noting that in 2019 Chris Rudd described the "Corded Crescents" quarter stater as having a hidden bearded head, so it appears that a bearded head is neither unique nor objectionable. 6 See ABC 788, 791, 794, 797, 800, most of the Iceni gold, the Dubonni gold, ABC 2237, 2246, 2249, 2252, and 2583 8 Figure 10: Coin News November 2019 (Copyright www.tokenpublishing.com) Find Spots The find spots for the coins (where known to be real) are mainly in the assumed territory of the Regini tribe although there are some outliers. Figure 11: Find spots Metallurgy A metallurgical analysis of one coin (CCI 96.3353) was performed by Dr. Peter Northover in 1996, and the alloy was found to be 24.8% gold, 44.5% silver, and 30.4% copper (Sills, email). The closest coins to this are: British C class 1 British G class 2a British G class 2b Horse Geometric Au% 28.9 28.2 25.5 24.8 Ag% 47.4 43.9 44.3 44.5 Cu% 23.6 21.2 30.3 30.4 Ag/Au 1.6 1.6 1.7 1.8 Ca/Au 0.8 0.8 1.2 1.2 Figure 12: Average alloy composition as % and ratio for various types 9 Figure 13: Alloys for British C class 1, British G class 2a and 2b, and Horse Geometric There's a limit to what can inferred from one sample but it confirms that there's nothing out of the ordinary with this particular coin's metallurgy. It's a very close match to the British G class 2b average, and close to the others, deviating by a maximum of 4.1% for gold, 2.9% for silver, and 9.2% for copper. The alloy from one coin is not always representative of the entire population. Coins from the same type can have a range of values, so we can't take the point value of 24.8% gold from this coin and say for certain that Horse Geometric quarters were lower fineness than British G class 2a (average 28.2%) and British C (average 28.9%). British G class 2a (62 coins recorded; 7 obverse dies, 15 reverse dies) is a good example of a type where the metallurgy varied considerably: 10 Au% Ag% Cu% Ag/Au Ca/Au Average Coin 1 Coin 2 Coin 3 Coin 4 Coin 5 Coin 6 Coin 7 Coin 8 Coin 9 28.20 43.90 27.77 1.6 1.0 35 43.8 21.2 1.3 0.6 30.3 44.6 25 1.5 0.8 36.3 41.8 21.8 1.2 0.6 31.7 39.7 27.6 1.3 0.9 26.8 44.5 28.7 1.7 1.1 28.1 41.3 30.6 1.5 1.1 26.5 43.9 29.5 1.7 1.1 21.6 46.5 32 2.2 1.5 17.5 49 33.5 2.8 1.9 Range 18.8 9.3 12.3 1.6 1.3 Figure 14: Alloy composition as % and ratio for British G class 2a Figure 15: Alloy for British G class 2a 11 The percentage of gold in the alloy drops from 36.3% to 17.5%, which is a range of 18.8%. The silver and copper content is also variable, and the ratios of the various metals are not consistent. British C class 1 (14 coins recorded; 2 obverse dies, 2 reverse dies) is a type where the alloy was well controlled, although that may be expected as it was a much smaller issue and wouldn't have had much time to suffer from debasement. We also have to consider that the coins analysed7 came from a single hoard which is likely to have contained the best coins. Au% Ag% Cu% Ag/Au Ca/Au Average Coin 1 Coin 2 Coin 3 Coin 4 Coin 5 Coin 6 28.88 47.40 23.58 1.6 0.8 29.8 30 48.6 48 21.6 23 1.6 0.7 1.6 0.8 30 49 22 1.6 0.7 29.7 48 22.4 1.6 0.8 27 45 25 1.7 0.9 26.8 45.8 27.5 1.7 1.0 Range 3.2 4.0 5.9 0.1 0.3 Figure 16: Alloy composition as % and ratio for British C class 1 7 There is one more British C stater that hasn't been included because it is underweight by 2g and has a significantly different alloy (27%, 14.8%, 55%). It is believed to be a plated copy. 12 Figure 17: Alloy for British C class 1 The ranges here are much tighter, and the ratios are quite consistent. British Da1 class 1d is another large issue (34 coins; 8 obverse dies, 10 reverse dies) which is closer to British G class 2a: 13 Average Coin 1 Coin 2 Coin 3 Coin 4 Coin 5 Coin 6 Coin 7 Coin 8 Coin 9 Coin 10 Coin 11 Coin 12 Coin 13 Coin 14 Range Au% Ag% Cu% Ag/Au Ca/Au 33 36 30 1.1 0.9 37 42 21 1.1 0.6 36 32 32 0.9 0.9 31 39 31 1.3 1.0 35 42 24 1.2 0.7 34 42 24 1.2 0.7 32 43 24 1.3 0.8 36 42 22 1.2 0.6 35 28 34 0.8 1.0 36 27 34 0.8 0.9 36 41 23 1.1 0.6 31 28 40 0.9 1.3 29 32 39 1.1 1.3 31 30 39 1.0 1.3 27 35 38 1.3 1.4 10.0 16.0 19.0 0.6 0.8 Figure 18: Alloy composition as % and ratio for British Da1 class 1d 14 Figure 19: Alloy for British Da1 class 1d The deviations in gold, silver and copper can be quite high, but the ratios between them stay well controlled. The following table summarises the ranges in the alloy compositions from the above examples. It can be clearly seen that coins from the same type can exhibit quite large deviations in gold, silver and copper content. British C class 1 (14 coins recorded) British G class 2a (62 coins recorded) British Da1 class 1d (34 coins recorded) Au% Ag% Cu% Ag/Au Ca/Au 3.2 4.0 5.9 0.1 0.3 18.8 9.3 12.3 1.6 1.3 10.0 16.0 19.0 0.6 0.8 Figure 20: Alloy ranges for three coin types 15 The Horse Geometric quarter that has been analysed sits at an unknown location8 within a range of unknown size. If the range is similar to British C class 1 then we can be relatively confident that the sample we have is representative of the type, but it could have maximum alloy components of 28% gold, 48.5% silver or 36.3% copper, which puts it in the range of British C, British G class 2a and British G class 2b, although in all cases the copper content would be high. It's possible that the copper content of the one coin analysed is already at the high end of the range, in which case it could drop to 24.5% and that's with all of their ranges. If the range is similar to British G class 2a's, then our coin could be at the low end of a wide range and the gold content at the start of the sequence could be up to 18.8% higher. This puts it at 43.4% and into the range of British B, British C, all of British D, British F and British G. Without access to the actual coins, or colour corrected photos, there's no way to judge how wide the range is. The coins are generally all described as coppery or base gold, and have been likened to British C in appearance, so the variation appears to be low. Metrology There are fifteen coins in the catalogue. The weight of two of them is unknown, one is a fragment of a coin, and four are only recorded to 1 decimal place. They are compared against eight other early quarter stater types. The bars indicate the percentage of coins of the type in each weight range. 8 When the coins are placed in order by die wear, it appears in position eleven of fifteen, so it may be one of the lower quality coins. 16 Weight g Insular Cf2 Class 1 Insular Cf2 Class 2 Ad2 Geometric Af2 Stippled Ad2 Nipple 1.65 1.55                                                    1.55 1.45                                                                                                                                                                          1.45 1.35                                                                                                                                                                                                             1.35 1.25                                        1.25 1.15                          1.15 1.05                 1.05 0.95              0.95 0.85       Weight g B2 Hampshire Thunderbolt British G2 Clacton de Jersey Horse Geometric Da2 Curdridge 1.65 1.55     1.55 1.45                                 1.45 1.35                                                                                                                                            1.35 1.25                                                                                                                                                                                                    1.25 1.15                                  1.15 1.05             1.05 0.95     0.95 0.85     Figure 21: Weight distribution compared to other types of quarter staters shown as a percentage. 17 Insular Cf2 Insular Cf2 Class 1 Class 2 Number of Coins Min (g) Max (g) Average (g) Median (g) 23 1.41 1.64 1.51 1.52 7 1.38 1.50 1.44 1.44 B2 Hampshire Thunderbolt Number of Coins Min (g) Max (g) Average (g) Median (g) 103 1.05 1.52 1.36 1.37 Ad2 Geometric Af2 Stippled Ad2 Nipple 52 0.90 1.63 1.40 1.43 31 1.00 1.55 1.38 1.43 43 0.99 1.48 1.35 1.39 British G2 Clacton de Jersey 68 1.10 1.52 1.36 1.37 Horse Geometric Da2 Curdridge 12 1.20 1.46 1.33 1.32 22 1.05 1.38 1.26 1.27 Figure 22: Weight statistics (in g) compared to other types of quarter staters Based on the average and median measurements, Horse Geometric quarters would appear to come shortly after British B2 "Hampshire Thunderbolts" and British G2 "Clacton de Jersey" quarter staters, and just before the British Da2 "Curdridge" quarter staters. This assumes that they were all struck to the same weight standard, all measurements are accurate, and the extant coins are representative of the original population. Associated Stater Many quarter staters types are associated with a stater type, and it is tempting to try and identify an associate for Horse Geometric quarter staters. The most obvious candidate is British C9. Both types are small issues from the same area, have similar metallurgy, and have novel iconography. British C does not have an associated quarter stater. British G is a slightly better match in terms of metallurgy, but the distribution is not compatible (North Thames vs. South coast) and it already has associated quarters. There are other staters from the same area which don't have associated quarters, but only the Cheesefoot Head comes close to having the correct metallurgy (31.7% gold, 28.1% silver, 38.0% copper), but the silver to gold ratio is wrong. The silver content of the Horse Geometric is 44.5% as opposed to 28.1% for the Cheesefoot Head. 9 The idea that the Bearded Head quarters are associated with British C is not a new one. 18 Horse Geometric Cheesefoot Head Au% 24.8 31.7 Ag% 44.5 28.1 Cu% 30.4 38.0 Ag/Au 1.8 0.9 Ca/Au 1.2 1.2 Figure 23: Alloy composition as % and ratio for the Horse Geometric quarter stater and the Cheesefoot Head stater Figure 24: Alloy for the Horse Geometric quarter stater and the Cheesefoot Head stater Metrology To compare the weight of Horse Geometric and British C, we can exploit the fact that the Horse Geometric's metrology is very close to British G2's. The metrology of their associated staters should also be similar. The following chart shows the weight distribution of British C class 1 and British G class 2a staters. 19 Weight g 6.60 6.50 6.40 6.30 6.20 6.10 6.00 5.90 British C Yarmouth Right Stater 6.50 6.40 6.30 6.20 6.10 6.00 5.90 5.80 British G class 2a Clacton Curved Exergue                                                                                                                                                                                                               Figure 25: Comparing the metrology of British C and British G class 2a. Bars are % of coins in that weight range These have a similar distribution between 6.4g and 6.0g (shown in black). British G Class 2a has a much larger sample size (62 coins) than British C class 1 (14 coins) so has more outliers (shown in grey). British G outliers at 5.64g, 5.12g and 4.52g have been excluded. The British C outlier of 4.07g has not been included because it is thought to be plated. The metrology suggests British C might be the associated stater. Metallurgy The metallurgy of British C class 1 is close to that of the Horse Geometric quarters, although the numbers don't quite match, and the copper component of the one sample we have is high: Horse Geometric British C class 1 Au% 24.8 28.9 Ag% 44.5 47.4 Cu% 30.4 23.6 Ag/Au 1.8 1.6 Ca/Au 1.2 0.8 Figure 26: Alloy composition as % and ratio for the Horse Geometric quarter stater and British C staters (average) The following table shows the alloy composition of each British C class 1 coin (ignoring the underweight plated coin), and the differences between them and the Horse Geometric quarter stater: Coin 1 Coin 2 Coin 3 Coin 4 Coin 5 Coin 6 Au% 29.8 30 30 29.7 27 26.8 Ag% 48.6 48 49 48 45 45.8 Cu% 21.6 23 22 22.4 25 27.5 ΔAu% 5.0 5.2 5.2 4.9 2.2 2.0 ΔAg% 4.1 3.5 4.5 3.5 0.5 1.3 ΔCu% -8.8 -7.4 -8.4 -8.0 -5.4 -2.9 Min Max Range 26.8 30.0 3.2 45.0 49.0 4.0 21.6 27.5 5.9 2.0 5.2 3.2 0.5 4.5 4.0 -8.8 -2.9 5.9 Figure 27: Alloy composition for British C, and difference from the Horse Geometric quarter 20 Figure 28: Alloy for British C and the Horse Geometric quarter In all instances, the gold and silver content of British C is higher (by at least 2% and 0.5%), and the copper content is lower (by at least 2.9%). These are close, but are they close enough? We can check by looking at the relationships between other staters and their associated quarters. British Da has been chosen as it has a similar metallurgy and only has data for one quarter stater. The quarter stater is class 1a which John Sills [Sills 2017] has associated with the stater class 1c. The tables below compare it against this class, as well as the preceding class (1b) and the subsequent class (1d). 21 Au% 33 Ag% 36 Cu% 30 ΔAu% ΔAg% ΔCu% Quarter stater Stater 1 Stater 2 Stater 3 Stater 4 Stater 5 38 37 38 37 40 45 46 44 44 41 17 17 19 19 19 5.0 4.0 5.0 4.0 7.0 9.0 10.0 8.0 8.0 5.0 -13.0 -13.0 -11.0 -11.0 -11.0 Min Max Range 37.0 40.0 3.0 41.0 46.0 5.0 17.0 19.0 2.0 4.0 7.0 3.0 5.0 10.0 5.0 -13.0 -11.0 2.0 Figure 29: Alloy composition for British Da1 class 1b, and difference from the British Da2 class 1a quarter stater. This is the preceding stater class Figure 30: Alloy composition for British Da1 class 1b and the British Da2 class 1a quarter stater. This is the preceding stater class 22 Au% 33 Ag% 36 Cu% 30 ΔAu% ΔAg% ΔCu% Quarter stater Stater 1 Stater 2 Stater 3 Stater 4 Stater 5 38 38 40 39 40 42 40 43 39 42 20 22 17 22 18 5.0 5.0 7.0 6.0 7.0 6.0 4.0 7.0 3.0 6.0 -10.0 -8.0 -13.0 -8.0 -12.0 Min Max Range 38.0 40.0 2.0 39.0 43.0 4.0 17.0 22.0 5.0 5.0 7.0 2.0 3.0 7.0 4.0 -13.0 -8.0 5.0 Figure 31: Alloy composition for British Da1 class 1c, and difference from the British Da2 class 1a quarter stater. This is the associated stater class Figure 32: Alloy composition for British Da1 class 1c and the British Da2 class 1a quarter stater. This is the associated stater 23 Au% 33 Ag% 36 Cu% 30 ΔAu% ΔAg% ΔCu% Quarter stater Stater 1 Stater 2 Stater 3 Stater 4 Stater 5 Stater 6 Stater 7 Stater 8 Stater 9 Stater 10 Stater 11 Stater 12 Stater 13 Stater 14 37 36 31 35 34 32 36 35 36 36 31 29 31 27 42 32 39 42 42 43 42 28 27 41 28 32 30 35 21 32 31 24 24 24 22 34 34 23 40 39 39 38 4.0 3.2 -2.0 2.0 1.0 -1.0 3.0 2.0 3.0 3.0 -2.0 -4.0 -2.0 -6.0 6.0 -4.2 3.0 6.0 6.0 7.0 6.0 -8.0 -9.0 5.0 -8.0 -4.0 -6.0 -1.0 -9.0 1.6 1.0 -6.0 -6.0 -6.0 -8.0 4.0 4.0 -7.0 10.0 9.0 9.0 8.0 Min Max Range 27.0 37.0 10.0 27.0 43.0 16.0 21.0 40.0 19.0 -6.0 4.0 10.0 -9.0 7.0 16.0 -9.0 10.0 19.0 Figure 33: Alloy composition for British Da1 class 1d, and difference from the British Da2 class 1a quarter stater. This is the succeeding stater class 24 Figure 34: Alloy composition for British Da1 class 1d and the British Da2 class 1a quarter stater. This is the succeeding stater British C class 1 vs. Horse Geometric quarter British Da1 class 1b Stater (preceding class) vs. Quarter British Da1 class 1c Stater (associated class) vs. Quarter British Da1 class 1d Stater (succeeding class) vs. Quarter ΔAu% 2.0 to 5.2 4 to 7 5 to 7 -6 to 4 ΔAg% 0.5 to 4.5 5 to 10 3 to 7 -9 to 7 ΔCu% -8.8 -to -2.9 -13 to -11 -13 to -8 -9 to 10 Figure 35: Summary of alloy deviation ranges between staters and associated quarters The deviations between the Horse Geometric and British C class 1 are 2% to 5.2%, 0.5% to 4.5%, and -8.8% to -2.9% for gold, silver and copper respectively. Deviations between British Da2 and its associated stater (British Da1 class 1c) are greater at 5% to 7%, 3% to 7%, and -13% to -8%. Comparisons to the other classes are worse. The metallurgy tells us that Horse Geometrics are a better match to British C class 1 than British Da2 is to British Da1. 25 Find Spots NOTE: More find spots have become available since the paper was originally written. These have been added to the map below, but the catalogue at the end of the paper has not been updated The Horse Geometric and British C find spots are similar if we ignore the hoard found in Yarmouth on the Isle of Wight. Figure 36: British C Stater (●) and Horse Geometric Quarter (○) find spots. Iconography The case for Horse Geometric quarters being associated with British C staters will be strengthened if there are some visual similarities to either each other, or to the type they derived from. 26 British C is derived from the GB-Ce staters10 [Sills 2017]. It features an abstract head of Apollo on the obverse, and a Belgic style horse on the reverse. The hair locks on the obverse are a talon shape, and the cloak is in the form of three parallel crescents. The face is a kind of L shape with three radiating lines for a nose. Figure 37: British C class 1 stater. Copyright © Trustees of the British Museum (CC BY-NC-SA 4.0) 10 Derek Allen, in his seminal paper "The Origins of Coinage in Britain: a Reappraisal" [Allen 1961], devoted one sentence to British C, saying "British A gives rise to a curious progeny, British C, of which a small hoard was found at Yarmouth, Isle of Man". He doesn't attempt to explain why he thought a Belgic style coin was derived from a Westerham style coin. This claim was repeated without analysis by Claxton [Claxton 1999]. Bean [Bean 1994, Bean 2000] states "It has long been recognised that British C is an off-shoot of British A (e. g. Allen 1960,106). Mackensen declined to identify either A1 or A2 as the parent". His comment could be read that Mackensen couldn't decide between A1 or A2 as the parent, but in fact Mackensen said it wasn't British A at all: "Such an evident typological relationship for example between British B1 and B2 cannot be noticed between British A1/A2 and C" [Mackensen 1974]. Bean did try to link British A2 to C via the direction of the leaves on the wreath, the metrology, and the ornament above the horse's tail, saying it looked like the horse's head from A2, but most of these don't stand up to analysis. The leaves do point down like British A2 (which is in fact two distinct series; British Ad and British Af), but the metrology doesn't match (average weights of 6.06g vs. 6.17g for A2 and C), and the object above the tail has an axe head where the mouth should be and three vertical lines where the A2 horse's head has two. Bean also comments that "British C appears to be produced from A2 [...], debased using fixed quantities of silver (about 11%) and copper (about 11%)", but this isn't obvious from the numbers. British C has an average alloy of 28.9%, 47.4%, 23.6% (gold, silver, copper). British Ad has an average alloy of 47%, 36%, 17% which is substantially too much gold. The other part of British A2, British Af, has an average alloy that starts as 55.4%, 34.2%, 10.4% but ends as 8.7%, 56.5%, 32.9%. Bean's mean values for A2 were 47.02%, 36.49%, 16.41% and for C were 28.94%, 47.27%, 23.73%, which show an approximate 18% decrease in gold, 11% increase in silver, and 7% increase in copper. Taking average weights into account, these normalise to 2.88g of gold, 2.23g of silver, and 1.0g of copper per A2 stater, and 1.79g, 2.92g, and 1.47g per British C stater. There's no obvious relationship between the alloys which would suggest C was produced from A2. Bean's analysis may have been influenced by a single British C stater which he lists with an alloy of 63.56% gold, 10.86% silver and 25.55% copper. This was a transcription error [Sills 2017] in Northover's paper [Northover 1992] and the correct value should have been 27%, 14.8%, 55%, which is the plated example. The only link between British A2 and British C is the direction of the leaves, but they also point down on GB-Ce, which has other features in common with British C. 27 The reverse features an articulated horse with windmill like forelegs. The tail is a triple crescent design and one of the arms of the forelegs ends with a triple claw. There's a pellet triad between the windmill arms, another below the horse, and a third merged with the tails A pellet rosette sits above the horse. There's a three stringed lyre like object above the horse, but it's not particularly clear what it might be, and no coin shows its full extent. Reverse 1 Reverse 2 Figure 38: Lyre object above horse's tail. Copyright © Trustees of the British Museum (CC BY-NC-SA 4.0) On reverse 1 it is straight edged with the hint of a curve and talon on the left hand side. On reverse two it's much more rounded and the talon nature is pronounced. The same shape as on reverse 2 appears on both the obverse and reverse of the Horse Geometric quarter stater: Horse Geometric obverse Horse Geometric reverse Figure 39: The "talon" from British C's "Lyre" appearing on the Horse Geometric quarter The talon shaped hair curls on the obverse are mirrored on the obverse of the Horse Geometric quarter in the horse's front legs. They can also be seen on the "hand" like objects, and the "axe" like objects next to them, on the reverse. 28 Talon like hair curl on British C obverse Horse Geometric obverse Horse Geometric reverse Figure 40: The "talon" like hair curl from British C appearing on the Horse Geometric quarter The triple lined nose on British C is similar to the triple lined tail on the Horse Geometric's horse, and the pellet triad appears below both horses. The claw like nature of the horse's front leg on the stater has hints of the two back legs on the quarter. This claw like device is repeated on the reverse of the quarter, as is the horse's tail. The iconography on both the stater and quarter stater can be described as bizarre. The same can also be said of GB-Ce, where both the staters and the quarter staters deviate from the norm in strange ways. One characteristic of GB-Ce (class 2) is short spikes added to many of the design elements. These can be seen on the British C hair locks, but they are more prevalent on the Horse Geometric quarter, where they can be seen on the horse's hind legs, the crescent in front of the horse, and on the "thunderbolt" on the reverse. The pellet field above the horse on the quarter is similar to that on a GB-Ce quarter: Figure 41: The pellet field above the horse on Horse Geometric and GB-Ce2. Photo Copyright ognnumismatique.com Both the British C stater and the Horse Geometric quarter stand out from contemporary coins due to their strange iconography, and there are enough similarities between them, and with the GB-Ce coins that British C is influenced by, that it's feasible that they are companions. 29 Concerns The previous section has worked with the assumption that the coins are genuine. However, the authenticity of these coins is doubted by some and concerns have been raised11. This section will address each concern in turn. The methodology used is to identify an authentic coin that shares the same concern. This nullifies the concern as it cannot be used to condemn a coin if it can also be applied with no detriment to a known authentic coin. All From the Same Reverse Die NOTE: A coin from a third reverse die was found after the original paper was published. There is no need to go through a similarly detailed analysis of it so this section has been left as it was. It has been noted that all of the Horse Geometric quarter staters share the same dies. While this may be acceptable for obverse dies, it is considered suspicious for reverse dies. The first point to note is that the coin auctioned by DNW in 2008 comes from a different reverse die from the other coins. Both photos have been sized and aligned on the upper middle large bar feature. This reverse (DNW) is the only one that can't be fully aligned with the others. It's possible that some of the feature changes at the edges may be caused by the reverse being extremely concave but the photograph doesn't suggest that it is. It's hard to imagine the mechanism that would cause some of the shapes to rotate clockwise like they have, rather than elongate and spread on their current axes. It's also hard to imagine the mechanism that would cause objects to change distance from each other, or change shape. Figure 42: Angles and positions change. Photo Copyright DNW 11 Although it should be noted that the concerns have not been published for scrutiny. The concerns addressed in this paper have been passed to me "second hand" by other interested parties. If there are additional concerns, they have not been made publicly. 30 Figure 43: Different join point Figure 44: Different attachment point and new segment Figure 45: Missing pellet 31 Figure 46: Missing lines at edge Figure 47: Change of shape and no longer attached to neighbouring object Figure 48: Change of shape and change of distance to neighbouring object 32 Figure 49: No longer joined to neighbouring shape Figure 50: The DNW reverse with the outline of die 1 overlaid in black There are other miscellaneous changes where the distance from the end of a crescent to another object changes. The last figure shows that the outline from die one cannot be made to overlay the design on die two just by scaling and rotating. It can be made to do so for every other coin. It could still be argued that having 93.3% (14 of 15)12 of the coins from the same dies is a concern, so attention is drawn to the British Qc quarter stater "Petersfield Wreath" (ABC 773). This type has between forty four and fifty three extant coins13 and between 93.2% and 94.3% come from a single 12 There are nineteen Bearded Heads listed in the CCI, which may or may not include the DNW coin. If it is included, then the worst case is 18/19 = 94.7% from one reverse die. If it isn't included then it's 19/20 = 95% from one reverse die. Either scenario is still better than the Petersfield Wreath. 13 Forty one coins with reverse die two are listed in Divided Kingdoms [Sills 2017] and a further nine were found online. There are ten coins in Divided Kingdoms that don't have publicly available photograph, so some of the nine new coins might already be included. That means there are between forty one and fifty coins from reverse die two, and forty four to fifty three coins from all reverse dies. The equates to 93.2% (41/44) to 94.3% (50/53) of coins coming from one reverse die. 33 reverse die (the remaining three coins all come from different reverse dies). The provenance data shows that they didn't come from a single hoard. In light of this, the number of dies cannot be used to cast doubt on the authenticity of these coins. They Don't Really Fit in With Anything Else It's not clear where this argument originates, and the iconography, metrology and metallurgy point to this being British C2 (see "Associated Stater" on p18). The reverse is a clear derivation of the standard geometric design used on a number of other quarter staters. The obverse design of either a bearded head or a horse is unusual, but no more unusual than the Gallo-Belgic A obverse was at the time, or the "à l'astre" obverse used on DK 180, or on any of the other coins which deviated from the standard "abstract Apollo's head"14. In light of this, the experimental obverse design cannot be used to cast doubt on the authenticity of these coins. They're All Curiously Well Centred This argument has been checked by measuring the displacement of the centre of the obverse and reverse dies from the centre of the flan, and comparing it against the British Qc "Petersfield Wreath" quarter stater (ABC 773)15. The full design for each die was drawn using multiple coins, and was overlaid on photos of every extant coin. The centre of each flan was calculated, and the distance between the centre of the die image and the centre of the flan was measured for each coin. 14 See ABC 542 548, 551, 623, 626, 629, 632, 635, 638, 770, 779, 782, 785, 788, 791, 794, 797, 800, 2237, 2246, 2249, 2252, and 2583, along with DK 319, most of the Iceni gold, and most of the Dubonni gold as examples of non-standard obverses 15 Only coins with reverse die 2 were used. There are forty one such coins listed in Divided Kingdoms, of which thirty one have publicly available photographs. In addition, nine coins were identified which weren't listed in Divided Kingdoms. Assuming that these are not duplicates of the ones where photos aren't available, forty of the fifty known coins were analysed. 34 Figure 51: Complete image of dies formed and the centre marked with a small square. This is the centre of the die Figure 52: Coins overlaid with die design to find centre of die on the flan 35 Figure 53: Coin photos converted to vector objects and a blue circle aligned to the centre. This is the centre of the flan (based on bounding box) Figure 54: A line is drawn (auto connected to centres) between the square marking the centre of the die and the circle marking the centre of the flan 36 Figure 55: The line is measured (auto snap to end points) to measure the distance in pixels between the centre of the flan and the centre of the die The magnitudes of the displacement vectors are in the following tables expressed both as a percentage of the die size and as the physical distance in mm. 37 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% Petersfield Wreath Obverse ● ● ●● Horse Geometric Obverse Petersfield Wreath Reverse Horse Geometric Reverse ● ●●●●● ●●●●●● ●● ●●●● ● ●● ●●●●● ●●●● ●● ● ● ● ● ●● ●● ●● ● ● ● ●●●● ●●● ● ●●●● ●● ●● ●●● ●●●● ●●●● ● ●●● ● ● ●● ● ● ● ● ●● ●●● ● ●●● ● ●● ●● ● ● ● ● ● ● ● Figure 56: Displacement vector magnitude for obverse and reverse images as a percentage of the die size mm 0.00 0.50 1.00 1.50 2.00 2.50 3.00 mm 0.50 1.00 1.50 2.00 2.50 3.00 3.50 Petersfield Wreath Obverse ●●●● ●●●●●●●●●●●● ●●●●●●●●●●● ●●●●●●●●● ●●●● Horse Geometric Obverse ● ●● ●●●●●●● ● ●●● Petersfield Wreath Reverse ●●●●●●●●●●●●●● ●●●●●●●●●●●● ●●●●●●●●● ●●●●● Horse Geometric Reverse ●●● ●●●●●● ●●● ● ● ● Figure 57: Displacement vector magnitude for obverse and reverse images in mm 38 It's clear from both of these tables that the Horse Geometric quarter stater has marginally worse centring on both sides than the Petersfield Wreath quarter stater. In light of this, the die centring cannot be used to cast doubt on the authenticity of these coins. Reverse Die Engraved with False Die Flaws It's unclear what the exact concern is here. Note that Sills [Sills 2017] says the following about the Ab1 Tarring stater: "An oval die flaw develops below the hair bar on obverse 1 and subsequently becomes incorporated into the design as a pseudo flaw", p102 "reverses 5 to 7 have small, irregular pseudo flaws in and around the body", p102 "Pseudo flaws are a characteristic feature of Tarring staters and their presence on Geometric quarters helps to equate these otherwise very different types", p103 It appears that false die flaws are both present and useful on genuine coins, so their presence on the Horse Geometric cannot then be used to cast doubt on their authenticity. Note that it's not clear which coins have the false die flaws, and what they are. None Have Been Found in Hoards with Other Coins The argument here appears to be that genuine coins should turn up in the hoard record, and by not appearing, the authenticity of a coin can be doubted. "Coin Hoards in Iron Age Britain" [de Jersey 2014] does not list the Horse Geometric type as appearing in any hoards. There are two other coins with similar populations that are considered genuine and don't appear in hoards with other coins. The first is ABC 1764 "Lincolnshire Boat Tree". There are at least fourteen extant coins of this type, and none have been found in a hoard. The second coin is the British C "Yarmouth" stater. This has been found in a single hoard, but there were no other coins included. The circumstances of discovery are unknown, as is the exact location and number of coins found. Despite this there is no doubt about their authenticity. John Sills [Sills 2017] (p178) says that it is “one of the most enigmatic British stater types”, and nine of the fifteen class 1 coins are in the British Museum (8) and the Ashmolean museum (1). It would be extremely useful to find a Horse Geometric quarter stater in a hoard with other coins, but it doesn't appear to be a requirement with other coins. The initial controversy about Horse Geometric quarters seems to have been caused by the initial finders lying about the find spot to conceal an undeclared hoard, and that doesn't seem too far removed from the British C situation; a hoard consisting of just that type from an unknown location. It might be argued that British C must be genuine because of the individual finds, but there are individual finds of the Horse Geometric quarters as well. The hoard argument cannot be used to cast doubt on the authenticity of these coins unless the authenticity of British C and ABC 1764 "Lincolnshire Boat Tree" are also under doubt. 39 The Finders All Know Each Other This is not correct. There are at least four distinct groups of finders who don't know each other: 1. The original finder(s) who sold coins to Chris Rudd 2. MG who published an article on finding his coin in the Feb 2019 Treasure Hunting magazine 3. JW, BH, & HF who found coins independently on different sites, but who know of each other 4. JB whose post about his Horse Geometric quarter stater was seen on Facebook This cannot be used to cast doubt on the authenticity of these coins. We Need Several Finds from Different Places See "Find Spots" on page 9 above. Five find spots in addition to the original false ones are now known. This cannot be used to cast doubt on the authenticity of these coins. None Have Been Recorded on PAS See SUSS-BCFB4B and SUR-C1DAE6. Note that one recent find was not recorded on PAS because the finder thought there would be no point sending it off just to be told it wasn't genuine. This cannot be used to cast doubt on the authenticity of these coins. Too Sharply Struck and Lack the 'Softness' of Genuine Iron Age Coins It's not clear what is meant by this objection. A Horse Geometric quarter stater has been compared in-hand against eighty three Belgic and British unscribed quarter staters and seventeen Belgic and British staters, and there is no obvious difference in either sharpness or softness (ignoring the effects of worn dies). These comparisons have been done using a 10x-45x stereo microscope, and from photographs taken using a 100mm macro lens with three extension tubes to allow closer focus. The following photograph is an unscaled segment of one of these photos. 40 Figure 58: Unscaled photo of part of a Horse Geometric reverse The next photo shows the Horse Geometric quarter stater (top left) compared against five other unscribed British quarter staters, including a Hampshire Thunderbolt, an Ad2 Class 3 Geometric, and a Duro Boat gold. 41 Figure 59: A Horse Geometric quarter alongside five other unscribed British quarters No visual difference is apparent. Without further clarification of the concern, this cannot be used to cast doubt on the authenticity of these coins. They Appeared Around the Same Time as a Large Group of Fake Cheriton Smiler Staters A large group of fake Cheriton Smiler staters appeared in the trade in the autumn of 1991. The first coin appeared in the trade in Münzen & Medaillen 1991 fixed price list. The concern that the Horse Geometrics were just another attempt at introducing modern fake coins into the market cannot be ignored or easily nullified. What we can say is that the six coins found in recent years are, on the balance of probabilities, genuine. The mechanisms for the recent single finds being fake would either be a number of trusted metal detectorists working in collaboration to risk their reputations over coins that aren't really that valuable16, or the forger burying coins in very random and remote locations in the 1990s in the hope that they would someday be discovered17. Neither possibility makes rational sense. If the recent finds are genuine, then at least one of the original coins must also be genuine as any fakes would have had to have been made from something 16 One dealer who believes they are genuine had one on sale for £1,485. The metal detectorist who sold it to him would have been paid less than this. One sold privately in 2019 for £500, and one shortly before that for £250. At these levels it's simply not worth getting a reputation for deliberately selling fake coins 17 Quite how he/she intended to monetize someone else finding his/her fakes is unknown. As it has taken 25-30 years and advances in detector technology for these new coins to be found, it would appear to have been a poor strategy. 42 real. Which of the finds from the 1990s might be fake is unknown; all have been well examined by experts and none have been condemned. This concern cannot be used to cast doubt on the type overall, and the existence of a fake coin can't condemn the type. Conclusions Concerns about the authenticity of the Horse Geometric quarter staters have lingered, unpublished and unchallenged, since shortly after Chris Rudd advertised one for sale in 1995. A quarter of a century has now passed and it is time to examine the arguments that have been made against these coins. This research has taken two approaches to resolving this situation. Firstly, it has examined the iconography, metallurgy, metrology, and find spots, and shown that they are all consistent with genuine coins, and that the evidence points to the Horse Geometric quarter staters being minted alongside the British C "Yarmouth" stater. There are enough similarities between these two types that considering one as genuine and the other as fake is a difficult position to take. Secondly, it has examined the concerns that have been raised about the Horse Geometric quarters, and shown that all can also be applied to genuine Iron Age coins. It follows that the concerns cannot stand. The argument that a type of coin is problematic for some reason while another isn't even though the same reason applies cannot hold. The Horse Geometric quarter staters fit well into the British unscribed series and none of the concerns raised survive close analysis. While doubts may have been justified at the time of the original finds, there is no longer any reason not to accept them as genuine Iron Age coins. While it would obviously be useful for a Horse Geometric quarter stater to be found by a professional archaeologist in an archaeological context, that is not a standard that has been applied to the plethora of other new types that have appeared in academic books and auction catalogues in the intervening time. If any credible doubts about the authenticity of these coins remain then they should be published in a freely available format so that they can be examined by all interested parties. 43 Appendix 1: Catalogue This work is based on the following fifteen coins. There are more coins in the Celtic Coin Index, but only information up until around 2004 was publicly available at the time the original paper was written. It was assumed the extra CCI coins did not change the conclusions and that was correct. The coins have been placed into order based on the deterioration of the dies, but there are limits to the accuracy of this based on photos alone, and the order may ultimately change. Reverse Die 1 Reference Photo & CCI Comments Weight and Provenance Diameter Found by "MG" on the Ridgeway "in the middle of nowhere on the downs on private land" MG Video is YouTube SUSSBCFB4B on 1.46g 11.87 mm Vosper 2020 1.3g Found by "JB" in Westhampnett ca.12mm 44 96.1797 1.3g 96.1798 (CCI Link) 1.3g Lat: 50.846839; Lon -0.751392 CCI Comment: "In trade (C Rudd 2161). False provenance of Westhampnett." CCI Comment: "Good condition. Genuine or fake?" 97.0807 (CCI Link) 1.36g CCI Comment: "Was in trade (C Rudd 2737)." 1.4-g CCI Comment: "In trade (Mike Vosper 1998 list 100, no. 4)." CCI Comment: "Slightly cracked obverse. Fake or genuine?" 98.0724 (CCI Link) CCI Comments: "Bearded head quarter - genuine? From the usual dies. See 96.3353 etc" 45 "Weight given to 1 dp." SURC1DAE6 1.2g Lat: 50.96473746; Lon: -0.84791877 10.7mm HF 1.33g Lat: 50.944490; Lon: -0.48540473 Found in Wiggonholt West Sussex near Pulborough. 97.0804 (CCI Link) 1.35g CCI Comment: "In trade (M&M Basle, PSL 549, Nov/Dec 1991, no. 137)." CCI Comment: "Fake or genuine?" 96.3353 (CCI Link) 1.42g CCI Comments: "Cracks in edge of flan. Fake or genuine?" "Analysis by Northover, 1996." 24.8% 44.5% and 30.4% gold, silver and copper 46 95.1195 (CCI Link) 1.3g Lat: 50.846839; Lon: -0.751392 Westhampnett, Chichester, West Sussex Chris Rudd list 19, 1995 CCI Comments: "False provenance of Westhampnett bypass." "Cracked across obverse. Genuine or fake?" 97.0806 (CCI Link) 0.71g CCI Comment: "Was in trade (C Rudd 2702)" CCI Comment: "Half of coin, much of obverse worn. Fake or genuine?" 97.0805 (CCI Link) CCI Comment: "In trade (Blancon, Hannover, PSL 13, 1993-94, no. 886)." CCI Comment: "Fake or genuine?" Reverse Die 2 Reference Photo & Information Weight and Diameter Provenance 47 DNW 29-092008 Lot 5127 1.25g From the collection of Michael J. Grover 48 Appendix 2: Die Offset Calculations The photos of the Horse Geometrics and the Petersfield Wreaths were all scaled to a common size in Photoshop. All of the photos which had CM scales were sized and aligned first, and then all of the ones without were scaled to these. Various photos were merged to show the full extent of each die, and a vector drawing of each die was created in Inkscape. Each coin had the die images overlaid and aligned, showing the centre of the die on the flan. Each coin was then converted to a vector object by Inkscape, and the centre marked by automatically aligning a small blue circle. The end result of this process is two marks for each side of each coin. One showing the centre of the flan, and the other showing the relative position of the centre of the die. These were connected using a line that snapped to the centres of these objects, and measured using the ruler tool. The vector lengths, measured in pixels, show the displacement between the centre of the flan and the centre of the die. These were converted to mm by measuring one of the CM scales from one of the images. The maximum of each die's width or height was used to calculate the "percentage of die" displacement. The 57 mm of CM scale measured as 4040.99 pixels. This was used to convert pixels to mm in the following tables. The maximum dimensions of each die were: PW Obv Die Diameter (pixels) PW Obv Die Diameter (mm) 880.88 12.43 PW Rev Die Diameter (pixels) PW Rev Die Diameter (mm) 788.47 11.12 BH Obv Die Diameter (pixels) BH Obv Die Diameter (mm) 806.04 11.37 BH Rev Die Diameter (pixels) BH Rev Die Diameter (mm) 821.76 11.59 The Die centring displacement for the Petersfield Wreath is shown below. Note that there is one extra reverse result because the obverse of one of the coins was too badly photographed to use 49 Petersfield Wreath Obverse Length (pixels) Length (mm) % of die 142.72 2.0 16% 67.95 1.0 8% 54.24 0.8 6% 47.65 0.7 5% 96.02 1.4 11% 18.61 0.3 2% 91.91 1.3 10% 81.44 1.1 9% 63.13 0.9 7% 67.89 1.0 8% 64.27 0.9 7% 36.23 0.5 4% 107.19 1.5 12% 46.66 0.7 5% 131.84 1.9 15% 102.56 1.4 12% 107.97 1.5 12% 97.14 1.4 11% 156.98 2.2 18% 104.05 101.58 89.86 160.15 74.81 13.46 73.77 42.43 24.25 39.69 81.87 122.25 113.54 47.19 132.26 41.12 155.51 123.83 127.1 108.27 1.5 1.4 1.3 2.3 1.1 0.2 1.0 0.6 0.3 0.6 1.2 1.7 1.6 0.7 1.9 0.6 2.2 1.7 1.8 1.5 12% 12% 10% 18% 8% 2% 8% 5% 3% 5% 9% 14% 13% 5% 15% 5% 18% 14% 14% 12% Petersfield Wreath Reverse Length (pixels) Length (mm) % of die 34.91 0.5 4% 18.31 0.3 2% 46.14 0.7 6% 17.17 0.2 2% 67.99 1.0 9% 99.24 1.4 13% 52.66 0.7 7% 72.87 1.0 9% 18.62 0.3 2% 27.08 0.4 3% 118.34 1.7 15% 94.8 1.3 12% 66 0.9 8% 26.47 0.4 3% 25.91 0.4 3% 62.04 0.9 8% 15.13 0.2 2% 46.15 0.7 6% 62.55 0.9 8% 78.57 1.1 10% 24.15 0.3 3% 70.84 1.0 9% 18.52 0.3 2% 24.77 0.3 3% 126.81 1.8 16% 34.09 0.5 4% 23.87 0.3 3% 17.29 0.2 2% 43.02 0.6 5% 77.77 1.1 10% 84.82 1.2 11% 108.22 1.5 14% 130.15 1.8 17% 43.09 0.6 5% 73.5 1.0 9% 96.26 1.4 12% 67.68 1.0 9% 65.73 0.9 8% 80.21 1.1 10% 123.95 1.7 16% 50 Die centring displacement for the Horse Geometric: Bearded Head Obverse Length (pixels) Length (mm) % of die Bearded Head Reverse Length (pixels) Length (mm) % of die 28.22 0.4 4% 146.57 2.1 18% 93.09 1.3 12% 44.97 0.6 5% 79.61 1.1 10% 68.63 1.0 8% 72.38 1.0 9% 34.89 0.5 4% 85.98 1.2 11% 87.68 1.2 11% 77.37 1.1 10% 62.61 0.9 8% 41.73 0.6 5% 78.63 1.1 10% 166.24 2.3 21% 210.25 3.0 26% 145.75 2.1 18% 72.36 1.0 9% 60.39 0.9 7% 31.41 0.4 4% 134.57 1.9 17% 52.1 0.7 6% 82.42 1.2 10% 17.7 0.2 2% 171.52 2.4 21% 54.83 0.8 7% 213.12 3.0 26% 60.99 0.9 7% 87.82 1.2 11% 51 References ABC "Ancient British Coins", Elizabeth Cottam, Dr. Philip de Jersey, Chris Rudd, Dr. John Sills. 2010 Allen 1961 "The Origins of Coinage in Britain: a Reappraisal", Derek Allen. Paper presented in 1958 and published in [Frere 1961] Bean 1994 "The coinage of Atrebates and Regni", Simon C. Bean. 1994 Bean 2000 "The coinage of Atrebates and Regni", Simon C. Bean. 2000 Claxton 1999 "An iconological analysis of British gold staters c. 80 BC - AD 45", J. Claxton. 1999 de Jersey 2014 "Coin Hoards in Iron Age Britain", P. de Jersey. 2014 Frere 1961 "Problems of the Iron Age in Southern Britain", S. S. Frere (ed.). 1961 Note that this is frequently cited as a 1960 publication Mackensen 1974 "Die älteste keltische Gold und Silberprägung in England", M. Mackensen. 1974 Mays 1992 "Celtic Coinage: Britain and Beyond. The Eleventh Symposium on Coinage and Monetary History". M. Mays (ed.), BAR 222. 1992 Northover 1992 "Materials issues in the Celtic coinage", P. Northover. 1992. In [Mays 1992] Sills 2017 "Divided Kingdoms: The Iron Age Gold Coinage of Southern England", Dr. John Sills. 2017 VA 1993 "An Anomalous Yarmouth Stater", Robert D. Van Arsdell. Published in Spink's Numismatic Circular. 1993 52