¹ In the preparation of this paper much valuable help has been received from Dr. Paul Jacobsthal, Dr. I. A. Richmond, Mr. R. Rainbird Clarke, and from the Department of British and Medieval Antiquities and Research Laboratory in the British Museum. The Castle Museum, Norwich, has kindly given permission to publish the helmet.

² Analysis of helmet: tin, 5·9; copper, 81·0; lead, 0·5; iron, 0·1; zinc, 12·5.

³ Strong, E., La scultura romana 11, fig. 5.Google Scholar

⁴ Ibid. pl. 6.

⁵ Toynbee, J. M. C., The Hadrianic School pl. 39, no. 2b.Google Scholar

⁶ Ibid. pl. 43, no. 2a.

⁷ Dalton, O. M., Catalogue of Early Christian Antiquities, etc., in the British Museum, pl. 18.Google Scholar

⁸ Matzulewitsch, L., Byzantimsche Antike pl. 19.Google Scholar

⁹ Brailsford, J. W., The Mildenhall Treasure: a Provisional Guide pl. 1.Google Scholar

¹⁰ Drexel, F., ‘Römische Paraderüstung’, Strena Buliciana 55–72Google Scholar.

¹¹ On the upper part of a bronze visor-mask said to have been found on the face of a skeleton in a grave at Nola, now in the British Museum, there are two sea-monsters, each with a three-pronged tail, comparable to those on the Worthing helmet. But since their heads have been destroyed their precise species cannot be determined (British Museum Guide to Greek and Roman Life 82, fig. 77 (on right); Mittheilungen über Heddernheim i, 42, figs. 44, 45). The rings marking the pupils of the eyes on this mask, and on a visor-helmet in Istanbul, catch the light when the mask is worn and give it a remarkably life-like appearance.

¹² Curle, J., A Roman Frontier Post and its People: the Fort at Newstead 164–173. Pp. 179–180Google Scholar contain a list of all the decorated Roman helmets known to the author in 1911, which forms a very valuable preliminary ‘corpus’ of the material.

¹³ British Museum Guide to Romano-British Antiquities 80, pl. 5; Benndorf, O., Antike Gesichtshelme und Sepulcralmasken pl. 4, 5, and 6, nos. 3a and 3b.Google Scholar

¹⁴ e.g. von Lipperheide, F., Antike Helme. 290, 291, 292, 293, 298Google Scholar. These gladiatorial helmets, although they have no face-mask visors, may possibly have exercised some influence upon the decoration of the military helmets.

¹⁵ Benndorf, op. cit. 12, no. 3a, b, c; Lipperheide, op. cit. 342, 520, 521; Jahrb. Kunsthist. Samml. des allerhöchst Kaiserhauses i, 1883, pl. 3, 40–53, fig. 8.

¹⁶ Arch. Anz. 1941, 151–187.

¹⁷ Benndorf, op. cit. pls. 7, 8, 15, no. 2.

¹⁸ e.g. Mittheilungen über Heddernheim i, 36, figs. 36, 37.

¹⁹ Ibid. pl. 4; Germania xiv, pl. 4 (opposite p. 153). Another very similar helmet, the visormask of which has disappeared, was found, according to Lipperheide (op. cit. 315, 316), in the Danube and passed into a private collection in Vienna. It has a mask on its down-turned peak and a crest terminating in the standing figure of a spread eagle. A snake adorns either side of the cap and there are two rows of conventional curls at the base of the crown. Eagles on Roman helmets are presumably emblems of victory. An exquisite griffin's head, cast in solid bronze, from the Vimose bog-find and now in the National Museum, Copenhagen, may have served as the termination of the crest of a decorated Roman helmet (Brøndsted, J., Danmarks Oldtid, iii: Jernaldern 211, fig. 208)Google Scholar.

²⁰ Braat, W. C., ‘Romeinsche Helmen in het Rijksmuseum van Oudheden’, Oudheidkundige Mededeelingen uit het Rijksmuseum van Oudheden te Leiden xx, 1939, 29–46Google Scholar, figs. 23–31.

²¹ Woelcke, K., ‘Der neue römische Paradehelm aus Heddernheim’, Germania xiv, 1930, 149–153Google Scholar, pls. 1–3, figs. 1–4. On the back of this piece are four animal-heads with long ears (mules or donkeys ?), each turned back upon a long tube-like body, recalling the bird-headed snakes of the Worthing helmet.

²² Braat, op. cit. figs 32–4.

²³ J. Déchelette, La Collection Millon: antiquités préhistoriques et gallo-romaines 253–8, pls. 43, 44.

²⁴ British Museum Guide to Romano-British Antiquities 77, 80, fig. 100, pl. 4. The incised decoration includes two snouted snakes. The snakes, which appear not infrequently on Roman decorated helmets of Classes I–IV, had presumably an apotropaic significance.

²⁵ Evelein, M. D., ‘Een Romeinsche Helm uit het Peel’, Oudheidkundige Mededeelingen uit het Rijksmuseum van Oudheden te Leiden v, 1911, 132–151Google Scholar, pls. 61–5. For another fine example of helmets of this class see Alföldi, A., ‘Eine spätrömische Helmform und ihre Schicksale im germanisch-romanischen Mittelalter’, Ada Archaeologica v, 1934, 101, 102, 105, 122Google Scholar, figs. 6a and b, 14, 15a, pls. 3, 4, 5.

²⁶ British Museum Guide to Romano-British Antiquities 77, pl. 4.

²⁷ Espérandieu, E., Bas-reliefs de la Gaule romaine vii, no. 5822Google Scholar.

²⁸ Op. cit. 172 f.

²⁹ ‘Then those of them who are conspicuous for rank or skill in horsemanship ride into the lists armed with helmets made of iron or bronze and covered with gilding to attract the particular attention of the spectators. Unlike the helmets made for battle, these helmets do more than serve as protection for the head and cheeks; they are made to correspond in every way to the faces of the horsemen, with openings at the eyes contrived so as to avoid exposing them, while admitting of a clear view. Attached to the helmets are yellow “manes” [dyed horse-hair streamers ?], designed less for use than for display.’

³⁰ It is evident that Ammianus Marcellinus regarded the wearing of face-mask visor-helmets by Persian horsemen in battle in 363 as odd and unpractical, since in such head-gear one could neither see nor breathe with absolute freedom (25, i, 12: ‘humanorumque vultuum simulacra, ita capitibus diligenter aptata, ut inbratteatis corporibus solidis, ibi tantum incidentia tela possint haerere, qua per cavernas minutas, et orbibus oculorum adfixas, parcius visitur, vel per supremitates narium angusti spiritus emittuntur’). For graffiti showing Parthian cataphractarii and clibanarii, helmeted and probably masked, from Dura-Europos, see The Excavations at Dura-Europos: Fourth Season of Work, pls. 21, no. 3; 22, no. 2. For clibanarii with visor-masks in the service of the Roman army in the fourth century see Julian, Or. in Constanti laudem I, 3 7: τὸ κράνος αὐτῷ προσώπῷ σιδηροῦν ἐπικείμενον.

³¹ A Hellenistic artist's version of a Celtic cavalry visor-helmet can be seen in a relief from the balustrade of the temple of Athene Polias Nikephoros (197–159 B.C.) at Pergamon. It depicts a trophy of captured Gaulish arms, which includes a horse's ‘chamfrein’ with a feathered fan-like erection on the top of the head—obviously a ‘show’ piece—and a helmet with a bearded mask (Altertümer von Pergamon ii, pl. 43). An Iberian version of the visor-helmet, this time surmounted by a pair of horns, appears on a denarius struck by P. Carisius in Spain between 25 and 22 B.C. (BMCCRE i, pl. 5, nos. 4, 5). Other possible representations of cavalry visor-helmets in Roman art are (i) in the trophy of arms to the right of the Victory between the scenes of the first, and those of the second, war on the Marcus Column (Brunn-Bruckmann's plates 64B and 65A), (ii) on the Flavinus stele in Hexham Abbey (27. Bericht der Römisch-Germanischen Kommission 1937, pl. 2, fig. 1), and (iii) on the Genialis stele at Cirencester (Archaeologia lxix, 186, fig. 13). I owe the second and third of these suggestions to Dr. I. A. Richmond.

The stele, found at Brigetio and now in the National Museum, Budapest, of M. Aurelius Avitianus, described as ‘milis [sic] leg(ionis) I ad(iutricis)’, shows the deceased standing with an oval shield and two long spears. Beside him on the ground, on his right, is a helmet. The drawing of this object in Reinach's, S. Répertoire des Reliefs (ii, 121Google Scholar, no. 1) indicates a face-mask, but a photograph (Barkóczi, L., ‘Brigetio’, Dissertationes Pannonicae ii, 22, 1944Google Scholar, pl. 7, no. 3) only shows what might be an outer visor-mask, covering brow, cheeks, and chin, like that of the crested Heddernheim piece, with the space occupied by eyes, nose, and mouth left hollow. No traces of ornament can be seen, and it is possible that the object is just a plain iron helmet with cheek-pieces curving round the chin (cf. Lindenschmit, L., Die Altertümer unserer heidnischen Vorzeit IGoogle Scholar, ix, pl. 5). The object appearing above the left shoulder of Q. Luccius, standardbearer of the Legio XIV Gemina, on his stele at Mainz, seems to be a headdress with an animal's mask and with animal's paws depending from it (ibid. I, iv, pl. 6), not a visor-helmet with a human face.

³² One of the Leiden helmets of Class III(a) is inscribed ‘t(urma) Pomponi Iusti’; and the word ‘Stablesia’ inscribed upon the Leiden potshaped piece suggests the ‘equites Stablesiani’ of the Notitia Dignitatum. Cf. the inscriptions on visor-helmets of Class I: ‘t(urma) Ges’ (Newstead brass helmet), 't(urma) Pipprisci (?), and ‘Vitalis t(urmae) Crispini’ (visor-mask from River Olt, Rumania, in Vienna).

³³ The elaborately decorated visor-less helmets worn by infantrymen, as well as by cavalrymen, on such historical reliefs as the fragments of the great Trajanic frieze, now surviving partly on the Arch of Constantine and partly in the Galleria Borghese, in the Louvre, and in the Altesmuseum, Berlin (M. Pallotino, Il grande fregio di Traiano pl. 1 and figs. 4 and 5; Strong, op. cit. figs. 92 and 94), and on such battle-sarcophagi as the colossal third-century Ludovisi piece in the Museo Nazionale Romano delle Terme (Strong, op. cit. fig. 200) do not prove that helmets of this kind were worn in actual warfare. These scenes, like those on funerary stelai (vide supra note 31), depict ‘ideal’, rather than real, battles, in which parade-dress is being worn b y the Emperor and his troops alike. For the distinction made in Roman times between parade- and battle-armour, between ὅπλα πομπεντήρια and ὅπλα πολεμιστήρια, see Dionysius of Hali carnassus, De admir. vi dicendi in Demosth. 32.

³⁴ The decorated visor-mask helmets found in royal burials just outside the Empire were probably made in Roman workshops and either sent as presents to, or specially ordered by, the princes who owned the tombs: or they may have been bought in markets near the frontier. Possibly they were executed by Roman military artists working abroad in these princes' service.

³⁵ The “Vitalis” on the Olt mask (vide supra note 32) reads like the owner's name.

³⁶ JRS xxvi, 1936, 250 f.Google Scholar

³⁷ The suggestions made in this paper as to the classification, chronology, and purpose of Roman decorated helmets must be regarded as purely provisional, pending the detailed study, based on a ‘corpus’ of all existing pieces, which the present writer is preparing. A number of related problems, which could not be discussed here, also await solution.

³⁸ Preliminary notes have appeared in Eastern Daily Press, 6th September 1947; The Times, 9th September 1947; Illustrated London News, 20th September 1947; Connoisseur, December 1947.

³⁹ It should be emphasized that the irregular growth of this timber detracts from its reliability, but the suggested dating is consistent with that of fragments of Samian ware (Form 37 by Casurius of Lezoux and Form 30 probably by Divixtus of Lezoux, c. 140 A.D.) found in 1948 in dredged material at this site.

⁴⁰ Shown on Ordnance Survey Map of Roman Britain, 1928.

⁴¹ Roman fetters of a similar type from Sicklesmere, Suffolk, are in the Ipswich Museum.