As B. makes clear from the outset, this is a very personal ‘Handbook’ and should be understood and used as such. It came about because he felt the need for a series of plans of Roman legionary fortresses to common scales and conventions, a need many will share and be very grateful that he has catered to with such clear and elegant drawings, albeit to quite a small scale. These could only be achieved, as he points out, by a great deal of work trying to reconcile plans, part-plans and sketches to a variety of scales and conventions, inevitably involving a certain amount of judicious ‘fudging’. Obtaining and reconciling the base-plans for so many fortresses from so many countries and in so many languages must have been an immense labour, one concealed with due modesty but which deserves to be saluted. These plans form the core of the book and will stand as a resource for many years to come; they are also available on the website accompanying the book – legionaryfortresses.info. The plans only cover the defences and the area within them, though extra-mural features are outlined in the brief Chapter 7. What is depicted and reconstructed on the plans is generally very judicious in evaluating the reliability of the information available and thus the extent to which information may reasonably be extrapolated. That said, Inchtuthil is there in its full, extrapolated and reconstructed glory (including the mess in the right latus praetorii), though the plan is usefully a version that shows where structures were actually encountered. The only example I noted where extrapolation seems to have run well away from the evidence was the basis for the reconstruction of eight cohorts-worth of barracks at Lincoln-Lindum.

With each fortress (including those known to have existed but for which no plan can yet be reconstructed) comes a gazetteer in summary form including: location in ancient and modern administrative units; dimensions; orientation; dating; ancient literary and sub-literary references; modern references as well as units attested. This last needs a note of caution; as B. states this is not a list of all units ever in garrison at the site, it is a list of units mentioned in the epigraphy from the site, so a single tombstone of a much-travelled soldier or veteran could boost this list significantly. Units attested as being in garrison are marked in bold. A huge amount of information is given, but equally hugely compressed.

The plans and gazetteers are prefaced by eight introductory chapters covering topics such as history and development of the bases, defences, internal buildings, infrastructure, construction (and demolition). Like the gazetteers these are also very compressed, 41 pages in all. What is said is again a model of clarity, but it is often what is left unsaid that would have been more interesting; topics such as the debate over valetudinaria or varieties of store-buildings or problematic buildings or the accommodation of auxiliary formations are adumbrated and then swiftly moved on from. At the end of the book is a series of appendices covering topics such as the movements of individual legions over time or the legionary locations listed in the Notitia Dignitatum. There are then the notes by chapter, followed by a magnificent bibliography which like the plans must also have taken an enormous amount of compilation. In the centre of the book is a fold of useful colour plates, largely taken by B. (what Dura-Europos is doing there is unclear, though it is nice to see wall-top traverses).

As will by now be appreciated, this book is an enormously useful resource, for which we should thank B. very warmly. It is therefore most unfair that one's principal reaction is to beg for more, especially since as noted above it is a Handbook that sets out to do what its author wanted it to do. But given the sheer amount of knowledge and expertise deployed here, this is rather like a recipe book that gives the ingredients but does not then actually tell one what to do with them. This works at various levels. One is the problem of what constitutes a legionary base. There are the well-known first- and second-century a.d. sites of around 20 ha., which could have held a full legion, but then there are other sites such as Usk, which is of the right general size but whose excavated internal buildings do not conform well to the ‘expected’ plan. Moreover, as Appendix 4 states, that curious penumbra class of ‘vexillation fortresses’ has been excluded. But they may relate to points which B. rightly makes explicit: did a legionary base only hold one complete legion (as so often implied), or were there also auxiliaries (yes, as at Neuss and elsewhere) or were parts of the legion out-stationed as when wall-building or in times of active campaigning (cf. the Vindolanda renuntium for an auxiliary unit half of which was somewhere else), and if a vexillation (?with auxiliaries) was detached for some time, what would the base for such a group look like? B. would probably argue that since these did not contain the legionary eagle they were not the base of the legion (but how does one tell where the eagle was?). There are other major questions that B. touches upon, briefly. The whole topic of when and where the ‘classic’ legionary fortress was devised; Spanish sites such as Cáceres el Viejo may be in the evolutionary line but the Augustan period still seems formative. Yet what of probably post-Caesarian sites in northern Gaul such as Blanc Mont at Folleville or Liercourt-Erondelle; what would their interiors yield? The extreme shortage of Republican bases in areas of long-term military commitment such as Spain remains a problem and a puzzle. At the other end of the chronological range, the book seems less comfortable once past the early-third century and sites such as Albano. The major changes to unit structures by the fourth century meant that though there were things called legions they did not look much like their early-imperial predecessors. Late legionary fortresses there were, such as El-Lejjun, Luxor, Palmyra or Richborough, but they were very different beasts, so do they belong in this book or in a separate one? What happened inside early-imperial fortresses through the third and fourth centuries is something we still need to know much more about in Britain and beyond. Another level at which we are left wanting more is that of the individual bases. Sometimes it is as simple as explaining oddities such as the embayment in the line of the defences at Anreppen or the extraordinarily wide streets at Haltern or Oberaden, or the odd things that sometimes seem to be going on in and behind the scamnum tribunorum. But mostly it is simply a question of identifying buildings, particularly unusual types such as (to take a couple of British examples) what Richmond (mis-)labelled the basilica exercitatoria at Inchtuthil or the long courtyard building behind the principia at Chester (a store-building? – as B. suggests on p. 26).

B. set out to write a book that does what he wants it to do, and in that it succeeds magnificently and we shall be in his debt for a long time. But of course what we want and need now is the companion volume, the one where he can put all his knowledge to use in a discursive discussion of all the points he raises here and more. To return to the culinary metaphor, this list of ingredients leaves us salivating for the delicious things that could be prepared and served up by someone who so obviously knows how to accomplish this.