² The principal modern discussion of the Atrium Libertatis is that of Castagnoli, F., ‘Atrium Libertatis’, Rend. Linc. 8, 1 (1946), 276–91Google Scholar. The sources are collected by Lugli, G.Fontes ad Topographiam veteris Urbis Romae pertinentes, VI lib. XVI, 79–85 (Rome 1965)Google Scholar. See also Anderson, J., The Historical Topography of the Imperial Fora (Brussels 1984), 21–6Google Scholar.

³ Bailey, D. R. Shackleton, ed. Cicero's Letters to Atticus, Vol. II (58–54 B.C.), 46–93 (Books III and IV) (Cambridge 1965)Google Scholar.

⁴ Shackleton Bailey ad loc. (n. 3) shows that the last sentence is indeed to be taken with what precedes it; the ‘nam’ that follows is not explanatory, and introduces the further details of the plan for the Saepta.

⁵ See, for example, Gros, P., ‘L'urbanizzazione dopo la guerra sociale’, Storia di Roma (Turin, Einaudi, 1990) II, 831–55 at 852Google Scholar, on the ‘significato… ‘capitolino’ di questo immenso cratere theatrale’; also Frézouls, E., ‘La construction du theatrum lapideum et son contexte politique’, Théatre et spectacles dans l'antiquité (Leiden 1981), 193–214Google Scholar; Gleason, K. L., ‘The Garden portico of Pompey the Great’, Expedition 32.2 (1990), 4–13Google Scholar.

⁶ Wiseman, T. P., ‘Rome and the resplendent Aemilii’, in Tria Lustra: Essays presented to the Editor of Liverpool Classical Monthly (Liverpool Classical Papers 3 [forthcoming 1992]), 181–92Google Scholar.

⁷ The identity of the two basilicas and the meaning of in medio Foro are not so certain, and have recently once again been the subject of much constructive discussion: Steinby, M., ‘Il lato orientale del Foro Romano’, Arctos 21 (1987), 139–84Google Scholar; Wiseman, cit. (n. 6) on ‘in medio foro’.

⁸ To be explicit: it would not be correct to translate the passage as ‘spreading out a Forum’. The context demands an allusion to the Forum Romanum, and the phrase ‘laxaremus et explicaremus’ would be very oddly applied to an isolated and separate construction ex novo.

⁹ Suetonius, DJ 44, 1Google Scholar ‘theatrum summae magnitudinis Tarpeio monti accubans’. A theatre of the size of that of Marcellus here would create a need for modifications to the topography of the approaches from the Comitium, Clivus Capitolinus and Asylum.

¹⁰ On the Forum Iulium see Thomsen, R., ‘Studien über den ursprünglichen Baues Caesarforums’, OpuscArch II (1941), 195–218Google Scholar; Hastrup, T., ‘Forum Iulium as a manifestation of power’ ARID 2 (1962), 45–49Google Scholar; Ulrich, R. B., The Temple of Venus Genetrix in the Forum of Caesar in Rome: the topography, history, architecture and sculptural programs of the monument, Diss. Yale 1984Google Scholar; Amici, C. M., Il Foro di Cesare (Florence 1991)Google Scholar.

¹¹ As this article was about to go to press, I read the stimulating discussion of Ulrich, R. B. ‘Julius Caesar and the creation of the Forum Iulium’, AJA 97 (1993), 49–80CrossRefGoogle Scholar. It was a great pleasure to find him so substantially in agreement with the views expressed here: e.g., 49 ‘the project that would result in the integrated complex of buildings known as the Forum Iulium was not foreseen when land was first bought in 54 BC.’ I have not altered my discussion except to add references to some of the points where his argument most closely touches mine. On the circumstances of 54: Ulrich notes that competition between Caesar and Pompey, famously, only became intense after the death of Iulia, the former's daughter, the latter's wife, and cites this as explanation of this text; but Iulia only died in September 54 (Cicero, , Q. fr. III 1, 17 and 25Google Scholar; Fam. 7, 9, 1). I would therefore add this to my reasons for not seeing the project of early summer 54 as even the germ of a decisive riposte to Pompey's aedificatio.

¹² See especially Coarelli, F., Il Foro Romano, II Periodo repubblicano e Augusteo (Rome 1985)Google Scholar, Ch. I.

¹³ Cicero's was, moreover, a project of which the cautious Atticus had long approved. A man who had advocated the demolition of the Comitium and the construction of a Curia Iulia in the early 50s would not have been likely to live long enough to see his daughter wedded to M. Agrippa!

¹⁴ Coarelli, cit. (n. 12), 236 notes that the tabernae of the Forum Iulium do avoid the site of the Curia Cornelia, not surprisingly, since the latter in its reconstructed form was still there until the last years of Caesar's dictatorship. Ulrich, cit. (n. 11) 56–7, cf. 72 envisages a plan to isolate the Curia Cornelia in a more spacious setting; but I still very much doubt that the demolition of the Elder Cato's Basilica, and the levelling off of all trace of the Comitium, could have seemed possible in July 54, even if the Curia was not threatened, and the isolation of an old building in a new space by the removal of encumbering structures sounds more like nineteenth or twentieth century urban planning than that of antiquity.

¹⁵ The words derive, rather appropriately, from the language of military manoeuvres: cf. Caesar, , BG II 25Google Scholar, 1 ‘manipulos laxare iussit quo facilius gladios uti possent’ (I owe this point to Nicholas Horsfall). Anderson, cit. (n. 2), 41–2, refers the words to function rather than topography, but their vivid practical colour speaks against this.

¹⁶ Appian, BC II 102Google Scholar, cf. Dio 43, 22, for the functional separation between the two fora. Nichols, F. M., The Roman Forum, a Topographical Study (London 1877), 255Google Scholar saw the need to refer the words to the Forum Romanum, in a sensitive account written before discovery of the exact location of the Forum Iulium. Ulrich, cit. (n. 11), 54–6 adds the telling argument that a new Forum, of all things, would have been so excitingly unprecedented a development in 54 that it would be very strange for no hint of its identity to emerge in Cicero's allusion to it.

¹⁷ Suet, . DJ 26, 2Google Scholar.

¹⁸ Appian, BC II 68–9Google Scholar. Since, moreover, the podium of the temple is cut from natural rock, little work at this end of the Forum can have preceded Pharsalus (I owe this observation to Albert Ammerman). For the view that Pharsalus represented a major moment in the formation of the project, Thomsen, cit. (n. 10), also Ulrich, cit. (n. 10), 12–14, and cit. (n. 11) 53, rightly denying that Caesar had the auctoritas for this sort of project before 48.

¹⁹ Dio 50, 2. Coarelli, cit. (n. 12), 134, n. 42 sees that this plan jarred with the intentions of Caesar. For the end of the Curia Cornelia, ibid. 235–6.

²⁰ Castagnoli cit. (n. 2) puzzlingly speaks of the commemoration of a ‘trionfo gallico’ in 54. T. P. Wiseman will argue in CAH IX² that Caesar was expected to make a decisive return to Rome in the winter of 54, and that might have been the moment intended for the public reception of whatever Cicero and Oppius had been up to. Naturally, no-one could expect much to have been completed of a plan such as the Forum Iulium was to be by that date, and the need to have something ready that was worthy of the reditus is another strong argument for seeing some more limited objective in Att. IV 16Google Scholar. For monumentum as a suitable word for a project that does not have a tight architectonic coherence, see Shackleton Bailey (n. 3), ad loc. Gros, cit. (n. 5), 853 comments on the inappropriateness of the moment and thinks of a smaller early plan for the Forum Iulium as the subject of Cicero's remarks.

²¹ The sum now spent was 100 million rather than Cicero's 60; Suetonius, , DJ 26Google Scholar; Pliny, , HN 36, 103Google Scholar.

²² Cicero and Atticus were both in Rome over the winter 55–4 (still in Rome 14th Feb., Q.fr. II 12Google Scholar [11] = Shackleton Bailey 16). It was only at this point that it became certain that Quintus would join Caesar's staff, and Cicero's reliability in carrying out Caesar's schemes could hardly have been assured before. They would no doubt both have known about any plans of this sort. Cicero went to Campania as usual in the early summer (Q fr. II 13Google Scholar [12] = Shackleton Bailey 17, suggesting that he planned throughout to return at the end of May), leaving Rome before Atticus, who went north to Cisalpina on his circuitous journey to Epirus on May 10th. Cicero seems to have left Rome about 1st May, about which time he received a letter from Quintus in Ariminum to which he replied with the letter cited above. Atticus had left later than intended because of illness (Att. IV 14Google Scholar = Shackleton Bailey 88). Cicero wrote during May, saying that he had heard of Atticus' departure, and this seems to have been the last letter before Att. IV 16Google Scholar: no doubt Atticus was essentially incommunicado while on the move. Cicero returned to Rome in the first days of June (A.d. IV non. Iun.). The next day (Q.fr. II 14Google Scholar [13] = 18 Shackleton Bailey) he received a letter from Quintus sent from Cisaipina (‘Blandenone’ in the MS is a puzzle; it is not easy to see what Caesar and Quintus on their way to Britain would have been doing on the Lago di Varese pace Shackleton Bailey ad loc., and it may be easier to go back to one of the emendations that see Laus Pompeia as the place in question). So the plan was formed after Atticus' departure and before Cicero's return, that is, in the last three weeks of May.

²³ Tortorici, E., Argiletum. Commercio, speculazione edilizia e lotta politico, dall'analisi di un quartiere di Roma di età repubblicana (Rome 1991)Google Scholar ingeniously suggests that Cicero is being deliberately bland because the project is indeed so controversial. Like other aspects of Tortorici's case it demands a level of duplicity that is hard to believe. The letter, we note, was not one of Cicero's more confidential ones, since it was dictated to a librarius. For the possibility that a theatre to rival Pompey's was being planned, above n. 5. For the rebuilding of the Curia Cornelia, Cicero, de Fin. 5, 2Google Scholar, cf. Mil. 90, with Asconius ad loc.; Dio 40, 59.

²⁴ Cf. Castagnoli cit. (n. 2), 283 n. 1 ‘su altri lati del Foro Romano non si possono ricercare ampliamenti’.

²⁵ See above, p. 126, and Steinby, cit. (n. 7).

²⁶ We do not know exactly what was under the Temple of Divus Iulius; cf Steinby cit. (n. 7).

²⁷ If Alt. IV 16Google Scholar does refer to this site, it is in connexion with the ‘basilica magnificentissima’ of Paullus, as argued by Wiseman, cit. (n. 6), 181–2.

²⁸ Tortorici, cit. (n. 23), 65 sees no problem with a plan for the Basilica Julia in 54, but that is largely because he believes that the Cicero passage refers to the Forum Iulium.

²⁹ Coarelli, F., Il Foro Romano, I Periodo arcaico² (Rome 1986), 199Google Scholar speaks of a ‘viadotto’ which he thinks is the ‘porticus ab aede Saturni in Capitolium ad Senaculum et super id Curiam’ of the confused passage Livy 41, 27, 7 (174 bc). On this see n. 99 below.

³⁰ Coarelli cit. (n. 12), 239, 241 for the way in which the Curia Cornelia extended into the space of the Comitium.

³¹ Cicero, Sest. 124Google Scholar (56 bc) ‘venit ut scitis a Columna Maenia: tantus est ab omnibus spectaculis usque a Capitolio, tantus ex fori cancellis plalisus excitatus’.

³² Festus 470L ‘ubi nunc est aedes Concordiae inter Capitolium et forum’; 430L ‘ara Saturni in imo Clivo Capitolino’; Servius, Aen. II 116Google Scholar ‘templum Saturni quod est ante Clivum Capitolinum iuxta Concordiae templum’. For the difficulties, Richardson, L., ‘The approach to the Temple of Saturn in Rome’ AJA 84 (1980), 51–62, at 52CrossRefGoogle Scholar. The orientation of the Temple of Saturn before the Augustan rebuilding was 4° more towards the Arx and away from the alignment of the Temple of Castor: Maetzke, G., ‘Struttura stratigrafica dell'area nordoccidentale del Foro Romano’, ArchMed 18 (1991), 43–200 at 58–75Google Scholar. The divisions appear clearly at Pliny, HN 19, 23Google Scholar ‘mox Caesar dictator totum Forum Romanum intexit Viamque Sacram a domo sua et Clivum usque in Capitolium’. See also Solinus I 13, for the other end of the base of the Capitoline, ‘pars infima Capitolini montis habitaculum Carmentae fuit’, whence the shrine and the name of the gate in the Forum Holitorium.

³³ Coarelli cit. (n. 29), 11–26; Carandini, A.,‘Domus e insulae sulla pendice settentrionale del Palatino’, Bull. Comm. 91 (1986), 263–78Google Scholar = Schiavi in Italia. Gli strumenti pensanti dei Romani fra tarda republlica e medio impero (Rome 1988), 359–87Google Scholar.

³⁴ Compare the odd anecdote of Procopius, Anecd. 8, 15–21Google Scholar. For recent archaeological work here, Maetzke, cit. (n. 32), and for a different account of the systematisation of the area, Richardson, cit. (n. 32), 61–2. For the Senaculum and Basilica Opimia in this area, see nn. 99 and 109 below.

³⁵ Cicero, Mil. 26, 64Google Scholar; it was alleged of Milo that ‘arma in villam Ocriculanam devecta Tiberi, domus in Clivo Capitolino scutis referta’. The land may have been among the lots around the Capitoline Hill sold to raise money in 88 bc, Orosius V 18, 27, cf. Wiseman, T. P., ‘Flavians on the Capitol’, AJAH 3 (1978), 163–78 at 170 and n. 26Google Scholar. For a votive deposit found at the corner of the temple of Concord, Maetzke, G., ‘Area nordoccidentale del Foro Romano’ BullComm. 91 (1986), 372–80Google Scholar; Sciorbino, I., Segala, E. ‘Rinvenimento di un deposito votivo presso il Clivo Capitolino’ Alaz X = QuadAEI 19 (Rome 1990), 38–42Google Scholar; Cristofani, M., ed., La grande Roma dei Tarquini (Exhib. Cat., Rome 1990), 63–8Google Scholar. The deposit is of the sixth century BC and does not tell us much about the situation in the late Republic.

³⁶ Cf. n. 11, ad fin.

³⁷ In this context note that Cicero's words are best related to a plan which visually complements Paullus' Basilica. The Forum Iulium, though splendid, was not an embellishment of the Forum Romanum; the logic of the letter requires that the ‘res gloriosissima’ should be just that.

³⁸ For Caesar's Rostra, Coarelli cit. (n. 12), 246–54. I thus fully agree with Ulrich, cit. (n. 11), 55 that Caesar had ‘agreed to fund a more general program of the enlargement of the old Roman Forum’; I am simply offering a different direction for the plan in question.

³⁹ See Shatzman, I., Senatorial Wealth and Roman Politics (Brussels 1975), 23Google Scholar; (A. Gellius 12, 12, 2, 20,000,000; Shatzman 291 14,800,000; Shatzman 308 3,700,000 min.). It is clear on any interpretation that hidden costs are involved in this transaction, since the surface area of even the Forum Iulium is quite insufficient to generate this sort of sum in normal terms. The nucleus of the domus in the Casa dei Cavalieri di Rodi — a mere nucleus of a house of this period — would fit 20 times into the ‘footprint’ of the Forum Iulium, and so the price of even these houses would have to be nearly the 3,500,000 that Cicero's house on the Palatine had cost him each to account for the sum mentioned. We note that the land for the Forum Iulium is alleged to have cost 100,000,000 in the end (above, n. 21).

⁴⁰ For an instance see Att. IV 13Google Scholar = 87 Shackleton Bailey, where Cicero feels bound to attend Milo's wedding in mid-November 55.

⁴¹ Pace Tortorici, cit. (n. 23), who smells here a whiff of‘edilizia abusiva’ and a plan by Cicero and Atticus to sell their own property in the Argiletum to Caesar at an inflated price, the involvement in the project of Caesar's agents and the real needs of political life at the time make this view seem anachronistic.

⁴² Bunsen, C., Le Forum Romanum expliqué selon l'état des fouilles le 21 Avril 1835 (Rome 1835), 286Google Scholar says of this building ‘le seul édifice grand qui nous reste de la republique, le seul édifice d'état de la Rome ancienne’. For the ‘Tabularium’, see Azzurri, G., Sull'antico tabulario Capitolino (Rome 1842)Google Scholar; Delbrueck, R., Hellenistische Bauten in Latium I (Strasburg 1907), 23–46Google Scholar and Plates III-IX. Lugli, G., Roma antica (Rome 1946), 42–6Google Scholar; Blake, M. E., Ancient Roman Construction in Italy (Washington D.C. 1947), 143–4, 331Google Scholar; Mura, A. Sommella ‘Il tabularium: progetto di consolidamento e restauro’ Alaz IV = QuadAEI 5 (Rome 1981), 126–31Google Scholar; Coarelli, F., Roma (Guida Archeologica Laterza, 2nd edn. Rome/Bari 1981), 32–6Google Scholar.

⁴³ For a brief account, Rodocanachi, E., The Roman Capitol (London 1906), 64–70Google Scholar.

⁴⁴ There has been some archaeological investigation in and around the ‘Tabularium’, revealing earlier Republican buildings on the gallery level, descending in terraces behind the later terrace wall towards the Clivus Capitolinus; Colini, A. M. ‘Aedes Veiovis inter Arcem et Capitolium’ BullComm. 70 (1942), 5–56, at 6Google Scholar. Colini attributed the mosaics he found (51) to an earlier public building. On traces of fire, Sommella, A. Mura ‘L'esplorazione archeologica per il restauro del tabularium’, Alaz VI — QuadAEI 8 (Rome 1984), 159–63Google Scholar, linked by the author to the Capitolium fire of 83, in which case Catulus' restoration work to this building would have been considerably prompter than his works on the Area Capitolina.

⁴⁵ The south end of this gallery is blocked by an imperial sewer.

⁴⁶ High elevations were important in much ancient architecture. Suggestive parallels for the appearance of the structure in the Augustan age may be sought in the great towers with which Herod embellished Jerusalem, especially the turris Antonia: Netzer, E., Greater Herodium (Jerusalem 1981), 79–84Google Scholar.

⁴⁷ Bunsen, cit., (n. 42), 9–10, refers to the discovery of fluted travertine columns with Corinthian capitals, alleged to be from the next stage up.

⁴⁸ Nicolò Signorili, Descriptio Urbis Romae (before 1431), edd. Valentini, R. and Zucchetti, G., Codici topografiche della città di Roma IV (1953), 207–8Google Scholar. The salt-warehouse was at the north end of the façade according to Delbrueck, cit. (n. 42), 24. The inscription is CIL I² 737 = VI 1314, cf. 31597 = ILS 35 = ILLRP 367: Q. LUTATIUS Q.F. Q.[N.] CATULUS CO(N)S(UL) |SUBSTRUCTIONEM ET TABULARIUM| DE S(ENATUS) S(ENTENTIA) FACIUNDUM COERAVIT [EI]DEMQUE PRO[BAVIT].

⁴⁹ Poggio Bracciolini, De Varietate Fortunae (1448), in Valentini/Zucchetti, cit. (n. 48), 232. Cf. Azzurri, cit. (n. 42), p. 9. Dr Patterson points out to me that there must be some question as to whether an inscription on tufa, like that which is still extant, could really have been corroded by salt.

⁵⁰ CIL I² 736 = VI 1313, cf. 31597 = ILS 35a = ILLRP 368: [Q, LU]TATIUS Q.F. Q.N. C[ATULUS CO(N)S(UL)|DE S]EN(ATUS) SENT(ENTIA) FACIUNDU[M COERAVIT] |EIDEMQUE [P]ROB[AVIT]; Canina, L., ‘Sulle recenti discoperte fatte nel grande edifizio Capitolino cognito col nome Tabulario’, Ann.Inst. 23 (1851), 268–78Google Scholar. Note ibid. 270, it was found close to where it was placed by Canina and where it remains today, forming the keystone of one of the doorways in the north range of the ancient building and visible from the road outside.

⁵¹ So already Jordan, H. ‘Il tabulario Capitolino’, Ann.Inst. 53 (1881), 60–73, at 67Google Scholar.

⁵² Poggio's ‘extant in Capitolio fornices duplici ordine novis inserti aedificiis… in quibus sculptum est…’ sounds like the Forum façade, but that would be incompatible with the salt. In the absence of any surviving remains that really resemble Poggio's vision, it must be confessed that he does not sound as if he knew what he was talking about.

⁵³ In fact, this has been tacitly recognised, since there has been no tendency to identify or to talk about ‘Catulus’ Substructio', or ‘the famous Substructio on the Capitoline Hill’.

⁵⁴ Misinterpretation of the abbreviation ‘tab. pub.’ on the inscription of AD 46 CIL VI 916Google Scholar = 31201, seen on the Capitol by the Einsiedeln author, as referring to curatores Tabularii Publici rather than tabularum publicarum, compounded the misunderstanding (thus still Lugli, Fontes, cit. [n. 2]. The diplomata CIL XVI 35Google Scholar and AEp 1974. 655Google Scholar, of AD 88 refer to a building in the Area Capitolina as ‘Tabularium Publicum’: this would not have been a sufficient designation if there had been another building with this or a similar title close by. For other tabularia in Latin epigraphy, n. 59.

⁵⁵ On Q. Lutatius Catulus, consul 78 bc, censor 65, Pauly Wissowa Lutatius 8. His special commission from Senate and People, after Sulla's death, for restoring the Capitolium: Cicero, Verr. IV 69Google Scholar; Gellius, NA II 10, 2Google Scholar. It should be noted i) that there is no evidence that the fire of 83 destroyed anything outside the Area Capitolina; ii) that there is no very obvious reason why the reconstruction of the most sacred temple of Rome should have been arbitrarily and gratuitously linked with an inevitably less prestigious project in the environs — at extra expense — whatever we identify with the Palazzo Senatorio structure; and iii) that such a link is actually excluded by the description of Catulus' work in the Signorili inscription as being decreed by the Senate, without reference to the People.

⁵⁶ Mommsen, T., ‘Sui modi usati da' Romani nel conservare e pubblicare le leggi ed i senatusconsulti’, Ann. Inst. 30 (1858), 181–212Google Scholar = Gesammelte Schriften III (Berlin 1907), 290–313Google Scholar.

⁵⁷ At first he referred the tabularium of Catulus' inscription to the tamieion of the aediles on the Area Capitolina (Polybius III 26), which he thought closely linked to the aedes tensarum. He thought at first that Poggio's testimony was wholly dependent on that of Signorili, but was eventually convinced of the document's authenticity by De Rossi (though it is worth noting that he took considerable persuasion) and accepted that a tabularium of some kind was to be sought under the Palazzo Senatorio. For another honourable exception to the prevailing credulity about the ‘Tabularium’, see De Martino, F., ‘Louis de Beaufort: l'incertezza nella storia’, Index 19 (1991), 3–17, at 10–11Google Scholar, questioning also the value of the fifteenth-century record.

⁵⁸ Jordan cit. (n. 51) was the main counterblast. For an example of implausibility in the argument, one may cite the view that archives were kept in the favissae of the Area Capitolina!

⁵⁹ Balty, J. C., Curia Ordinis. Recherches d'architecture et d'urbanisme antiques sur les curies provinciales du monde romain (Brussels 1991), 151–61Google Scholar, collects the archaeological evidence for city tabularia, many of which can hardly be regarded as certainly identified. Among epigraphic attestations, see for example AEp 1966.67 (Formiae) ‘L. Paccius C.f. aed. |cuuriam tabul. armamentar. portic. muincip. [sic] d.’: cf. Colombini, A., ‘Un edile di Formia costruttore di alcuni edifici cittadini’, Ath. 44 (1966), 137—41Google Scholar. The relative modesty of these structures is nicely illustrated by AEp 1959.310 = AEp 1944.100 (Drobeta), recording the building of a tabularium by two slave vilici. Also ILS 140 (Pisa); ILS 5515 (Castrum Novum); AEp 1952.73 (Vienna); AEp 1972.268–9 (Munigua); AEp 1988.558 (Lucus Feroniae).

⁶⁰ Brief, general account in Posner, E., Archives in the Ancient World (Cambridge, Ma. 1972), 160–85CrossRefGoogle Scholar; some remarks on the nature of administrative record-keeping in Purcell, N., ‘The Arts of Government’, in Boardman, J., Griffin, J., Murray, O., edd. The Oxford History of the Classical World (Oxford 1986), 560–91Google Scholar. Culham, P., ‘Archives and alternatives in Republican Rome’ ClPhil 84 (1989), 100–15Google Scholar stresses the limited efficiency of Roman archival practice. Wenger, L., Die Quellen des römischen Rechts (Wien 1953), 754Google Scholar, speaks of ‘der Doppelfunktion der Tabularia als Aufbewahrungsstellen für öffentliche Urkunden and als fiskaliche Verwaltungsstellen’.

⁶¹ Thus the tabularium now attested in the Basilica Iulia by the Lex de portoriis provinciae Asiae: Engelmann, H. and Knibbe, D., Das Zollgesetz der Proving Asia. Eine neue Inschrift aus Ephesos ( = Epigraphica Anatolica 14 [1989], 2Google Scholar (grammatophulakion). Places like this are the routine setting whose debilitating effect on oratory is deplored at Tacitus, , Dial. 29, 1Google Scholar ‘quantum virium detraxisse orationi auditoria et tabularia credimus in quibus iam fere plurimae causae explicantur?’.

⁶² Georg. II 501–2Google Scholar.

⁶³ For strong words on the subject of such misunderstandings of ancient mentalités, attributed to the petit bourgeois world-view of academics, see Coarelli, cit. (n. 12), 126–7!

⁶⁴ Cf. the view of Bunsen quoted above (n. 42). The view of Jordan cit. (n. 51), that ‘The Tabularium’ was in some senses the Aerarium, or an extension of it, was an attempt to meet this problem, and was adopted by Sachers, ‘Tabularium’, Pauly Wissowa (1932), cols. 1962–9. Culham, cit. (n. 60) also favours this solution and sees in the design a mainly scenographic and architectonic purpose.

⁶⁵ Ovid, , Fasti IV 624Google Scholar. Since the publication of the Fasti Antiates (NSc 1921, 92Google Scholar) which give 13th April as the date also of the dedication of the Temple of Jupiter Libertas, some (e.g. Platner-Ashby ad loc.) have thought that Ovid confused the Atrium and the Temple. In that case the prominence of the Atrium in the Augustan period is interestingly confirmed. But we know of another inaugurated Atrium (the Atrium Vestae, since the Senate could meet there, Servius, Aen. VII 153)Google Scholar, and it is better to take this as an informative parallel rather than Ovidian ineptitude or a coincidence.

⁶⁶ The parallel is naturally the cult or cults on the Aventine (with its ancient plebeian ties) of Libertas and/or Jupiter Liber or Jupiter Libertas: one of these was built by Ti. Sempronius Gracchus, cos. 238, from fines (Livy 24, 16, 19).

⁶⁷ Livy 34, 44.

⁶⁸ Livy 45, 15, 1–5. For the censors and the Atrium in general, see Nicolet, C., The World of the Citizen in Republican Rome (Eng. tr. 1980), 63–4Google Scholar.

⁶⁹ Livy 43, 16, 12–13. Cicero, Arch. 8Google Scholar ‘hic tu tabulas desideras Heracliensium publicas, quas Italico bello incenso tabulario interisse scimus omnes?’ The alternative is to take this as the record-office of Heraclea, but it seems unlikely that Cicero would have expected everyone to know about that. See also Tabula Heracleensis 144–5, ‘eos libros census …accipito…exque ieis libreis quae ibei scripta erunt in tabulas publicas referunda curato, easque tabulas eodem loco ubei ceterae tabulae publicae erunt, in quibus census populi perscriptus erit, condendas curato’ (Bruns, C. G., Fontes Iuris Romani⁷ [Tübingen 1907], 110Google Scholar). It follows, however, from the survival of the bronze cadaster of the Ager Campanus long enough to be tampered with by Sulla (see n. 76 below) that those records at least were not kept in this particular depository).

⁷⁰ Milo's slaves: Cicero, Mil. 59Google Scholar. Villa Publica: Livy 30, 21, 12 (202); 30, 24, 5 (197); Josephus, BJ 7, 5, 4Google Scholar.

⁷¹ Livy 25, 7, 12. In 281 BC the Aerarium is attested as the residence of hostages, Zonaras VIII 3.

⁷² Livy, cit., (n. 69); cf. 29, 37, 12 ‘escenderunt in aerarium’. In AD 69 soldiers were billeted there (Tacitus, , Hist. I 31Google Scholar).

⁷³ Cicero's use of it as a reference point in the letter with which we began clearly reflects its size and importance: cf. Anderson, cit. (n. 2), 23. See also Ovid, Fasti IV 621–24Google Scholar ‘populo dignissima nostro atria’: Servius, Aen. I 726Google Scholar: ‘alii atria magnas aedes et capacissimas dictas tradunt, unde Atria Licinia et Atrium Libertatis’. For atria as public buildings, see now Coarelli, cit. (n. 12), 147–9; Gaggiotti, M., ‘Atrium regium — Basilica (Aemilia): una insospettata continuità storica e una chiave ideologica per la soluzione del problema dell'origine della basilica’, ARID 14 (1985), 53–80Google Scholar; Zevi, F., ‘L'Atrium Regium’ ArchCl 43 (1991), 475–87Google Scholar.

⁷⁴ Bonnefond, M., ‘Le Sénat républicain dans l'Atrium Libertatis?’ MEFRA 91 (1979), 600–622CrossRefGoogle Scholar, based on Sherk, R., Roman Documents from the Greek East (Baltimore 1969), no. 20.Google Scholar (Thasos): 606–7 for the argument about the pomerium.

⁷⁵ Cassiodorus, , Var. III, 6, 1Google Scholar (AD 509–11) to IX, 25, 3 (AD 533); also III 11, 3; 33, 3; V 21, 3; VI 4 3; 15, 2–3; 16, 3; VIII 10, 11. For discussion, see nn. 100–102 below.

⁷⁶ Livy, cit. (n. 69); Granius Licinianus 28, 15 Flemisch, the forma agrorum of the centuriation of the Ager Campanus, 165 BC.

⁷⁷ Festus 277L: Malcovati, H., ed., Oratorum Romanorum Fragmenta (Turin 1953), 89Google Scholar. The speech De auguribus is not datable.

⁷⁸ Ovid, , Tristia III, 1, 70–2Google Scholar ’nec me quae doctis patuerunt prima libellis/atria Libertas tangere passa sua est’. The primacy claimed here is altogether different from the primacy asserted for the work of Pollio: n. 82 below.

⁷⁹ If the state had, for instance, owned copies of the translation prepared at its expense of the Carthaginian agriculturalist Mago, or of the De agricultura of Cato, this would have been a natural place to keep them.

⁸⁰ See, e.g., Wenger, cit. (n. 60), 424, for the usage of the papyri.

⁸¹ Cf. n. 61 above. Senate-meetings: Talbert, R. J. A., The Senate of Imperial Rome (Princeton 1984), 117–8Google Scholar; Castagnoli, F., ‘Sulla biblioteca del tempio di Apollo Palatino’, RendLinc 8, 4 (1949), 380–2Google Scholar, the porticus and libraries of the Augustan Palatine (called Atrium Palatii, Servius, , Aen. 11, 235Google Scholar). We note a meeting of the town-council of Suessa Aurunca in 193 in a library provided by Matidia, , CIL X 4760Google Scholar.

⁸² The earliest source (Ovid, cit. [n. 78] is making a different point, about the collection of learned books, not about public access) is Pliny, , HN 7, 115Google Scholar, ‘quae prima in orbe ab Asinio Pollione ex manubiis publicata Romae est’, explicitly allotting Pollio's ‘first’ in library-history to the process of publicare ex manubiis; at HN 35, 10Google Scholar ‘qui primus bibliothecam dicando ingenia hominum rem publicam fecit’ gives him another primacy, for the idea of setting up portraits of the authors. Suetonius, DJ 44 ‘bibliothecas Graecas Latinasque quas maximas posset publicare data Marco Varroni cura comparandarum ac dirigendarum’ confirms that there was no monumental, decorated public library at Rome before Caesar. Isidore, Etym. VI, 5, 1–2Google Scholar derives from Suetonius: ‘Caesar dedit M. Varroni negotium quam maximae bibliothecae construendae. Primum autem Romae bibliothecas publicavit Pollio, Graecas simul atque Latinas, additis auctorum imaginibus, in atrio quod de manubiis magnificentissimum instruxerat’, and Suetonius, Aug. 29Google Scholar confirms that the ‘atrium’ which commemorated Pollio's triumph was the Atrium Libertatis. Strocka, V., ‘Römische Bibliotheken’, Gymnasium 88 (1981), 298–329Google Scholar has recent bibliography: 307 on the library of Pollio.

⁸³ Pliny, , HN 36, 23–5Google Scholar; 33: Becatti, G., ‘Letture pliniane: le opere d'arte nei monumenta Asinii Pollionis’, in Studi in onore di A. Calderini & R. Paribeni (Milano 1956), III, 199–210Google Scholar. For the decoration, n. 82 above.

⁸⁴ That Varro was still alive dates the completion of the project to before 28 bc. Nicholas Horsfall points out to me how significant the choice of the place where the cadasters of public land were kept might be in the context of the politics of land allotments at the time. It has been suggested that Pollio was in some sense carrying out the intentions of Caesar, and some of the points made in the sources (n. 82) might support this; but Pollio had his own interests (Pliny, HN 36, 33Google Scholar ‘Pollio Asinius ut fuit acris vehementiae, sic quoque spectari monumenta sua voluit’) and despite the possible connexion of Caesar and the Atrium Libertatis implied in the letter of Cicero with which we started, it is probably better to take this as a new initiative.

⁸⁵ Dio 55, 8, 1; HN 36, 28Google Scholar meetings in the Porticus Octaviae; cf. n. 81.

⁸⁶ See n. 65 for the link between Jupiter Victor and the Atrium Libertatis. For the location of the former, below, n. 98. Dr Lintott points out to me that the years around 238 were the years of success of C. Lutatius Catulus cos. 242 and Q. Lutatius Cerco cos. 241, cens. 236. If they had invested in the ideology of Libertas through founding or restoring the Atrium, that would fit well with the involvement with it later of a Q. Lutatius Catulus. A dedication to Libertas recorded on the Capitol, n. 120 below.

⁸⁷ Livy I 8, 5 makes it clear that the locus saeptus which formed the precise evocation of the Asylum in Augustan times was beside one of the ways up to the plateau, probably in the vicinity of the Temple of Veiovis (cf. Ovid, , Fasti 3, 340Google Scholar).

⁸⁸ Cf. n. 72 above, and note also the word ‘extemplo’. The element of climbing involved in the journey to an Atrium Libertatis near the Porta Fontinalis is not such as to be worth pointing out incidentally in a passage of this kind: see n. 117 below.

⁸⁹ Livy 24, 10. For Cato's speech, above, n. 77. Castagnoli cit. (n. 2), 288, n. 4 already saw the desirability of the link between the passages (since otherwise Cato's fire was either in the rather distant past or strangely omitted by Livy), and was forced to suggest that the site he preferred for the Atrium on the Capitol/Quirinal saddle could be called ‘in Capitolio’ in order to make it work, a solution thought too difficult by Bonnefond (cit. n. 74), 614, who remained struck by the coincidence. For the alternative nomenclature, we may note the pairing of Atrium Libertatis and Villa Publica in the repairs of 194 (n. 67).

⁹⁰ For the Aerarium Militare, AEp 1978.658. For the possibility that the Capitol was the base in late antiquity for the comes sacrarum largitionum, below, n. 95.

⁹¹ For the fire, n. 69; for its possible traces, n. 44.

⁹² Livy 45, 15, 1–5 is especially important for this role. The God of the Asylum was not named, but a link with Jupiter seems obvious, and the ties between Libertas of the Atrium and Jupiter are explicit: Dion. Hal. II 15, cf. Plutarch, Romulus 9Google Scholar on the theos asylaios (cf Piso fr. 4P quoted by Servius, , Am. 2, 761Google Scholar ‘Lucoris’). For other associations of the Capitoline Hill with the world of the slave, see Coarelli, F. ‘Iside Capitolina, Clodio e i mercanti di schiavi’, Alessandria e il mondo ellenisticoromano. Studi in onore di A. Adriani (Rome 1984), 461–75Google Scholar.

⁹³ Anderson, cit. (n. 2), 95 sees this as a function of the Atrium despite siting it elsewhere.

⁹⁴ For the location of the rupes Tarpeia, Wiseman, ‘Flavians on the Capitol’, cit. (n. 35), 169; ‘Topography and rhetoric: the trial of Manlius’, Hist. 28 (1979), 32–50Google Scholar. For ‘punishment-topography’, Coarelli cit. (n. 12), 80–7, linking the Atrium Libertatis with the Lautumiae; 84–7 on incarceration in arcae as characteristic of the prison-zone; also David, J.-M., ‘Du Comitium à la roche tarpeienne…sur certains rituels d'exécution capitale sous la république, les règnes d'Auguste et Tibère’, in Du chatiment dans la cité. Supplices corporels et peine de mort dans le monde antique (Rome 1984), 131–6Google Scholar. The use of the Atrium for the torture of Milo's slaves (n. 70) has a certain irony if Milo's house was so close at hand. Is the surviving gallery part of the ‘road leading to the prison’ that Sejanus' lictors took when they were separated from him by the crowds in the Clivus Capitolinus at Dio 58, 5? It is difficult to see how else they could have got from the approach to the Area Capitolina to the Scalae Gemoniae.

⁹⁵ Sidonius Apollinaris Ep. I 7, 4, cf. the scene of Arvandus strolling nonchalantly on the Area Capitolina, which seems by this time to have become a kind of luxury bazaar, ibid. 8.

⁹⁶ For the security angle and the case of Herod, see above, nn. 71 and 46.

⁹⁷ AEp 1974.655, very similar to CIL XVI 35Google Scholar.

⁹⁸ Torelli, M., Typology and Structure of Roman Historical Reliefs (Ann Arbor 1982), 12Google Scholar sees that the demands of censorial practice should require the Republican Atrium to be ‘near the curia, the tabularium, and the aerarium’, which might be a description of this site. In succession crises, the senate met on the Capitoline (see Talbert, cit. (n. 81), 120). In AD 41, Suetonius, Gaius 60Google Scholar, Dio 60, 1, 1 and Josephus, BJ 2, 205Google Scholar refer to the Capitol, and Josephus, AJ 19 248Google Scholar specifies the location as the Temple of Jupiter Victor (which shared a dedication date with the Atrium Libertatis, above n. 65). Wiseman, T. P., Death of an Emperor (Exeter 1991), 96Google Scholar takes the meeting of 41 to be in the Capitoline Temple itself, like that of AD 238 (Herodian 7, 10, 2–3). But why should Josephus use an allusive epithet if that is what he intended? For the Athenaeum in which the Senate met in 193 (Dio 63, 27, 2b), see however Callmer, C.Athenaeum’, OpRom 7 (1969), 277–84Google Scholar. For the connexion between the Curia and the Atrium Minervae, Balty, cit. (n. 59), 149–60; Atrium Vestae, n. 65; Atrium Palatii, n. 81. For the Atrium Libertatis and the Senate in the fifth century, n. 75 above, 100–2 below.

⁹⁹ Now that the zone between the Asylum and the Forum loses the burden of accommodating a huge Tabularium, it becomes available for the location of public buildings associated with the Atrium Libertatis. This might help interpretation of the much disputed passage Livy 41, 27, 7 ‘clivum Capitolinum silice sternendum curaverunt et porticum ab aede Saturni in Capitolium ad senaculum et super id curiam’. Compare the connexion between Basilica Opimia and Senaculum at Varro LL v 156, and the proposal for the identification of the former in n. 109 below.

¹⁰⁰ For the Atrium in Cassiodorus, above, n. 75 with Bonnefond, cit. (n. 74), 610; Tortorici, cit. (n. 23), 75 and 97. A different view holds that the late sources do refer to a building separate from the Curia Iulia, though not far removed from it, and are reduced to postulating a new site entirely because the Republican building was thought to have been destroyed and its site occupied: Welin, E., Studien zur Topographie des Forum Romanum (= Acta Inst. Reg. Sueciae 6, Lund 1953)Google Scholar. The name occurs in inscriptions as well as in literary texts: one, Bartoli, A., ‘Il Senato Romano in onore di Ezio’, Rend. Pont. XXII (1946–1947), 267–Google Scholar72 = AEp 1950.30 (AD 437), is in flowery language not unlike that of Cassiodorus, and likewise topographically unlocatable. The other is much more concrete: Bartoli, A., ‘Lavori nella sede del Senato Romano al tempo di Teodorico’, BullComm 73 (1949–1950), 77–88Google Scholar = AEp 1953.68, a major addition to CIL VI 1794Google Scholar = 31933 = ILS 825 (in Lugli, Fontes cit. [n. 2] only in its unsupplemented state):

(after the imperial titulature [of AD 493–518] and the name of the benefactor)

The new fragment was found under S. Adriano, and was therefore attributed by Bartoli to the Curia Iulia: but note Colini, cit. (n. 44), 46, brickstamps of Theodoric found in the ‘Tabularium’, and the clear if tantalising reference to the Capitolium in the last line but two.

¹⁰¹ Sidonius Apollinaris, Ep. I 7, 9Google Scholar.

¹⁰² De Francisci, P., ‘Per la storia del senato e della Curia nei secoli V e VI’, Rend.Pont. XXII (1946–1947), 275–315 at 314Google Scholar, and Degrassi, A., ‘L'iscrizione in onore di Aezio e l'Atrium Libertatis’, BullComm 72 (1946–1948), 33–44 at 39Google Scholar argue that the Curia proper was unusable for much of the fifth century, and that the Atrium must therefore have been a real place. Against these, Lugli, G., ‘“Atrium Libertatis” e “Libertas” nella Roma del tardo impero’, Synteleia. Sts. V. Arangio-Ruiz II (Naples 1964), 807–15Google Scholar.

¹⁰³ For the inimicitia, Pliny, HN pr. 31Google Scholar; cf. Richardson, cit. (n. 32), 56. Coarelli, F., at Epigrafia e Ordine Senatorio I (Rome 1982), 237Google Scholar proposes to recognise architectural terracottas similar to those of the Temple of Apollo Palatinus found on the slopes of the Arx near the Museo del Risorgimento (Di Mino, M. R., ALaz IV (Rome 1981), 119–25Google Scholar) as the product of Tusculan clayworks of Asinius Pollio, and attributes them to the Atrium Libertatis. I owe this reference to Dr Patterson.

¹⁰⁴ Orosius VII 16 ‘fulmine Capitolium ictum, ex quo facta inflammatio, bibliothecam illam maiorum cura studioque compositam, aedesque alias iuxta sitas rapaci turbine concremavit’, which suggests that the library was both ancient and famous; Jerome s.a. (Eusebius, ed. R. Helm [Berlin 1956], 209). Lydus, , Mag. III 29Google Scholar (ed. A. C. Bandy [Philadelphia 1983], 176) shows that an aule Kapetolidos ( = ‘Atrium Capitolii’) was the seat of learning of Constantinople in his day. Also Cod. Theod. XIV 9, 3Google Scholar (AD 425). CIL VI 1708Google Scholar, a strange text probably to be dated to AD 337, is an honorary dedication to a senator of high learning (philosophus) dated by the 381 years ‘post Caesariana tempora’: it was seen ‘in Capitolio’ by the Einsiedeln author, but the tempting connexions with various parts of the case proposed here are too tentative to be worth setting out.

¹⁰⁵ The Temple of Veiovis is probably that vowed by L. Furius Purpureo in 192 and dedicated in 192 (Livy 30, 41, 8); but the cult-place may be older, and the archaeological remains seem to support that.

¹⁰⁶ Thus Zanker, P., The Power of Images in the Age of Augustus (Ann Arbor 1988), 21Google Scholar. We note the possibility that the senate met in the Atrium Libertatis in 80 (above n. 74).

¹⁰⁷ See n. 74 above.

¹⁰⁸ Coarelli, F., ‘Public building at Rome between the Second Punic War and Sulla’, PBSR 45 (1977), 1–19Google Scholar. See also Gros, P., Architettura e società nell' Italia romana (Rome 1987), 114–18Google Scholar.

¹⁰⁹ For the elusive Basilica Opimia: Purcell, N., ‘Rediscovering the Roman Forum’, JRA 2 (1989), 161–2Google Scholar. We know only that it was in close contact with L. Opimius' Temple of Concordia (Varro, , De LL V 156Google Scholar), and that public slaves were based there (CIL VI 2338–9Google Scholar: compare the servi publici of Livy 43, 16, 12–3), which fits a frontispiece to the Atrium Libertatis, at the heart of the topography of public business, very well. The stoa-like plan, complete with taberna-like embrasures opening off the back, is suggestive, and there is little enough room for the Basilica anywhere else. For Basilicae as façades for Atria: Coarelli, cit. (n. 12), 59–63 on the Basilica Porcia and the Atrium Maenium and Atrium Titium; for the original basilica as a version of the Atrium Regium, Gaggiotti, cit. (n. 73), and Zevi, F. ‘L'Atrium Regium’, ArchCl 43 (1991), 474–87Google Scholar; for the area of the Atrium Vestae, Steinby, cit. (n. 7).

¹¹⁰ Castagnoli cit. (n. 2) showed decisively that the idea that it was behind and adjacent to the later Curia Julia fitted neither the literary, nor the archaeological, evidence, and this is confirmed by Morselli, C. and Tortorici, E., Curia, Forum Iulium, Forum Transitorium (Rome 1989 [1990]), 218–37Google Scholar.

¹¹¹ Castagnoli, cit. (n. 2), 282.

¹¹² Route to Campus Martius: Castagnoli, cit. (n. 2), 281, Bonnefond, cit. (n. 74), 612, citing Livy 35, 10, 12. Tortorici, cit. (n. 23), 76 ‘in alto sulla sommità della sella’. Anderson, cit. (n. 2), 25 observes that Livy 35, 10, 12 records a porticus built only a year after the 194 restoration of Villa Publica and Atrium Libertatis which would almost join them if the Atrium was near the Porta Fontinalis: but in the end takes this, no doubt rightly, as a coincidence.

¹¹³ Tortorici cit. (n. 23), 76–7 on ‘inequivocabili rapporti di continuità’ between Atrium Libertatis and Forum of Caesar. It was, on this view, Caesar who demolished the old Atrium; Pollio's works were part of the new layout. It is worth noting that the reason for this solution is the impracticability of a site between the Temple of Venus and the Republican city-wall.

¹¹⁴ Ovid, , Ars Am. III 951–2Google Scholar; Rem. 659–60. Pliny, , HN 36, 33Google Scholar.

¹¹⁵ Lugli, , Fontes cit. (n. 2), 408–27Google Scholar, from AD 319 on the laws; note also A. Gellius 11, 17, 1, praetor's edicts available there in the second century. Sidonius Apollinaris 2, 544–6 for a very loose tie between the Forum and libertas. On LIBERTATIS, Carrettoni, G., Colini, A. M., Cozza, L., Gatti, G., La Pianta Marmorea di Roma antica (Rome 1960), frag. 29, with p. 72Google Scholar; Almeida, E. Rodriguez, Forma Urbis Marmorea. Aggiornamento generale 1980 (Rome 1981), 109–10Google Scholar. For the libraries, Richardson, L., ‘The architecture of the Forum of Trajan’, ArchNews VI 4 (1977), 101–6Google Scholar, suggesting that against its normal practice in labelling, the Forma Urbis had ATRIUM in its other apse, indicating that the two halves of Pollio's library were set out in the two apses before the actual libraries were ready; Anderson, cit. (n. 2), 173–7, more cautious.

¹¹⁶ Suetonius, Galba 20Google Scholar. Tacitus, Hist I, 31Google Scholar attributes the delay more plausibly to their indecision: the ignorance of the way sounds like a face-saver. We note that when the messengers were sent to summon help from the troops in the Atrium Libertatis and the Porticus Vipsania, two were needed. Hence the places were not in the same direction. But the direct route to the Porticus Vipsania passes the west end of the Forum Iulium! Even supposing that they really did get lost, the point is that we do not know where the messengers who had been sent from the Area Palatina were ordered to tell the reinforcements to assemble. They had no way of knowing that Galba would need them most in the accessible spot of the Lacus Curtius in the middle of the Forum. For all we know they were headed off in a quite different direction to a rendezvous that never took place, and got lost on the way there.

¹¹⁷ For the size, see n. 73, and Tortorici cit. (n. 23), 32 with my n. 113. von Gerkan, A., ‘Die republikanischer Stadtmauer Roms zwischen dem Kapitol und dem Quirinal’, RM 55 (1940), 1–26Google Scholar. The Porta Fontinalis appears close enough to the Forum Romanum for the house of Cn.Piso, ‘imminens Foro’ at Tacitus, , Ann. III 9Google Scholar, to be described as ‘supra portam Fontinalem’ in the new SC de Cn. Pisone patre. And the argument of Bonnefond, cit. (n. 74), 606–7 makes it very likely that the Atrium Libertatis was within the pomerium: thus also Thomsen, cit. (n. 10). Lugli, cit. (n. 102), 810 takes the inscription CIL VI 10025Google Scholar to describe the location of a tomb ‘post Atrium Libertatis’, and therefore to describe the latter as close to the pomerium. But the phrase is clearly a work-address, like ‘post aedem Castoris’ in CIL VI 9872Google Scholar, 10024, 30748 = 363.

¹¹⁸ A reductio ad absurdum of this theory is the untroubled statement of Lugli, cit. (n. 102), 815 that the Atrium Libertatis had four sites and four functions in the Roman period!

¹¹⁹ CIL 10025 (on which cf. n. 117) shows that there were cheap shops in the vicinity of the Atrium in the imperial period. Recent work on the so-called Basilica Argentaria shows that is not in the least suitable as a candidate for a replacement of the huge and complex Atrium: Amici, cit. (n. 10), 101–21.

¹²⁰ Rodriguez Almeida, cit. (n. 115), 109–10. The idea of Libertas was a prominent theme of the principates of Nerva and Trajan, and there was room for more than one evocation of it. In this context we may note the dedication ‘SPQR Libertati restitutae’ of 18th Sept. AD 96, found under Martina, SS e Luca, , CIL 470Google Scholar = ILS 3780. Another such dedication, of 18th Sept. AD 96, CIL VI 916Google Scholar = ILS 274, was seen on the Capitol by the Einsiedeln author. VI 471 has an altogether different provenance.

¹²¹ For the citation of Sherlock Holmes and its place in contemporary historical methodology, see Wickham, C., ‘Marx, Sherlock Holmes and late Roman commerce’, JRS 78 (1988), 183–93 at 186Google Scholar.