Roman history is arguably full of the sound and fury generated by military conflict, with events like the Battle of Cannae being of considerable importance for both ancient and modern observers. However, Roman battlefields as a general rule produce very little in the way of material remains, certainly not enough to allow battlefield archaeology to produce meaningful results. Indeed, the lack of sites that could be positively identified as locations where a battle — as opposed to a siege — had taken place was for a very long time an important if somewhat disappointing characteristic of Roman military history. This changed in 2008. Following a chance find of a Roman hipposandal, the site of a major military action was finally identified in what may be described as the one of the most unlikely of places: at the Harzhorn in Lower Saxony, deep inside the barbaricum and around 95 miles beyond the frontier of the German provinces. Between 2008 and 2020, survey work as well as seven excavation campaigns produced a wealth of material offering important insights into how the Roman army operated on the battlefield — a battlefield about which literary sources are totally silent. There is some irony in the fact that while the Harzhorn site is so far the only battlefield from Roman history where battlefield archaeology is possible, the political and military events leading to the battle are unknown.

The excavation report produced by the editors presents the result of 12 years working on the site, which is still far from fully explored. The publication marks an important step in interpreting the site; so far, the only major publication prior to the excavation report was the 2013 catalogue of an exhibition at the Landesmuseum Braunschweig, where finds from the battlefield were first presented to the public; the catalogue covered only the work done between 2009 and 2012. The report comes in three (consecutively paginated) volumes of over 1,000 pages in all, two volumes of text and one of plates; the text volumes are copiously illustrated as well; there is hardly a page without a high-quality map or a high-resolution picture.

The first of the two text volumes begins with a brief introduction to the project (pp. 1–29), including short but very valuable notes on the methodology of battlefield archaeology by Michael Meyer (25–9). The remainder of the first volume falls into two parts: the first provides an extensive overview over the different archaeological campaigns between 2008 and 2020 (31–131), while the second focuses on the topography of the site and how it changed over time (133–235). It includes a detailed explanation by Geschwinde, Meyer and Lönne of how the ancient landscape was reconstructed, based both on historical and archaeological evidence (pp. 211–19).

The second volume provides an extensive discussion of the finds. Apart from a small number of early Iron Age and early medieval pieces (237–49), the vast majority of the material is of Roman origin; much less Germanic material has been found, of which a group of arrowheads (351–63) is the most prominent. Perhaps unsurprisingly, weapons and armour play a prominent role, but tools, personal equipment, coins, waggon gear and equipment, general horse gear, boxes, nails, shoe nails (most of which are post-Roman in date, however), pottery and bones are also presented in considerable detail. Some individual finds stand out, such as significant fragments of a lorica hamata (375–99), or a well-preserved dolabra with an inscription indicating it once belonged to soldiers of legio IV Flavia felix (413–15). As one would expect from a battlefield, the site yielded large numbers of ranged weapons spent during the action; these include at least four pila (326–31) and 55 arrowheads identified as Roman (332–50). Of particular relevance for anyone interested in Roman army operations is a chapter where Michael Geschwinde discusses 138 ballista bolt heads found at the site which are of three types, one of which has so far been poorly attested in the archaeological record (303–26).

The discussion of the finds is followed by several analytical chapters. Michael Geschwinde and Michael Meyer discuss the difficulties of dating the material (619–31). In a brief but important section, Michael Sommer combs through the numismatic and literary evidence for a campaign of emperor Maximinus during which the battle may have taken place (631–49). Yet while the event can be dated to the reign of Maximinus, it is at present not possible to connect the available evidence from the ground with the literary evidence for Maximinus’ campaign; whatever role the Roman force at the Harzhorn had in the context of the campaign lies in the dark. Michael Meyer discusses the Germanic warband that found itself fighting the Romans; unsurprisingly, very little can be said about where it came from or what its intentions were (655–71). The last major section of the second volume provides an in-depth attempt by the editors at reconstructing the events as they unfolded on the battlefield, based on the archaeological evidence (673–728). Given the almost total lack of historical context, this is a daunting task, and the editors cautiously speak of developing models of what might have taken place rather than trying to prove a firm reconstruction.

An English summary (733–49), an extensive bibliography (751–90) and 16 appendices conclude the second volume. The third and final volume is a catalogue of all the finds discussed in the second volume; it lists 2,806 entries and concludes with 85 plates.

The excavation report of the Harzhorn site provides a wealth of material presented in an excellent way. This quality of presentation has a price, however. Not only are the volumes so large and hefty that they lend themselves to being used in resolving not only scholarly disputes; their price — while justifiable by the sheer mass not only of material inside but of the actual books — is not likely to allow for a wide circulation beyond libraries. Still, they will be indispensable for anyone interested in how the Roman army operated in the field. Given that the Harzhorn site is at present unique as the only Roman battlefield allowing battlefield archaeology, it should feature in any future discussion of the Roman army on the battlefield, and prominently so.