In the autumn of 1921 a scheme for the extension of the Exchange at Bordeaux led to the demolition of certain properties in the Rue du Pont-de-la-Mousque. During the operations the workmen encountered the original Roman wall of the town, built about A.D. 300 as a result of the great invasion of A.D. 276. The Chamber of Commerce, on the enlightened initative of their President, M. Étienne Huyard, decided to follow up the discovery by a systematic and careful excavation. The decision was justified by its happy results. Many interesting objects which had been utilized as building material were brought to light. These included a votive altar with an inscription which is of peculiar importance because of the light which it throws on the early relations of Gaul and Britain. A brief account of it may not be unwelcome to readers of the Journal of Roman Studies.

For many years the controversy as to the nature of this mysterious document, though not entirely extinct, was rendered almost dormant by the general acceptance of the famous hypothesis of Savigny and Mommsen. It was, they suggested, the concluding portion of a comprehensive law, entitled the lex Iulia Municipalis, by which Caesar in 45 B.C. regulated municipal institutions, not only in all communities of Roman citizens within the Italian peninsula, but partly at least in Rome herself, as after all merely the foremost municipality of the empire. It is idle to deny that the hypothesis, so stated, rests upon a somewhat slender basis of established fact. Only two general municipal regulations are contained in the Table, one with certainty, the other with great probability, attributable to Caesar. It has to be assumed, therefore, that only an insignificant part of the whole law has been preserved. As for the reason suggested for including provisions about Rome in such a municipal law, Prof. Reid's dictum can hardly be gainsaid that never was so profound an idea presented in a more trivial form.

Ammianus Marcellinus has in this country long suffered undue and unfortunate neglect. No edition of him, so far as I know, has ever been produced in England. A translation executed between thirty and forty years ago by the late Professor C. D. Yonge, for Bohn's Classical Library, is, I think, not now readily accessible. Mr. Fisher, before he was called from the University of Sheffield to a larger and more laborious sphere, had bestowed much study on him; some of its results appeared in an article in the Quarterly Review for July 1918, which was last year republished in his volume of Studies in History and Politics. The Teubner text edited by Gardthausen in 1874–5 has long been out of print.

The subject is a large one, and in the treatment of it, from considerations of space, it is necessary to omit some of the less important aspects. The main purpose of this article is to show reason for thinking that the estimate generally prevalent of Tacitus as a historian is unreasonably high, and that the brilliance of his style dazzles the eyes of his readers and blurs them to imperfect consciousness of his many great defects. There are three great Roman writers, born within a space of twenty years, who strongly resemble each other in their general outlook upon the Roman state, and whose power of emitting epigrams of a biting, penetrating and even savage character is much the same. These are Lucan, Tacitus and Juvenal. All of them deal with history. There is the same ‘saeva indignatio’ gnawing at the hearts of all the trio. They write as though they were living in the worst of all possible worlds. All three look back with longing eyes on the great glories of the old Republic. All have a feud with the gods, who take so little care for mankind as to impose upon them the inexorable dominance of the imperial system. All three again feel that the Romans of their day are degenerate and corrupt. It may be reasonably said that the combination of pessimism with a capacity for savage epigram is not a promising equipment for the historian. Tacitus was saner than Swift, but if Swift had written the history of our great rebellion and of the reign of Charles the Second we should doubtless have found many points of comparison between him and the great Roman. Napoleon showed a sound instinct when he described Tacitus as ‘the detractor of humanity.’

The excavation described below was undertaken by the writer on behalf of the Ministry of Works and the Royal Commission on Historical Monuments, but the present article is his own sole responsibility.

The defences of the Legionary Fortress at York were first scientifically examined by the late S. N. Miller in 1925–8 and the report on the work in 1925–7 appeared in this Journal. As a result of his work we have a satisfactory framework into which to fit subsequent and previous discoveries. The structural sequence arrived at by Miller can be briefly summarized as follows:—

Period 1. A campaigning base deduced by Miller from occupation earth in what he considered to be part of the clay rampart of Period 2 and dated by him to the beginning of the governorship of Petillius Cerialis (A.D. 71–4).

Period 2. The first defences of which Miller recognized structural evidence consisted of a massive rampart of clay, dated by him to the close of the governorship of Cerialis.

Period 3. In the early part of the second century stone gates and towers were built and linked by a stone wall. The south-east gate was built in A.D. 108–9, but some work was still being done in Hadrian's reign. The clay bank of Period 2 remained behind the stone wall.

Period 4. At the end of the second century there was an extensive restoration under Severus. A new stone wall was built, at any rate on the south-east side, immediately behind the wall of Period 2 which was completely removed except for the foundations.

Period 5. At the turn of the third and fourth centuries the south-west and north-west sides and part of the north-east side were reconstructed, with projecting multangular corner and interval towers on the river front (fig. 9).

In 214 b.c., the army of Ti. Sempronius Gracchus defeated Hannibal's Carthaginian forces near the town of Beneventum. Gracchus, proconsul with imperium in Apulia, had led his troops from Luceria in the North-East, while Hanno, Hannibal's lieutenant, arrived with his forces from Bruttium in the South, and a pitched battle was fought by the river Calor. The Romans were victorious. According to Livy, the Carthaginian force of more than 18,000 was routed, less than 2,000 survived, and 38 standards were taken; but the truly striking fact about Gracchus' victory is that his army was largely comprised of slaves. This had been necessary, in contradiction of Roman law and custom, following the tragic and massive casualties suffered in the previous years' battles, most famously at Cannae. Exceptional circumstances called for exceptional measures: pueri donned men's armour; libertini were called to serve; criminals, too; then slaves, who were purchased to fight for the state. The status of such troops posed a significant problem, both legally as well as socially, a problem that was to have a long history.

An interesting bowl in terra sigillata (form 37) with the cursive signature ACAVNISSA was found last year in excavating Birdoswald; and, since the only other known bowls by this potter, were found at Mainz and Cologne, a description of these two bowls may be of some interest, especially from a chronological point of view.

The glaze of the Birdoswald bowl is poor and the figures are in places somewhat blurred. The decoration consists of an alternation of two panels separated by bead-rows ending in rosettes (pl. IV, nos. 1 and 2). In the first panel there is a medallion of ‘cog-wheel’ pattern with inner concentric ring; it encloses a bird (to right) with upstretched wings, similar to Déch. 1010, but slightly smaller. This bird was first used at Lezoux by the Trajanic potters LIBERTVS and BVTRIO and later by the Hadrian-Antonine potters ALBVCIVS, ADVOCISVS, MERCATOR (not the South Gaulish potter of this name), PAVLLVS, PATERNVS, SERVVS and CINNAMVS; in its smaller form it occurs on a bowl by MAPILLVS (37 MAPILL OF at Corbridge).

At rex sollicitus monstris oracula Fauni,

fatidici genitoris, adit lucosque sub alta

consulit Albunea, nemorum quae maxima sacro

fonte sonat saevamque exhalat opaca mephitim.

hinc Italae gentes omnisque Oenotria tellus

in dubiis responsa petunt.

The traditional view from the time of Servius onwards has been that this passage refers to a shrine in the neighbourhood of Tibur, and that the waters described are those of the sulphur lakes of Albula, white and strong-smelling, which lie two miles away in the open country. This view was first challenged by Bonstetten in 1805, and his theory has been further developed by Carcopino. The main objections to the old view are, firstly, that Albula lies about thirty miles north in the Sabine region, well away both from the district in which is set the action of the last six books of the Aeneid, and from the home of Latinus, whether the latter was at Laurentum (the existence of which is now doubted by some scholars) or at Lavinium; secondly, the difference in name, for that the springs near Tibur were called ‘Albula’ is well attested by ancient notices; and thirdly, the difficulty of access. Both Bonstetten and Carcopino are in agreement in locating the oracle at the sulphur springs near Pratica di Mare, the ancient Lavinium : neither, however, seems to have investigated the site as fully as was desirable, and it may be of interest to give the results of a further exploration.

The term Galatia is here used in the widest sense, equivalent to of Γαλατιϰὴ ἐπαρχεία C.I.G. 3991.

It is a remarkable feature that there is hardly any allusion in the inscriptions of Galatia to πραγματευόμενοι ʽΡωμαῖοι or ὁ δῆμος ϰαὶ οί ϰατοιϰοῦντες Ρωμϰῖοι. I do not remember one; but they may occur rarely and have escaped my notice or memory. In the province of Asia these cives qui ibi negotiantur occur very frequently.

The solitary allusion of any approximate character that I know in Galatia is ὁ δῆμος οἵ τε συμπολιτευόμενοι Ρωμαῖοι, which indicates a closer connexion and more equal rights between natives and cives Romani than was the case in Asia. Paul was a Cilician, and the same rule seems to have held there as in Galatia.

It is well known that in the second century B.C. the Roman Senate was frequently called upon to compose disputes between Greek states, but of its procedure in such cases there has hitherto been little to say. It is therefore a matter of some interest that fresh light is thrown on this question by a new inscription from Crete. The dispute which is here described arose between the towns of Hierapytna and Itanus, both of which claimed possession of an inland territory adjoining the sanctuary of Zeus Dictaeus, and of certain adjacent islands. The history of this quarrel was already known to us in part from an older inscription containing an arbitral award by the city of Magnesia-on-Maeander; but from the few stray references which that document made to the Romans it was impossible to reconstruct the account of their intervention. In future I propose to call the old inscription A, the new one B.

Perhaps the most remarkable of the various forms of decorated ‘Samian’ vessels is the jug or vase which is the main subject of this paper. Previously, the precise shape of these vessels was a matter of doubt, the recorded fragments being incomplete, generally lacking neck, handle and base. An attempt was made in 1933 to restore fragments existing in London on the lines of the St. Rémy green-glazed form, Déchelette 62, but this restoration has been found to be only approximately correct for the particular fragments which belong to Hermet's decorated form 15, as shown in his work La Graufesenque (Condatomago).

In connexion with the preceding paper, a summary of the numismatic evidence as to the currency of Britain, in its relation to that of the Continent, till the establishment of the Saxon scheme, based on the silver penny, may be found useful. It is derived almost entirely from published material.

1. Greek and Roman coins of the third to first centuries B.C. are found sporadically in England: there is no sign that they were used for the purpose of currency, or related in any way to the British and Gaulish money.

2. Late Republican and Augustan coins occur fairly often, and their types are borrowed by British mints.