extremely interesting and valuable inscribed slabs from Lanchester here given (see woodcuts on preceding page), on which are recorded the re-erection, during Gordian's reign, of several important buildings of the fortress there. These works were carried out by Marcus Aurelius Quirinus. I will give the exact concluding words of the text :

2

PR COH I L GOR

There cannot be the slightest doubt about these letters, and the stops are all there, perfect and distinct. Doubtless the L GOR, of the first cohort, of which Quirinus was captain, and the LON GOR of our altar, are identical.

Who can be meant by the expression? It does not add to the simplicity of the matter that, in the one case, it seems to be intimated that they were a people furnishing several cohorts, or regiments, to the Roman army, and that, in the other case, they are mentioned without any limitation as to numbers or divisions.

The Suebians, or Suevians, were a people that inhabited ancient Germany. They are described as being the most ancient, great, and warlike people of all that country. Tacitus says they were divided into several tribes, amongst which he enumerates the Semnones, the Longobardi, the Angli, etc.

Can the abbreviations LON GOR on our altar stand for LONGOBARDORVM GORDIANORVM or LONGOBARDI GORDIANI?

There can be little doubt of the signification of the GOR. It means, it would seem certainly, a title derived from the emperor's name, assumed through affection and devotion to him, by the troops in question. The doubt is confined to the signification of the L of the Durham inscriptions, and the LON of ours.

Is it vexillarü

Then what is the nominative to SOLVERVNT? Sueborum Longobardorum Gordianorum, 'The veterans of the Gordian Lombard Suevians,' as we should say, or is it Longobardi Gordiani, the Gordian Lombards.' Does pro salute stand alone without a genitive after it, or do the abbreviations VEX SVEBORVM, tell specifically for whose health and safety the erectors of the altar expressed their gratitude?

2 See Lapid. Sept. Nos. 699 and 700; and C.I.L. vol. vii. Nos. 445 and 446.

TO THE GODDESS 'GARMANGABIS.'

321

For my own part I do not remember to have seen the word vexillatio used in any Roman inscription found in Britain, except with the name of a legion following it. I, therefore, incline to think that VEX must stand for vexillarii, the veterans, those who were serving the last four years of their military life. In this Dr. Hübner agrees with me.

6

A friend has suggested that the final M stands for milites, the soldiers,' and indicates the nominative to the verb solverunt. Others think M is the usual abbreviation for the adverb merito, or the adjective meritis.

There was a numerus Longovicariorum at Longovicus. And it has been thought that our LON here may be an abbreviation for Longovicariorum. But, if so, from the inscriptions at Durham it would seem that there must have been several cohorts of the Longovicarii, which hardly seems feasible.3

The monument is, undoubtedly, of the age of the emperor Gordian. He was slain in the East, by Philip, who succeeded him as emperor, in A.D. 244.

Unless, as there is a Lanchester and a Lancaster, there was in Roman times a Longovicus and a Longovicium, and two cohorts of Longovicarii, one at the eastern fortress and the other at the western.

(B) BY F. J. HAVERFIELD, M.A., F.S.A.

[Read on the 27th day of September, 1893.]

THE Newcastle Society of Antiquaries is much to be congratulated on the discovery of yet another important Roman inscription. This is a fine altar, unearthed last July at Lanchester in the course of some digging connected with the Workhouse water supply, about two hundred yards north of the Roman fort, and close to the line of the Roman road. In size it is above five feet high (including a loose base) and two feet wide; the lettering is well preserved and large, the letters being three inches tall in the first line, two and seven-eighth inches in the second and third lines, and two and three quarter inches in the other lines. Besides the usual knife, dish, etc., on the sides, the altar is ornamented with an unusual profusion of that Roman geometrical ornament which sometimes reminds one of the later Norman work. Accounts of the discovery have appeared in the New

castle Daily Chronicle (on July 24), in the Academy (by myself, August 19), in the Proceedings of this Society (by Mr. Blair, F.S.A., vi. 55-7), and elsewhere. I have myself examined the altar, which is now in the porch of Lanchester parish church; I have to thank Dr. Hooppell and Mr. Blair for photographs and information.

The reading of the stone is, I think, beyond doubt. Expanded and completed it is :-Deae Garmangabi et n(uminibus) [G]o[rdi]ani

D E A E GAR
MAN G A B I

ETNO│| |│
ANI AVG PR [0]

SAL VEX

.

SVE BO

RVM LON GOR.Vo

Aug(usti) nostri), pr[o] sal(ute) vex(illationis) or vexillariorum) Sueborum Lon. Gor(dianorum), votum solverunt merito) or m(ilites).

With respect to the text I may remark that the first line certainly has DEAE, not DEEE; in the third and fourth lines the name Gordiani has been intentionally erased, but o and ANI can still be detected, and in the sixth line there are distinct stops before and after GOR. In the fourth line there is a fracture after PR.

TVM SOLVERVNT M

The interpretation of the stone involves several points of interest. (1) The name of the goddess, Garmangabis, or whatever the nominative may be, seems wholly unknown. The second part of the name can be compared with two German titles, Matronae Gabiae and dea Idban. Gabia,1 the latter mentioned on an altar found near Cologne, in both of which the syllable gab has been conjecturally connected with geben 'to give.' But the resemblance is not very close, and neither Dr. Stokes nor professor Napier can help me any further.

(2) The name erased in the third and fourth lines is that of Gordian III. (A.D. 238-244), a detail which dates the inscription, and is noteworthy for another reason. Emperors' names were not seldom erased on Roman inscriptions, but the erasures were limited to definite emperors, of whom Gordian was not one. Until the Lanchester altar was found, only one instance, I believe, was known in which his name had been deleted.2 We must explain the present erasure as a result

1 Ihm, Bonner Jahrbücher, lxxxiii. 28 and index; Zeitschrift für deutsches Alterthum, xxxv. 317. The etymology assigned to Gabia is, after all, little

more than a guess.

2 The instance is a milestone on the road from Carnuntum to Vienna (C.I.L. III. 4644). Two other instances are sometimes quoted, but both are due to error. One (C. II. 3406) is a slip in indexing; the other (Lapid. Sept. No. 22; C. VII.

TO THE GODDESS 'GARMANGABIS.'

323

of ignorance, such as caused the erasure of the names Pupienus and Balbinus on a Benwell inscription, but it is none the less extraordinary. It is proved however both by the traces of the lettering and by the GoR of line 6, which can only be Gordianorum.

(3) The regiment or detachment which erected the altar is indicated in the words pro salute vexillationis (or vexillariorum) Sueborum Lon. Gordianorum. The terms vexillatio, vexillari, are used very frequently in Roman literature (for instance, in Tacitus) and on inscriptions of the first two or three centuries, to denote troops, usually legionaries, sometimes veterans, occasionally auxiliaries, who had been detached from their proper organizations for some temporary purpose, and were 510) is the Benwell altar mentioned above. On it we have an ala I. Hispanorum Asturum . . Gordianae; Prof. Hübner (who saw the stone) supplies the gap (a definite erasure of some seventeen letters) as Severianae Alexandrianae and says that Gordianae is in erasure. After examining the stone with my friend Mr. A. H. Smith, M.A., F.S.A., I have satisfied myself that Gordianae has never been erased, and the actual erasure is doubtless Pupienae Balbinae, as Mommsen suggested (C. III. Suppl. 6953), though no lettering can be discerned. Even on the Vienna milestone (which I have examined myself) the erasure is very half-hearted.

VOL. XVI

PP

under a separate command and flag (vexillum). But in the second. century another sense appears, which takes us somewhat deeper into the secrets of the Roman military system. That system, as set up by Augustus, consisted of a uniform series of legions, cohorts and alae, without much distinction of race. The auxiliaries bore tribal names, but the recruiting soon ceased to be tribal. In the reigns of Trajan and Hadrian this began to alter, and a fresh set of auxiliaries levied, armed and drilled on a tribal basis, began to arise beside the regular army. These new troops are generally called numeri; they become plentiful during the second and early third centuries, and with them the tribal name has full meaning. Instead of numerus, we get other terms used occasionally, and among them cuneus and vexillatio. Examples of cuneus will be cited lower down; for vexillatio we have such examples as

Ala et vexillatio equitum Illyricorum (Dacia, A.D. 129; C. iii. pp. 876, 1977). Vexillatio militum Maurorum Caesariensium Gordianorum (Lambaesis in Africa, A.D. 255; C. viii. 2716).

Vex. equitum Maurorum in territorio Auziensi praetendentium

=

'camping'

(Auzia in Africa, A.D. 260; C. viii. 9045-7, and Cagnat Armée d'Afrique, pp. 253, 306).

Vex. Brit. (Holland, undated tiles; Brambach C. I. Rh. 4, 128, 139). And so in Britain

Vex. Germa. utriusque), R(aetiae) D(almatiae?); Lapid. Sept. 811, and C. vii. 303.

Vexil. Raetor(um) et Noricor(um); C. vii. 212.

We cannot, indeed, be quite certain that all these represent separate regiments. The national principle represented in the numeri seems to have, to some extent, invaded the regular forces, and we find at Birrens (C. vii. 1068) Raeti militantes in cohorte II Tungrorum, and at Carrawburgh (Eph. iii. 103) Texandri et Sunici vex. cohortis II Nerviorum, very much like the Germani, cives Tuihanti, serving in the cuneus Frisiorum, which erected the great Housesteads altars to Thingsus and the Alaisiagae. But we may be sure that in most of the cases, and probably on the Lanchester altar, separate troops are meant, and we may take vexillatio here to be hastening on from its classical sense to that which it acquired in the army of Diocletian and Constantine, that of a troop of horse in the field army.

(4) Sueborum affords a puzzle. In the first century A.D., as in

3 Arch. Ael. vol. x. pp. 148-172.