1. Plin., HN 16.212; Ulrich 2007, 245 ff. That boxwood was the preferred wood species for combs is most evident from the classical literature in which the term ‘buxus’ was used as a metaphorical synonym for a comb, e.g. Ov., Fast. 6.229; Juv. 14.194; Mart. 14.25. In this article, abbreviated references to classical authors follow those in the Oxford Classical Dictionary, 3rd edition, Oxford 1996.
2. We owe this observation to Dr. Laura Kooistra (BIAX Consult, Zaandam).
3. For Roman period bronze combs: Donaldson 1818, part 2, pl. 78 = Clarke 1847, 323 f); Pugsley 2003, 23, referring to an example from Chesterford; silver combs: Virgili 1989, 74, without quoting examples! Lead: Antiquarium Communale in Rome: inv. no. 16246; Virgili 1990, 99, cat. no. 147.
4. E.g. the wooden comb from the Neolithic lake settlement of Auvernier-‘Brise-Lames’: Schifferdecker 1977, 17, fig. 16. Such combs are to be distinguished from the so-called ‘long handled combs’ which predominantly date from the Iron Age and were used in manufacturing textiles. Cf. Tuohy 1999, spec. 97 ff.
5. For rare examples of one-piece double-sided Roman combs made from bone, see Overbeck 1884, 453, fig. 252i; De Caro 1996, 276, online available at http://marcheo.napolibeniculturali.it/itinerari-tematici/image-gallery/RA134); Petit & Santoro-Bianchi 2007, 67 (all from Pompeii); Simonett 1972, 116 f. (cemetery Locarno-Muralto, loc. Passalli, grave 37); Virgili 1990, 113, cat. no. 221/19 (Costanza, Archaeological Museum, inv. no. 12772). Even more rare are similar types made of horn (Pugsley 2003, 23, citing an example from Fishbourne), amber (Koster 2010, 183 and pl. 12.40), and silver and lead (cf. above note 3). The Pompeian bronze combs are single-sided.
6. Cf. Pugsley 2003, 16: ‘The situation may reflect the scarce degree of attention given in general to these artefacts. Far too many combs are inadequately published, if at all.’
7. Pugsley 2003, 145 ff (appendix 1: C001-C060; C132). In popular publications on Vindolanda, the excavator mentioned ‘over twenty fine box-wood combs’ without describing them individually (Birley 1977, 123). The extraordinary preservation conditions at Vindolanda are further illustrated by the discovery of one of the combs in its original leather case (Birley 2009, colour plate 7).
8. There is some discussion as to whether the boxwood trees in southern France were tall enough, those which presently grow in the garrigues definitely are not. On the methodological problems of establishing the extension of the tree’s natural habitat, Decocq et al. 2004.
9. Of course, raw material was transported (cf. the box logs in the Comacchio ship: Berti 1990, 53, and fig. 2), but this mostly involved short distances from the tree’s natural habitat to the workshops in town. According to the ancient sources, the boxwood at the Altinum workshop (see below) was readily available in the town’s hinterland (Ferrarini 1992, 192, with the observation that the box tree belonged to the ‘elementi consueti del paesaggio cisalpino’).
10. From Pula we know a faber pectinarius (CIL V 98) and from Asti a refector pectinarius (CIL V 7569; Pugsley 2003, 25 f, 177-9). Other pectinarii are known from Reggio nell’Emilia, Brescia and Ateste, but through the contextual association with lanarii and/or carminatores it will have to remain uncertain whether in those cases the term referred to wool combers rather than comb makers. Cf. Frayn 1984, 150, and note 4.
11. In the Diocletian price edict, the maximum price for a female’s (!) comb was set at 14 denarii.
12. Even if reported finds from the eastern provinces are scarce today, with the Egyptian desert for obvious reasons constituting an exception (cf. for instance, National Museum of Ireland, inv. no. 1904:572: one double-sided boxwood comb from Oxyrynchus; and Gazda 1983, fig. 44: two combs from Roman Karanis, online available at http://www.umich.edu/~kelseydb/Exhibits/Karanis83/KaranisExcavation/KaranisExcavation.html), an iconographic by Waelkens (1986) suggests a widespread circulation of such combs in this part of the Roman empire. Moreover, several references in the classical sources point to Paphlagonia as a source of boxwood combs: cf. Ovid., Met. IV.311, Cat. 4.13; Verg., Georg. II.437.
13. The same is true for Roman Egypt: during the 2009 excavation campaign at Karanis in the Fayum carried out by the Universities of Groningen and Los Angeles under the direction of Prof. R. Cappers and Prof. W. Wendrich, four double-sided comb fragments were found, three of which are imported combs made of boxwood (inv. nos. 14718, 15296, 16544), while the fourth (inv. no. 14207) is a locally manufactured product made from olive (Olea europea), a tree common in the area. Pers. comm. C. Vermeeren (BIAX Consult, Zaandam).
14. The find of a coarse animal hair, perhaps of a cow, in a comb from Vindolanda (Birley 1978, 149, Abb. 60; 143 f.) must thus have been the result of secondary use (MacGregor 1985, 205, note 1; Pugsley 2003, 25). For the grooming of horses and other animals with the help of brushes, Pugsley 2003, 24; Fellmann 2009, 80.
15. The point is nicely illustrated by Birley’s short discussion of the ‘over twenty fine box-wood combs’ from the pre-Hadrianic forts at Vindolanda. According to him, these were all ‘of the sort one would associate with fashionable women’s hair.’ Birley 1977, 123; cf. also ibid, 77.
16. Guide 1920, 138 f., no. 400. Modestina is the supposed owner of the comb. The meaning of the abbreviation at the end is unclear. Some have suggested reading the farewell as VALE. Such a reading, however, assumes an engraving error on the comb maker’s part and suggests that the implement was especially commissioned for the funeral, both of which seem unlikely.
17. Combs with lice were found in the army camp of Ribchester/Bremetenacum: Fell 1996; Pugsley 2003, 148 f: C119-C120; the small town of Oberwinterthur: Brem et al. 1999, 132; and the rural settlement of Vlaardingen-Hoogstad: Schelvis 2003. Isolated headlice from Roman contexts are further known from the colonia of York (Kenward 1999, 912) and from Herculaneum (Capasso & Di Tota 1998).
18. One could think of medieval cloister communities (Bernström 1966) or modern school classes.
19. If one was willing to accept that combs were also used for cutting the hair, the problem remains essentially the same.
20. Whether they were bought centrally by the army and handed out to every soldier, perhaps with deduction of costs, or purchased by the soldier himself as part of his private possessions remains to be seen. The variety in comb forms may suggest the latter but, alternatively, the army may have bought combs from different workshops.