Coin hoards have always formed a major subject of study for numismatists. Research can focus on the hoards themselves, what their distribution in time and space indicates about the processes and motivations for hoarding, or it can examine what the contents of those hoards can tell us about production, circulation and the coinage pool.
The Beau Street hoard was found during excavations in Bath in 2007. It contained 17,660 coins, a mixture of denarii, radiates and debased radiates. Most of the coins date between a.d. 222 and 274. It consisted of at least eight bags of coins along with a large number of loose coins that may have been in bags that had completely decayed. This volume consists of five short introductory chapters followed by the main catalogue.
Chapter 1 by Abdy provides an overview of the hoard. A brief summary of the relevant monetary history is followed by a discussion of the hoard contents compared to other hoards. The term ‘radiate’ is used for the pre-debasement coins up to 260 and ‘debased radiate’ for those struck after 260. The hoard has comparable quantities of earlier coins (77%) compared to the Stevenage (75%) and M1 (85%) hoards, but Cunetio (21%) and Bassaleg (45%) have far fewer. This is clear evidence of the deliberate selection of earlier, more intrinsically valuable coins. Abdy then compares the hoard to Pot IV from the Chalfont St Peter hoard. It would have been informative to have had the breakdown of the contents of Pots I–III presented as well. Abdy uses the French word thésaurisation (to accumulate, hoard) in his discussion but does not define in what sense he is using it, probably meaning ‘[coin] selection’.
Chapter 2 by Mark Corney describes the find-spot of the hoard and its archaeological context. In some places a greater integration between text and image is needed (where is heated suite A, for example?). The hoard was found in a stone-lined cist below floor level, although the floor had been destroyed.
Chapter 3 by Benedict Sayers discusses the contents of the eight bags and the loose coins. Each bag contained mainly one type of coin: bag 6 had mainly denarii, bags 1, 3, 4 and 5 silvered radiates and bags 2, 7 and 8 debased radiates. The loose coins look very similar to the bags of silvered radiates. Edward Tufte (The Visual Display of Quantitative Data, Cheshire, CT, 1983) writes that ‘A table is nearly always better than a dumb pie chart; the only worse design than a pie chart is several of them, for then the viewer is asked to compare quantities located in spatial disarray, both within and between pies’ (p. 178). This chapter provides an excellent example of the problem. Ten pie charts are spread over seven pages adding nothing to the reader's comprehension. A second table showing the data given on page 14 as percentages with the rows in the same order as the discussion would have been a much clearer illustration. Sayers discusses the chronological distribution of coins using line graphs and provides a distribution by mint. It is impossible to assess these patterns without comparative data from other hoards. Comparison against a background pattern is one of the essential lessons taught by Richard Reece in his many publications. Many questions remain unanswered: do the individual bags look like hoards with the same closing date or are they unusual in composition? Does the overall hoard look unusual? Does the pattern of bag compositions match the sequence in which they were deposited in the cist?
Julia Tubman describes the micro-excavation, recording and conservation of the coins. The dismemberment of the bags to preserve the groups of coins adds hugely to the academic value of this hoard. The only omission is a discussion of the possible sequence in which the bags were placed in the cist. The final chapter is a note by Kevin Butcher and Matthew Ponting on their ongoing work on the archaeometallurgy of Roman coinage.
The remainder of this book is a detailed catalogue with copious plates that are extremely welcome to anyone dealing with less well-preserved material, although some are rather too flatly lit.
The second volume is a consolidated version of five Presidential Addresses given by Bland and published in the British Numismatic Journal, topped and tailed by additional chapters and completed by an extensive ‘check list’ of Iron Age and Roman hoards which occupies two-thirds of the volume. We should join Bland in acknowledging the enormous contribution of Eleanor Ghey in collecting the data from which this check list was created, and on which Bland's analyses are based.
The volume is attractively typeset, although there are several editing errors, and copiously illustrated with 33 colour plates at the back. The photographs vary in quality from the glorious colour plates of hoards such as Sandridge (pl. 30) to the photograph of the Chalfont St Peter hoard which consists of three grey blobs on a grey background with no scale (fig. 2.10).
My biggest complaint is the presentation of the data. Thankfully, ‘dumb pie charts’ are rare, but extensive use has been made of pointless 3D effects. These effects in single series graphs (e.g. fig. 5.30) make seeing the values fiddly, but not impossible, unlike the graphs arranged in three series (e.g. pl. 8) where some of the bars hide those behind, making it impossible to see the data. They would be more effectively presented as simple 2D side-by-side bar charts such as pl. 33. The late Mike Baxter – a statistician with extensive interests in archaeology – argued that 3D graphics such as these should be banned from academic publications, a sentiment with which I wholeheartedly agree. Additionally, none of the graphs have their y-axes labelled. Mostly one can work it out from the text or the captions, but the point stands. The maps are excellent, but it is frustrating when trying to follow the flow of the argument to have some maps in the text and some in the maps section at the back, along with some graphs in the text and some, needlessly in colour, at the back.
Chapter 1 provides an overview of the development of coin-hoard studies. Bland acknowledges the contribution of key scholars including Richard Reece and John Casey but omits to note (p. 4) that Casey devised an alternative set of 29 periods to Reece's 21.
Chapter 2 discusses some of the key themes in hoard studies. In answer to the question ‘what is a hoard?’, for example, Bland opts for ‘a group of objects which have been deliberately brought together.’ It is worth considering, however, the value of the objects. The late Sara Champion pointed out to me that a single low-value bronze coin is likely to be an accidental loss, whereas a single gold coin is unlikely to be. The dating of hoards is also an issue. The hoard ontology developed by the Nomisma project defines the ‘closing date’ of a hoard as the date of the latest coin and the terminus post quem as the date after which the hoard was deposited taking all the possible sources of evidence into account. The Bredon Hill hoard discussed by Bland (p. 18, cf. p. 69) is an excellent example of this problem.
Bland also discusses the enduring argument that links periods of unrest with the incidence of hoarding. He cites the work of John Kent and Edward Besly on the hoards of the English Civil War and their competing views as to how the events in that war are reflected, or not, in the evidence. This particular case study cries out for formal spatial analysis to examine underlying factors, such as areas of recruitment or heavy fighting.
The chapter concludes with an examination of two historic examples of hoarding in times of trouble: Pepys hoard from 1667 and the Hackney hoard from 1940. The latter consists of two gold hoards that were recovered in 1952 and 2007. They were buried in 1940 in two jars, one containing 82 $20 US gold coins dating to 1890 and the other 80 similar coins dating between 1853 and 1913. Beyond Bland's observation that the hoard was concealed for safekeeping, the facts that the closing date of the two hoards precedes the actual concealment date by 50 years in one case and 37 in the other, and that the coins were not currency, having ceased to be struck in the US in 1933, are both salutary warnings against the simplistic interpretation of hoard data. In both examples, the immediate impetus for concealment was the threat of invasion which, in neither case, actually happened.
Chapters 3 to 7 examine successive periods and possible interpretations for the observed patterns. Chapter 3 looks at the Iron Age to Roman transition and the vexed issue of the relationship between Iron Age and Roman coinage. Chapter 4 looks at hoards from a.d. 69 to 238 and examines the possibility that the pattern matches historical events. Chapter 5 examines radiate coin hoards of the third century and competing economic versus ritual explanations. Chapters 6 and 7 look at late Roman and post-Roman hoards, and their relationship to the end of the province. The volume concludes with Ghey's valuable check list, which is usefully indexed.
These two volumes contain much valuable information, and a series of ideas which could be pursued in more detail. I do hope, however, that more effective means of data analysis, such as relative risk maps will be bought to bear on these data (see A. Bevan, ‘Spatial methods for analysing large-scale artefact inventories’, Antiquity 86 (2012), 492–506). As the size of our dataset continues to grow, we need to adopt more effective means of data analysis than the rudimentary bar chart and dot-distribution map if we are to make the most of this rich resource.
