gave back the statues of which Caligula had despoiled the cities, restored the Temple of Castor and Pollux to its original form, replaced the name of Pompey on the stage-building of the Stone Theatre, and added his own name "not because he had built the scene, but because he had dedicated it." He was careful to confine his inscription to the only part of the theatre in which he could justly claim an interest ἄλλῳ δὲ οὐδενὶ ἐνεκόλαψεν. The chapter ends in a similar vein with a reference to the modest deportment of the emperor at the dedicatory celebration.

1

A more explicit recognition of the right of a restorer to affix his inscription to a building in conjunction with the titulus of the founder is furnished by Suetonius, certainly no mean authority on antiquarian matters. In speaking of the buildings which Domitian restored after the fire of Titus, the biographer says: Plurima et amplissima opera incendio absumpta restituit in quis et Capitolium quod rursus arserat; sed omnia sub titulo tantum suo ac sine ulla pristini auctoris memoria (Domit. 5.)2 The animus underlying this comment is unmistakable. Domitian is criticised not because he did not renounce entirely the right of inscription. Suetonius singles out for particular notice the fact that Domitian inscribed his own name only. We must, therefore, surmise that if Domitian had prefixed to his own inscriptions those of the original builders, he would not have overstepped the bounds of a proper policy according to contemporary opinion.

1 I follow without hesitation Reiske's reading οὐχ ὅτι κατεσκεύασεν ἀλλ' ὅτι καθιέρωσεν. Boissevain approves this version, although he did not receive it into his text. The vulgate kai kaliéρwσev does violence to all available data. Kataokeváčely is Dio's regular word for a solo reficere. There is no ground for assuming a second conflagration in the few years subsequent to the completion of the building by Caligula (Suet. Cal. 21.). Had Dio supposed that Claudius built the stage anew, he must have accounted for the necessity of the act just as he does in referring to the restoration by Tiberius. Possibly ovx ŐTɩ Kal KαTEσ KEÚAσEV was the original order- to hazard a conjecture of my own. The transposition was brought about by confusion with the common non modo sed etiam combination.

2 For the bearing which this apparently neglected passage has upon the controversy concerning the date of the inscription of Agrippa on the Pantheon, cf. AJA, IX (1905), p. 449.

By means of this survey of the data I have sought to establish the fact that the method which I have classified as "second" was not employed sporadically in inscribing restored buildings, but was in vogue at all times and on all sorts of structures. To be sure, the inscriptions show no token of its use by the emperors on consecrated edifices until a relatively late period. Yet Augustus by his choice of words in the Monumentum Ancyranum showed plainly that he was conscious of the possibility of appending a notice of his repairs. to the inscription of Catulus on the Capitol. I have previously quoted the phrase sine ulla inscriptione nominis, in which the Princeps alluded to his rejection of the right of inscription. If it had been merely a question of keeping the name of the founder instead of substituting his own, some expression like manentibus titulis eorum qui opera fecerant would have been the natural formula to use. Furthermore, the Stone Theatre, by a clever fiction, was dedicated by Pompey as a temple of Venus Victrix, and therefore stood in close relation to the cult and to the natal-day celebration on August 12.1 Aulus Gellius, 10, 1, 7-9, indicates that the dedicatory inscription on the stage building served for the entire structure. Technically, therefore, the inscriptions placed by Claudius on the scene were located on a consecrated building. Lastly, Suetonius, in the passage just cited, recognizes no restrictions in the application of the method.

If an emperor desired to perpetuate the memory of a former builder, he had, it would seem, at his disposal two modes of procedure, each sanctioned by usage. The bearing of this conclusion upon the question at issue is, I hope, patent. The theory that the emperors looked first to the preservation of the exclusive sanctity of the dies natalis fails to explain a really significant fact of policy, the feature of imperial conduct upon which the historians lay particular stress; namely, why the emperors chose oftentimes to refrain from subjoining to the original inscription a second, thereby rejecting an indubitable prerogative. For, in a choice between the first two methods, considerations attendant on the 1 Tertull. de Spect. 10; C.I.L. 12, p. 324.

dies natalis were equally negligible. If the restorer had elected to add an inscription of his own to the founder's titulus, alteration in the Fasti would no more have been an inevitable concomitant than if he had passed his work in silence. In neither case was the original dedication obscured. Whenever the Fasti register a shifting of the dies natalis or the addition of a subsidiary festival in consequence of the restoration of a temple, it is possible in nearly every case to show that the shrine in question passed through a reconstruction a solo. In this event a new dedication was, of course, the rule and the inscription of the founder was never retained.1 The views of scholars differ as to the significance of the festivals referred to the Temple of the Dioscuri2 and the Temple of Minerva on the Aventine. One of the two celebrations in honor of Quirinus undoubtedly commemorates a restoration. The only question is, which is the date of foundation, February 17 or June 29? The Forum Temple of Concord, the Shrine of Janus near the Theatre of Marcellus, the Temple of the Lares on the Sacred Way, also acquired new dates of celebration. In all these cases, irrespective of any perplexities attached to them, we can prove that a rebuilding occurred together with a new dedication and a new inscription. Thus, the temples of Quirinus, of Minerva, and of the Lares are classified in the Monumentum Ancyranum 5 as new buildings. I have shown that Augustus kept the name of the founder on none of these." The temples of the Dioscuri, of Concord, and of Janus were restored and dedicated by Tiberius. Dio Cassius expressly informs us that the edifices first named were dedicated under entirely new

4

1 Except in cases of abnormal behavior such as that of Hadrian; cf. AJA, vol. IX (1905), p. 448.

2 Cf. Aust, op. cit. p. 43; Wissowa, Religion und Kultus, p. 217; Mommsen, C.I.L. I2, p. 308; Richter, Top. p. 86.

8 Cf. Aust, op. cit. p. 42; Wissowa, op. cit. p. 203; Mommsen, op. cit. p. 312; Richter, op. cit. p. 208.

4 Aust, op. cit. p. 41; Wissowa, op. cit. p. 140; Mommsen, op. cit. p. 310; Richter, op. cit. p. 286.

6 Op. cit. p. 431.

7 This building was begun by Augustus and finished by his successor; cf.

Tac. Ann. 2, 49.

inscriptions; cf. Hist. Rom. 55, 27, 4; 56, 25, I. On the Temple of Janus Tiberius probably wrote the name of Augustus in consonance with his usual policy; cf. Dio Cassius, 57, IO, I. At any rate the name of C. Duilius, the founder, was not repeated. I need not linger over certain doubtful cases. cited by Aust, such as the Temple of Flora and the Temple of Consus, where the literature and the Fasti do not meet on common ground. Certainly available data allow us to lay down the law that a change in the dies natalis was accompanied by reconstruction of the building and obliteration of the name of the founder.

It is possible to go a step farther in criticising Marquardt's suggestion. As a matter of fact the status of the dies natalis need not have been conditioned on the preservation of the inscription of the dedicator. There was another means by which the stability of the festival could be secured even if the restorer desired to stand as sole sponsor for the temple. The ceremonies attendant on the new dedication had only to be celebrated on the annual festival. Thus the date of the new foundation would coincide with the old, no matter whose name was to be read on the architrave. This device was actually resorted to.1 Yet the procedure of the emperors is scarcely so rigid as to indicate that they were affected by overweening concern to avoid tampering with the Fasti. If their attitude had been quite uncompromising, the arrangement desired could have been secured always by a postponement of the new dedication or by rushing the work on the building. It was simply a matter of adjustment. But the Fasti themselves, as we have just seen, prove that the imperial restorer frequently did not deem the game worth the candle.

In view of these facts it seems hardly credible that the emperors took serious cognizance of the dies natalis when they chose the method of inscription to pursue on a restored temple. Marquardt's interpretation of the motives of Augustus solely in terms of the dies natalis is inadequate, even if it is not to be dismissed entirely. As an element in forming

1 Wissowa, op. cit., p. 406.

the decision of the restorer its influence may be disregarded in comparison with another factor which Marquardt's explanation ignores. The prime consideration by which the emperors regulated their conduct in inscribing sacred edifices as well as secular, was what may be termed conveniently the commemorative impulse. The name of the original builder was identified in a memorial sense with his structure, which was thereby endowed with a monumental character. A decent homage to the rights of the founder thus appeared to be an act of homage, a pious obligation. The tribunal of public opinion to which even an emperor was amenable, took this point of view for granted in handing down its verdicts. Thus Augustus, Tiberius, and Claudius went down to posterity as restorers who had displayed loyalty to the memory of the founder. Zonaras, II, 17 (Dindorf, vol. III, p. 53), attributes to Vespasian a like policy.1 Hadrian's biographer tells the same story of him. Even if the traditional estimate of his method must be modified, as I have shown elsewhere is probably the case, he, like the others mentioned, wished to show honor to the great men of former times.

We have seen in the case of Domitian that an emperor who transgressed the rules of chivalrous behavior did so in the face of public opinion and courted criticism. A passage in Ammianus Marcellinus, 27, 3, 7, depicts the popular view of an emperor who treated lightly the obligations resting upon a restorer. The historian is speaking of Lampadius, a certain prefect, who made the restoration of a building a pretext for placing his own inscription alone on the architrave and thus masquerading as the founder. The same shortcoming is ascribed to Trajan in the following words: quo vitio laborasse Traianus dicitur princeps, unde eum herbam parietinam iocando cognominarunt.2 The word vitium is considered none too strong to characterize the conduct of 1 Probably to be taken cum grano salis and not to be accepted as universally

true.

2 Space does not permit me to discuss here the correctness of this description of Trajan's policy. There is reason to believe that the criticism is uncalled for. It is merely as an index to the feeling which instigated the criticism that I wish to utilize the passage.