This book is a significant and valuable addition to small finds research; being a comprehensive, lavishly illustrated study and catalogue of a type of padlock known largely only by those with an interest in locks and keys or, like the authors, an interest in mechanical puzzles.

The subject matter are 156 small (25–42 mm high) copper-alloy rotary padlocks each with a face (mask) on the front cover. In some instances, another mask is present on the back of the padlock. Each padlock requires manipulation of up to three component parts (e.g. lifting a small latch) before the lock can be opened with its key. Determining and finding what ‘tricks’ to manipulate and how is the ‘puzzle’ that, with the mask(s), gives the objects their modern name.

Having highlighted the importance of this work, it must be pointed out that the text is mainly written in an informal style. For example, the authors employ casual phraseology such as ‘… turn into a pile of rust …’ and describe the items to be of bronze despite the results of a metallurgical analysis of 31 (20 per cent) of the 156 padlocks showing copper alloy to be a far more appropriate term. There are also proofing errors and somewhat disappointing editing (e.g. ambiguities, disjointed writing, missing in-text references, inconsistencies, inaccurate cross-referencing, needless repetition — including near-identical paragraphs within a chapter (e.g. pp. 30 and 38 on a hoard of denarii)). These are suggestive of a rushed work and ill-serve this study.

Yet the objects are so fascinating, and their study so important, that such editorial issues with the text can be forgiven. Following a brief summary of previous research, a useful four-part typology is defined. Three of the categories are based on the different shackles encountered (Finger Ring, Square Shackle and Chain Shackle); the fourth on the padlock having two masks (Janiform). But why the authors decided to place the 67 locks (43 per cent of the assemblage) whose type cannot, for reasons of damage/loss of material, be determined with certainty into what was already the most populated category (Finger Ring) is, well … puzzling. Similarly, the classification of a padlock with a single mask (Cat. No. JF-01) as a member of the two-mask (Janiform) category because the facial features are similar to other examples in the category is at odds with my understanding of typological classification. It is also important to note that many of the classifications seem to have been made from drawings and photographs rather than physical examination of the objects. There is also an appendix that classifies 128 (82 per cent) of the padlocks into typological groups based on mask appearance and decoration but, frustratingly, the members of each group are not listed.

The authors struggle to discuss the dating of puzzle padlocks, largely because their material includes so many unprovenanced examples from museums and private collections, but argue that the earliest rotary locks were created in the second to first century b.c., with the majority dating to the later first to third century a.d. They also wrestle with the objects’ cultural affinities, which combine high-level technical design with ‘Celtic’ facial features. Findspots are known for 77 (49 per cent) of the padlocks and these have been plotted to enable us to see the distributions. Strikingly, the one-trick (and supposedly earliest) Janiform locks have a more eastern (including finds from Israel, Lebanon and Turkey) distribution than the three other types, which are mainly from north-eastern Italy (e.g. Aquileia) and the northern frontier (e.g. Brigetio and Carnuntum). However, in a number of instances, this is at no finer granularity than country. Furthermore, some locations are where an object was purchased, potentially distorting the distributions.

Much time and effort has also been spent on well-executed reconstructions that allow us to better understand the padlocks and explore their possible uses. In the main, these padlocks appear to have been a more expensive, but arguably quicker to work and more secure, alternative to seal-boxes. Their main function may have been to secure leather bags but they could also have secured cupboards or boxes. Both seal-boxes and padlocks could have been easily broken by force, but their purpose was to alert the owner to tampering and to prevent opportunistic theft.

The catalogue illustrates all 156 padlocks (almost all by at least one photograph), none of which appear identical. Important data for each object are included (although this is not uniform and frequently sparse) and the omission of weights is disappointing. The volume concludes with sections on metallurgical analysis and care of mask padlocks (the latter perhaps aimed at private collectors), and a guide on how to identify forgeries.

In conclusion, the authors express the hope that this work ‘… will stimulate further study and research about these amazing devices’. It is certainly to be hoped that improved recognition, even of component parts, will increase numbers from Roman Britain, where to date only three examples of mask puzzle locks are known.