The curious case of the praefectus legionis sextae

Examining a mysteriously vague command listed in the roster for the Roman army in Britain during the late 4th century AD.

Prefect of the sixth, or prefect at Sextae? 

Late Roman commands outlined in the Notitia Dignitatum (a comprehensive roster of Roman military units) are often an unfortunately vague affair, compounded by the lack of any extant original copy of the text. Not only are we relying on medieval or renaissance-era copies, which may contain errors, we are also at the whim of the original scribe (or scribes). The document is an admirable attempt to compile all regiments of the late 4th or early 5th century army in order of their commanding officer, in both halves of a known-world-spanning empire, in a pre-industrial world without access to modern communications. As such, some errors on the part of the initial author are almost certain to have crept in (indeed, this will be addressed below when discussing the Comes Britanniarum). 

For the uninitiated, an entry in the text lists the commander of a field army or region, and the deployments of the commanders beneath them. As an example, there is a roster for the Magister Peditum (the general of foot) and the regiments in his field army, and a roster for the Dux Britanniorum which lists the sub-commanders beneath his station (the Prefect of Dalmatian Cavalry, the Prefect of the Barcarii, and so on).

Terminology in the ancient world is also more fluid than translators would like. As an example, whilst it has been assumed until now that the term cataphract refers to heavily-armoured shock cavalry, and therefore any regiments in the N.D. listed as cataphracti are equipped in such a fashion, the term can also refer to just a cuirass rather than a fully panoply. Accordingly, in this case it may very well simply mean “armoured” – I.E. the regiment is not specifically a fully-armoured unit of lancers, they are instead identified as potentially any mainline cavalry not relegated to scouting or skirmish duty. 

Here we see the crux of the matter with identifying the command Praefectus legionis sextae in the N.D. The command could mean either “the Prefect of the Sixth Legion” or “the Prefect of the legion stationed at Sextae”. First, we must attempt to establish which of these is most likely. We must also note that in the case of the former, the VI Legion were known to have been stationed at York (Eboracum) – if this is the same unit, we could possibly expect them to remain there into the late period.  As a side-note, it bears noting that the original N.D. scribe may have taken names of postings from a textual list, and the reproduced version in the N.D. is not necessarily indicative of fortification and barracks names at the time the document was originally compiled. Taking an example from the roster of the Dux Provinciae Valeriae ripensis, the third entry reads Cuneus equitum Constantianorum Lusiono nunc Intercisa, i.e., "the Constantinian cavalry squadron of Lusonium, now at Intercisa", while the caption for the third fort reads not plain Intercisa but Nuncinercisa i.e. "now at In(t)ercisa", which only makes sense if it was copied, without understanding, from the textual list. This may mean the use of Sextae is a clerical error, as will be addressed below.

Listed below is an excerpt from the Parisian copy of the document (page 153, shown in the image below), labelling the stations under the command of the Dux Britanniorum (limitanei commander in Britain). The stations are listed as follows:

Sextae, Presidium,

Dano, Morbio, Arbeia, Dictim,

Concangios, Lauatres, Verteris, Braboniaco,

Magloue, Magis, Longouicio, Deruentione.

Each of these corresponds to one of the fourteen prefects reporting to the dux that are not stationed on Hadrian’s Wall. If we follow this line of reasoning, the prefect’s command is stationed at Sextae rather than his regiment being the Sixth Legion. The presence of Sextae on the map alongside the other prefect stations lends a lot of strength to this theory, suggesting the unit’s number (if it ever possessed one as many units, even limitanei, do not) was omitted. 

The second option is that the officer’s posting is absent, and the text is referring to the Sixth Legion rather than the legion from Sextae, and the use of Sextae rather than Eboracum is simply a clerical error on behalf of the original author. As stated above, we should expect the VI legion (by this time, fully-titled as Legio VI Victrix pia fidelis Britannica) to be stationed at Eboracum. This however relies on the assumption that Roman high command has not redistributed the regiment around the country or even returned them to Europe sometime during the 3rd or 4th Centuries. Indeed, detachments from all legions are known to have been distributed around provinces such as Britain and Germany, and often seconded to other forces, best evidenced in the case of the Sixth by an inscription on a dedication to Apollo: “…Gordian’s Own Unit of Sarmatian cavalry of Bremetennacum Aelius Antoninus, centurion of the Sixth Legion Victrix, from Melitene, acting-commander and prefect” (RIB 583). Elements of the Sixth were also sent North to aid in construction of Hadrian’s Wall (RIB 1, 1427) and the Antonine Wall, as well as stationed at Corbridge (RIB 1, 1137). In later service, the Sixth were transferred variously to lower Britain (Britannia Inferior) by Septimius Severus, then again to Carpow alongside II Augusta, before returning to Eboracum. Evidence is extant of a detachment from the Sixth at Hadrian’s Wall serving at Piercebridge under a centurion from the II Augusta (AE 1967, 259). 

This may go some way to explaining how the reformed late Roman army originally created the smaller regiments we see listed in the N.D., in that constituent parts of the old VI Legion from Eboracum were split from the unit and distributed around the province to take up new permanent postings. The nomenclature is also unhelpful, as whilst the word legio may initially imply the unit is a legion in the old-style, a later legion is noted to have been a significantly smaller number of total soldiers (larger estimates placing a 4th Century legio at 2,000 men, and more realistic estimates at 1,000 – around five times smaller than the early imperial legions of 5,000 or more). This allows a lot of room for the constituent parts of the regiment to be split off and sent elsewhere, possibly suggesting the legionis sextae is the “original” part of the legion scaled down, although this is entirely supposition. 

It is worth noting that the occupation of the VI Legion at York can be confirmed up to the middle of the 3rd Century, after which their posting is unknown. It is possible the legion ceased to exist after this point (at least in nomenclature), that the legion was redistributed elsewhere either in constituent parts or wholesale, or that our records are simply incomplete as is often the case in late antiquity. Some suggest that this legion were those who proclaimed Constantine the Great emperor in the early 4th Century as the declaration took place in York, however I have failed to find any evidence stating the designation of the unit. Theories to this effect seem to be relying on the location of York, for many years the home of the sixth legion, as the only indicator that these were the men responsible. As Constantine was campaigning in Britain at this time, and indeed was operating after the reforms of Diocletian, it seems entirely plausible that his forces (originally those of his father Constantius) were either composed of legionaries from Gaul or were new-style post-Diocletian regiments and the Sixth were no longer extant. Again, this is entirely supposition.

There is evidence of other legions from Britain “disappearing” over time. Of the two other legions known to have been present in the province during the 3rd century, only one remains extant in records. Elements of Legio II Augusta are found within the N.D., as part of the Comes Britanniarum’s roster (the Secundani iuniores). Of the other British legion, the XX Valeria Victrix, nothing further is depicted. It is possible this was a consequence of Carausius’ rebellion at the end of the 3rd Century and the unit’s nomenclature changed accordingly, although this is only one possibility. It is worth noting as well that also under the Comes Britanniarumroster is a unit listed as the primani iuniores – mysteriously, no “I Legio” is known to have operated in Britain, so the presence of a “first legion” at this date is unclear. It can at least therefore be concluded that the numerical designation of units is a somewhat untenable means to use in identifying a regiment with a service history lasting into the later period. The following possible explanation for these units is taken from Luke Ueda-Sarson’s excellent study of the Notitia Dignitatum, which can be found at the following link (click here): 

“Looking to Egypt, under the Dux Thebaidos, we find listed a Legio prima Valentiniana (56/7.15) and a Legio secunda Valentiniana (56/7.18). These may have been created by Valentinian I (reigned 364-375). They are less likely to have been old units renamed by Valentinian I, as the only usurper during his reign, Procopius, does not appear to have raised any new units, and the units he is recorded by Ammianus as bringing over to his cause are all attested with their names in the Notitia. Creation under the young Valentinian II is a possibility (375-392), but unlikely given he was in the west, and until 383 was very much a minor influence under Gratian (and there are already seemingly too few units named after Gratian), while from 379 the Eastern Augustus was the more vigorous Theodosius. Valentinian II did however suffer from a major usurper, the 5-year rebellion of Magnus Maximus who came from Britain. It is possible that these two units are ex-units of Magnus': as he controlled not only Britain, but at one point Spain, Africa and Gaul, and almost all of Italy as well, they need not be British units. However, the stationing of the ala quarta Britonum (56/7.24) under the Dux Thebaidos suggests that they could be. Speculatively, I would suggest the Secundani iuniores could be part of the second legionary unit that Magnus 'created', and that the first part, the seniores part, is the Legio II Valentiniana of the Dux Thebaidos, renamed because they had received a title such as Legio II Magnecensium under Magnus Maximus which had then to be expunged. The Primani iuniores would by this hypothesis then be the rump of the old Legio VI formerly stationed at York, the seniores part of which would have been Magnus' first legion, and consequently renamed Legio I Valentiniana when Maximus' rebellion against Valentinian II was finally crushed by Theodosius. Magnus' core British troops would under this hypothesis have been sent to far off Egypt to keep them out of trouble, in the same manner that Constantius II had sent Gallic legions of Magnentius and Decentius' off to Syria in the 350s. However, this is all very speculative!” (Useda-Sarson, 2015)


The problem of the Comes Britanniorum

The rather contradictory evidence for the legionis sextae may indicate that the unit was in the process of being transferred whilst the N.D. was being constructed, although this is again entirely speculative. The above list under the Dux Britanniorum (and shown in the attached image) is composed exclusively of limitanei regiments. The confusion surrounding this part of the roster is compounded by an apparent duplication of some units between the Dux Britanniorum and his immediate superior (possibly, as discussed below) the Comes Britanniorum, whose roster we would expect to include comitatenses regiments. The Count’s own roster also presents further questions, as some of his units do not corroborrate with other parts of the document. As an example he is listed as having a unit of equites cataphractarii iuniores, which are apparently drawn from the units under the command of the Magister Equitum. However, no such unit is described in the Magister’s roster. Ueda-Sarson suggests that this unit is in fact the same as listed in the Dux Britanniorum’s list as the command of the Praefectus equitum carafactariorum (stationed at Morbio), and the Magister Equitum and Comes Britanniorum’s lists are not consistent with each other, likely due to an organisational change of some kind during the time the document was penned. 

All this is to say that there is no small amount of evidence for the reshuffling of British legions during the time the N.D. was being compiled, leading to contradictory information and a rather confusing picture of which units are stationed where (including our legionis sextae). A valid hypothesis is that the British units were being deconstructed to reinforce the Gallic field army, supported by the writings of Claudian in AD 402 who states that the Magister Militum Flavius Stilicho withdrew “a legion” (of unknown designation) that was previously guarding Britain against the Picts back to Europe to reinforce his armies fighting the Goths. This may be a specific reference to Stilicho pulling a single regiment back to support his Gallic army, or it may be a more stylistic attempt to suggest multiple units were being stripped piecemeal from the British province. As ever with Latin literature, particularly poetry such as Claudian, the “correct” answer remains elusive and over-speculation of the specifics can be a dangerous practice. Suffice to say, a general transfer of military forces from Britain to Europe was taking place at the time. 

The section of the N.D. concerning the Comes Britanniorum, and by extension the rest of Britain, is clearly somewhat confusing and contradictory. It is of note that, according to the N.D., the province of Britain boasts two Comes postings (the Comes Britanniorum and the Comes litoris Saxonici per Britanniam, or “Count of the Saxon Shore”) – a unique feature amongst the provinces of the western empire, thereby suggesting either that Britain was more militarised than any other part of the empire or, more likely, that there was a reorganisation of army forces taking place at the time the N.D. was penned and this development was not recorded correctly, leading to the confusion. 

The most likely event, in my opinion, is that the Comes Britanniorum was a temporary or mobile post created to gather regiments that the army could spare from Britain before being reassigned to Gaul to reinforce the Gallic army. After this, the post may have been abolished or, at the very least, transferred to Europe on a more permanent basis.



I would posit in conclusion that, perhaps unsurprisingly, we cannot be certain about the nature of the Praefectus legionis sextae’s command. It is sadly not possible to determine if this posting is of the remainder of the VI legion from Eboracum, an entirely new unit posted at Sextae, or indeed some blend of the two. My personal preference would be to suggest the second or third option, as I find it unlikely that the sixth legion survived the reforms of Diocletian and a two-century posting in Britain entirely intact, although this is of course speculation. 

Our search is not helped by the confusing nature of the British units in the N.D. rosters, but it seems a safe suggestion that these problems stem from a general re-shuffle of the army in Britain during the 4th century. This further supports my own conclusion that the sixth were unlikely to survive as we know them and may have been broken up or reassigned to a different province entirely if not simply scattered around Britain. As stated above, we have no concrete evidence of the sixth legion remaining in York after the mid-3rd century, and the unit were subject to a series of redeployments as well as transfer of sub-detachments throughout the 2nd and early 3rd centuries. It is of course entirely possible, as outlined above, that the use of Sextae (the entire crux of this issue) is simply a mistake by the author or a later scribe producing a copy and is indeed a reference to the Sixth Legion rather than a settlement of Sextae, still stationed at Eboracum. I cannot find further references to a settlement of Sextae, either military or civilian, and as such there is a chance this mystery is simply the result of careless Roman or Medieval scribes. I have therefore put this in the conclusion as, whilst a valid hypothesis, it is a rather anti-climactic answer. In a similar vein, and indeed emphasisng the issue with variable terminology made in the introduction, is the suggestion made by Francis Hagan of The Barcarii that Sextae may actually be a name for Roman York, replacing Eboracum either over time or for a short period (or simply on the whim of a particular author). The late Roman historian Ammianus, writing about the Great Conspiracy of AD 367 - 368 refers to Londinium (London) with the phrase “'which later times (is) called Augusta”. It has been suggested that the 'Augusta' name is derived from the garrison of Legio II Augusta now using Londinium as their main administrative base. If this is the case, it is possible Sextae is the name given to Eboracum, now renamed in the mid-4th century after the garrison legion. Both labels then possibly fell away over time, perhaps only lasting for a generation, and the older names of the cities reasserted themselves. Again, this thought is primarily speculation but seems plausible and indeed aligns with several other examples of variable or inconsistent use of terminology in the original Latin causing problems for modern researchers.

As ever, a clear answer eludes us, and we can at best hypothesise on the most likely explanation whilst considering all possible angles. Alongside many elements of late Roman army hierarchy, organisation, and even basic equipment, it comes down to each individual choosing the option they feel is most appropriate. 

Recommended reading: 

The curious case of the praefectus legionis sextae
Ross Cronshaw
By Ross Cronshaw
Documentation, Organisation, and Strategy