Journal of Archaeology in the Low Countries 2-2 (November 2010)Ton Derks; Wouter Vos: Wooden combs from the Roman fort at Vechten: the bodily appearance of soldiers

4 Production and distribution of boxwood combs

The choice of the wood species combined with the length of the combs – the incomplete comb no. 1 has a longest preserved length of 89 mm – provide important clues to the site of production. In the Roman period, boxwood does occur in Northern Europe, but only as shrubs which remain too small to produce billets of the right size (Van Rijn et al. 1998, 6 ff.). For combs with the length of our examples logs with straight grains and no knots are required. These conditions are only met by tall trees in an environment of dense forests in the mountainous areas of the Mediterranean (Mille 2006). Such habitats are to be found in Greece and Asia Minor (especially the Black Sea coast) in the eastern part of the Mediterranean, and in northern Italy, northern Spain and perhaps also the western part of southern France.[8] Moreover, as boxwood is best worked in green condition, transport of raw material over long distances seems unlikely.[9] We thus conclude that all Roman box comb finds from the northwestern provinces were imported as finished products from the Mediterranean (contra Pugsley 2003, 23) and northern Italy seems to be the most likely source of origin.


Fig. 4: Other boxwood products from the same workshop at Altinum, partly (1-3) made on the lathe. 1 Leg (?) of furniture; 2-3 handles of equipment; 4 sole (after Ferrarini 1992 , 197, fig. 9.15 and 198, fig. 10.16, 10.18, 10.26).

Judging from the small number of inscriptions which mention the profession of comb maker (pectinarius), workshops that produced these combs were concentrated in northern Italy.[8] Interestingly, the only workshop that has been identified through finds of waste material of wood chips as well as half-finished combs is situated in exactly the same area, i.e. in the ancient town of Altinum, close to present-day Venice (Ferrarini 1992). The waste material, found in a stratified context dating from the first half of the 1st century AD, contained a total of 59 combs, all but one double-sided, in various stages of production. They allow for a detailed reconstruction of the manufacturing process. Since this unique find assemblage has not yet been discussed in the Anglophone literature, a short description of the five main stages may be useful. First, rectangular tablets were produced, all of about the same length (7-8 cm), width (5-6 cm) and thickness (1 cm) (fig. 3.1) Second, one of the long sides was bevelled, first on the front and subsequently also on the back (figs 3.2 and 3.3). The next step was to do the same for the other long side (fig. 3.4). What followed was the most crucial stage in comb manufacture, the sawing of the teeth. As fine teeth were more likely to break than coarse ones, the former were sawn first (fig. 3.5). Finally, the item was finished by sawing the coarse teeth on the other long side of the comb.

The finds from Altinum are not just important for the reconstruction of the chaîne opératoire of comb manufacture, they are also instructive for the organisation of the workshop. For besides the combs, a whole range of other boxwood finds comprising parts of furniture, sarcophagi, vessels, handles of equipment and even a small sole (fig. 4), were found in the same stratified deposit. If we assume that all waste material originated from one and the same workshop, its owner (and probably most of his fellow craftsmen as well) seems to have specialised in the working of boxwood rather than comb manufacture. The close grain and hardness of the wood species made it exceptionally well suited for working on the lathe. The skills and craftsmanship which were required for this in fact created the professional specialisation. Given the less sophisticated skills needed for comb manufacture, combs may well have constituted just a less valued side product, but one that perhaps still brought a nice income thanks to the high sales figures.[11]


Fig. 5: Boxwood comb from Clermont-Ferrand bearing on both sides two (!) identical production stamps SAIIRISSAT or SAERISSAT (after Mille 2006 , fig. 1).

Even though the workshop and its boxwood merchandise may have served a rather specialised niche of the market, it must have had quite a number of competitors both in the area and also beyond,[12] at least as far as comb making was concerned. This can be deduced from the first published find of a boxwood comb with a production stamp (fig. 5) as well as from the incredible numbers of box combs – as best exemplified by the Vindolanda and Vindonissa cases – that must have circulated in the Early empire. In view of the many close-knit cultural ties that connected northern Italy with the Rhineland, set in motion by military recruitment and veteran settlement, there seems little doubt that the combs from Vechten actually originated from one of these northern Italic workshops.

Finally, it is interesting to note that the imported specimens incidentally served as a model for local imitations in wood species which were readily available in the region (table 1).[13] However, at the same time, the importance of box combs is an indication that the success of this local comb manufacture was limited. The advantages boxwood had both in the manufacture process as well as when using the comb – box combs normally would not split and thus would not catch individual hairs when combing – easily outbalanced the higher costs of transport.


Type of site

Wood species



military settlement urban settlement


unknown number of finds kept by the National Museum of Antiquities, Leiden; Van Rijn 1993, 181 f


military settlement

Buxus sempervirens / boxwood

12 specimens; catalogue, this study

Valkenburg aan den Rijn

military settlement (vicus)

Buxus sempervirens / boxwood

2 specimens; Van Rijn 1993, 179-181, 185, 49-50, fig. 27.49, 27.50


small town

Buxus sempervirens / boxwood

publication in prep.; oral information P. Seinen, December 20, 2009


rural settlement

Buxus sempervirens / boxwood

Van Rijn 2006, 172-173


rural settlement

Cornux mas / kornoelje

Therkorn 2004, 36, 303-4 [table 7a,c], 327


rural settlement

Betula / birchwood

Schelvis 2003

Frisia, exact findspot unknown

rural settlement (terp?)

Unknown, probably boxwood

National Museum of Antiquities Leiden, vdT zn 85 (described as medieval!); unpublished

Table 1. Roman wooden combs from the Netherlands according to finds context.