Although finds of weapons, belts/baldrics and horse gear are generally viewed as characteristic of military sites, the material from the eastern Rhine delta shows that these categories of finds also occur frequently outside army camps and guard posts. The majority of the c. 2,700 objects in the inventory come from rural settlements, ranging from simple hamlets to larger settlements with villa-like structures (fig. 3.1, appendix 1). In addition, a sizeable quantity of material is documented from rivers, the urban centres in Nijmegen and cult places. Only occasionally do objects occur in graves. Thanks to the extensive data set from the research region, we are able to carry out analyses to give us an idea of the use and significance of weaponry and horse gear from different non-military contexts. These analyses are designed to identify patterns in the material that will be elaborated further in subsequent chapters.

This chapter, which looks at the material at different levels of scale, consists of two parts. The first contains chronological and geographical analyses of the material from the region as a whole.1 The chronological analysis presents the material in various bar charts to give an idea of developments in composition. The geographical analysis focuses on the distribution pattern of the different categories of finds in the eastern Rhine delta, paying special attention to the archaeological context in which the material was found. The research region will be placed in a broader, northwest European context in order to establish whether the patterns observed also apply to material from neighbouring areas.

The second part examines the finds at the site level. In recent decades, various excavations of rural and urban settlements, their associated cemeteries, and cult places in the civitas Batavorum have yielded weaponry and horse gear finds. I will firstly examine the excavated sites in terms of the composition and spatial distribution of the finds before giving an overview of the specific archaeological context of the individual objects. Given the lack of context data, I will only examine the composition of some of the larger river assemblages.