A Roman sports helmet was found at Ribchester, Lancashire, in the summer of 1796. It formed part of a hoard of some thirty items and was bought by the collector Charles Townley (1737–1805) of Townley Hall, Burnley, Lancashire. The hoard was published by the Society in Vetusta Monumenta.Footnote ¹ Publication took the form of a letter from Charles Townley to John Brand, Secretary of the Society, with illustrations provided by the Society. As a postscript to the account, Charles Townley provided information about a further item in the hoard in a letter dated 17 December 1798.Footnote ² The letter was accompanied by a statement by the Revd T D Whitaker, a well-known antiquary. The text of the two documents is here reproduced:

The letters intimate that the crest had the body of a lion and the face of a woman, namely a sphinx, was made of bronze, curved on the lower side and attached to the helmet by solder. The helmet was discussed by W Thompson Watkins in his Roman Lancashire, where he also described the ‘sphinx’, quoting Whitaker's account and concluding that ‘the suggestion that the sphinx was attached to the top of the helmet is not at all improbable’.Footnote ³ He also suggested that the helmet, ‘from its delicate and fragile nature, could not have been made for active warfare, the question of arises, for what was it used? We are as yet in the dark on the subject, though my own idea is that it was used processionally.’ It is, of course, now fully accepted that the Ribchester helmet was a cavalry sports helmet.Footnote ⁴

The late Ben Edwards, FSA, published a booklet on the Ribchester hoard in 1992. In this, he concluded that it was ‘improbable that the item was a crest’.Footnote ⁵ The discovery of the Crosby Garrett helmet, however, caused him to change his mind, and he communicated this to the present author shortly before his death, leading to this short paper.

The adornment of a helmet with a crest of plumes is well known from Roman imperial-period iconography. The crest was held in a crest-holder that was slid or twisted into a plate on top of the bowl, as is most commonly seen in first- and second-century infantry helmets.Footnote ⁶ Such crests were worn by ordinary soldiers as well as officers. In some cases, a representation of the crest is rendered in metal. In the case of the cavalry helmet from Theilenhofen, this has an eagle at the front, though here the crest is integral to the helmet design.Footnote ⁷ The placing of a separate, possibly detachable, crest on a helmet is unusual. Robinson in his detailed treatment of crests does not offer a single example.

The Crosby Garrett helmet was found by a metal detectorist in 2010.Footnote ⁸ It is of the type known as Robinson Cavalry Sports C, whereas the Ribchester helmet was categorized as Cavalry Sports B.Footnote ⁹ Both series date to the late first century through to the mid-third century. The crest on the Crosby Garrett helmet was in the form of a winged griffin made of copper alloy. It had been attached to the top of the helmet with traces of the lead solder surviving on the peak of the helmet; the solder survived on the concave lower side of the Ribchester sphinx. In 2010 a second example of a griffin was found at Vindolanda in a third-century context in the extra-mural settlement. The base of this is not curved and its function is not clear.Footnote ¹⁰ Jackson has suggested that the object on the Ribchester crest may have been confused with another hybrid creature, a griffin.Footnote ¹¹ The contemporary description of the item as having the body of a lion and the face of a woman suggests that this is unlikely.Footnote ¹²

Watkins made the suggestion that the crest may have been attached to the top of the helmet. This, however, seems unlikely. The helmet is covered in repoussé decoration so that there is no space for the crest to rest, while the uneven surface of the helmet ensured that a permanent fix between the helmet and the crest would be difficult. Moreover, the helmet has ‘split spiked loops at top and base of the head-piece probably for securing a detachable crest box’.Footnote ¹³ The probability must therefore be accepted that the crest did not belong to this helmet.

Jackson drew attention to the varied composition of the hoard, which contained ‘a) some pieces that were apparently complete and intact on burial, b) some pieces that had been de-silvered prior to burial, and c) some pieces that appear to have been buried in a worn, damaged or broken state’.Footnote ¹⁴ Appreciation that the sphinx crest was not attached to the helmet reinforces Jackson's conclusion that the ‘hoard was a miscellaneous assortment of equipment, some functional and some not, which was being stored for future use or for reuse as scrap’.

Jackson reviewed the evidence for the findspot as described by Townley and concluded that ‘the hoard might have been buried immediately outside the fort, in or beneath the rampart, or just inside the fort. In fact, probability suggests the latter.’Footnote ¹⁵ Roman barrack-blocks are normally arranged so that the quarters for the senior officer lay at the rampart end of the building. A location for the Ribchester hoard ‘just inside the fort’ may suggest that it was buried within the officer's quarters (in this case, as this was a cavalry regiment, a decurion, an officer in charge of a troop of thirty-two soldiers). The problems of interpreting Townley's plan, however, as Jackson has noted, renders any such conclusion uncertain, yet there are specific links between the helmet and an officer rather than a soldier.

The helmet was surmounted by a corona muralis. Professor Valerie Maxfield, FSA, advises me that Roman military decorations were only normally awarded to Roman citizens.Footnote ¹⁶ Centurions and decurions in auxiliary units normally rose through the ranks, and therefore were not Roman citizens, though transfers from legions, which were composed of Roman citizens, are known, albeit few are attested.Footnote ¹⁷ On this basis, therefore, the helmet is likely to have been owned by the commander of the unit, or by a former legionary who had been decorated and perhaps as a result promoted to the post of decurion in an auxiliary regiment, or by a former legionary who was decorated while a decurion.Footnote ¹⁸ There is one further possibility, pointed out to me by Professor Maxfield, which is that the Ribchester helmet does not reflect a real award of a corona muralis, but was worn by a soldier playing a role in a military exercise, and in support of this possibility may be cited sports helmets at Eining which bear female hair styles.Footnote ¹⁹ On balance, this seems to be the most likely explanation for the appearance of the corona muralis on the Ribchester helmet.

Detached crests of the type found at Ribchester and Crosby Garrett are extremely rare and accordingly it might be assumed that these two examples would also be worn by officers, but we should be careful about drawing conclusions from such a small number of objects; even ‘normal’ sports helmets are comparatively few in number.

Two cavalry units are attested at Ribchester in the second and third centuries, the Second Cavalry Regiment of Asturians and the Sarmatian Cavalry Regiment. The former had moved from Ribchester by about ad 180 when it is recorded at Chesters, while the latter may have been drawn from the 5,500 Sarmatians sent to Britain in ad 175.Footnote ²⁰ Jackson prefers the Asturian as the unit based at Ribchester when the hoard was buried on the basis of the date of the helmet and the possible link between the name on the helmet, Caravus, and Spain.Footnote ²¹

The nearest fort to Crosby Garrett is Brough-under-Stainmore, 8km (5 miles) to the north east. This fort is famous for the many lead sealings found there, including those recording the Seventh Part Mounted Cohort of Thracians, possibly based at Brough.Footnote ²² Kirkby Thore lies 17.5km (11 miles) to the north west. A cavalry unit was stationed here but its name is unknown.Footnote ²³

In conclusion, the sphinx found in the Ribchester hoard in 1796 may well have been a crest, but it is unlikely to have been attached to the helmet in the same hoard. This reinforces Jackson's suggestion that the hoard was composed of material saved for reuse. There is no need to link either the helmet or the crest to an officer. The location of the hoard is too vague to confirm its burial in a decurion's house; a more likely location for its deposition would be in the intervallum space in the fort where a variety of activities took place, though there can be no certainty for this.