¹ Teubner, Scriptores Historiae Augustae, edited E. Hohl, 1965, 190–1.

² See C. Lessing, Scriptorum Historiae Augustae Lexicon, under the relevant entries; solearis, cameratio and mechana are in fact omitted from the Oxford Latin Dictionary.

³ The passage was rejected entirely by Domaszewski, , Die Topographie Roms bei den SHA, Heidelberg 1916, 7Google Scholar. On the other hand, in his review of Erika Brödner's Untersuchungen an den Caracalla Thermen, in JRS 1953, 210–12, J. B. Ward Perkins wrote: ‘For the present the text of the SHA must remain the touchstone for the validity of any restoration.’

⁴ Guattani, G. A., Della Gran Cella Soleare nelle Terme di Antonio Caracalla, Rome, 1783, 35 ffGoogle Scholar.

⁵ Canina, , Gli Edifiii di Roma Antica, Vol. III, Rome, 1851–1856, 80–1Google Scholar.

⁶ Iwanoff, S., Architektonische Studien III, Berlin, 1898Google Scholar, Tafel IX–XIII.

⁷ Blouet, A., Restauration des Thermes d'Antonin Caracalla à Rome, Paris, 1828, 17Google Scholar.

⁸ Lanciani, R., The Ruins and Excavations of Ancient Rome, London, 1897, 537Google Scholar.

⁹ Middleton, J. H., The Remains of Ancient Rome, Vol. II, London, 1892, 163Google Scholar.

¹⁰ G. Aitchison, RIBA, 23 May 1889, 282.

¹¹ Hülsen, Ch., Architektonische Studien III, Berlin, 1898, 33–4Google Scholar.

¹² Brödner, E., Untersuchungen an den Caracalla—thermen, Berlin, 1951, 14–18Google Scholar.

¹³ Cf. Guattani, op. cit., 35.

¹⁴ de Pachtere, F. G., ‘Sur la Cella Soliaris des thermes de Caracalla’, MEFRA XXIX, 1909, 401–6CrossRefGoogle Scholar.

¹⁵ Ward-Perkins, op. cit., 211; Brodner, op. cit., 18.

¹⁶ For example, Rivoira, G., Roman Architecture, Oxford, 1925, 171–2Google Scholar; Anderson, W. J., Spiers, R. P. and Ashby, T., The Architecture of Ancient Rome, London 1927, 104Google Scholar.

¹⁷ Maiuri, A., ‘Significato e natura del solium nelle terme romane’, La Parola del Passato, 5 (1950), 223–7Google Scholar.

¹⁸ I am indebted to John Patterson for this suggestion.

¹⁹ I have already discussed these briefly in J. DeLaine, ‘An engineering approach to Roman building techniques: the Baths of Caracalla in Rome’, Papers in Italian Archaeology IV: Part iv, Classical and Medieval, BAR International Series 246, 1985, 195–206.

²⁰ Published by Giuliani, C. F., ‘Volte e cupole a doppia calotta in età adriánea’, RM 82, 1975, 329–42Google Scholar.

²¹ Using the numbering system of the plan published in ARID Suppl. X. Tav. II. This area of the Domus Aurea was published by Zander, G. ‘La Domus Aurea: nuovi problemi architettonici’, Boll Centro 12, 1958, 35–64Google Scholar, but he does not mention the cavities in the vault.

²² The designation of the area is that of Licht, K. de Fine, ‘Untersuchungen an den Trajansthermen zu Rom’, ARID VII Suppl., 1974Google Scholar. He does not discuss the cavities in the vault, 40–3, nor are they shown on his elevation, Taf. III.

²³ See Sear, F. B., Roman Wall and Vault Mosaics, Heidelberg, 1977, 127Google Scholar; confirmed by personal observation.

²⁴ Sear, loc. cit. Traces of mosaic over the tile lining are still visible at the springing of the vault over the west pier.

²⁵ Archivio Centrale dello Stato: Ministero Pubblico Istruzione, Direzione Generale Antichitá e Belli Arti, Antichitá e Scavi, Roma. Rapporto Pellegrini, 1873–4, Busta 102, Fasc. 134.

1873: Terme Antoniniane o di Caracalla

7–12 Aprile: Proseguendosi a sterrare fino al suo piano la gran cella caldaria o vasta cella centrale, non s'incontrano fin qui che i massi caduti della sua volta; e non si é rinvenuta alcuna cosa relativa a frammenti della sua decorazione architettonica, e ne anche de scultura. Fra detti massi si é ritrovato qualche pezzo delle impernature o grappe di ferro che servivano a reggere gli ornati come altre in questo luogo discoperte.

14–19 Aprile: Proseguendosi lo sterro fino al suo piano della gran cella calidaria, o vasta cella centrale delle terme, si sono ritrovate fra i massi della sua volta caduti, tre chiavi, o palettoni di catene di ferro, il primo rotto alia lunghezza di cm70, e gli altri due a cm54…. Item 14: Tre frammenti delle chiavi di ferro nominate disopra.

Elsewhere part of the ‘gran cella calidaria’ is described as being ‘verso il propnigeo da cui si entrava al grande laconico di forma rotunda’. It cannot, therefore, be the room usually designated the caldarium but must refer to the frigidarium.

²⁶ See I. Gismondi, ‘La sala del Planetario nelle Terme Diocleziane’, Architettura ed Arti Decorative, 1929, 385 ff. I am grateful to Dttssa Daniela Candilio and her associates at the Baths of Diocletian for the opportunity to examine the dome during restoration in December 1984.

²⁷ Cf. Note 16.

²⁸ Berlin, Staatlichen Kunstbibliothek, 4151.

²⁹ Suetonius, Nero, 31.

³⁰ The date is A.D. 115. It is interesting, although probably of no significance, that this inscription was found ‘in vinea ad Thermas Antoninianas’.

³¹ See Licht, K. de Fine, The Rotunda in Rome, Copenhagen, 1968, 48–58Google Scholar. He suggests that there was a vaulted false ceiling either of bronze plates on a wooden base or possibly entirely of bronze suspended from the roof truss.

³² Pausanias, V. 12.6.

³³ As suggested originally by Dessau, H., ‘Über Zeit und Persönlichkeit der S.H.A.’, Hermes XXIV (1889), 337Google Scholar, and convincingly restated on numerous occasions by Syme, most recently in Historia Augusta Papers, Oxford, 1983, 219–20Google Scholar.

³⁴ See Birley, E., ‘Africana in the Historia Augusta’, Bonner Historia Augusta Colloquium, 1968/69, Bonn, 1970, 79–90Google Scholar.

³⁵ For example: camera: Vitruvius 5.10.3 and 8.6.3; Valerius Maximus 6.7.2. concameratio: Vitruvius 2.4.2 and 6.8.1; Frontinus de Aquis 126.

³⁶ Maximus et Balbinus 9, 4; Quadrigae Tyrannorum 4, 3.

³⁷ Hadrianus 4, 10.

³⁸ F. G. de Pachtere, op. cit., 402.

³⁹ F. G. de Pachtere, loc. cit.

⁴⁰ C. F. Giuliani, op. cit., 337–9.

⁴¹ Of these two possibilities, the latter is more probable on technical grounds. Iron has a much higher Young's modulus than bronze or copper (30 × 10⁶ compared with 16 × 10⁶) and thus an iron bar would suffer only half as much elongation under a given dead load as a bronze or copper bar of the same cross-section. There is therefore a considerable structural disadvantage in using bronze or copper in place of iron. De Pachtere's suggestion (loc. cit.) that bronze or copper would have the advantage of not oxidising in the warm, moist atmosphere as easily as iron would only be relevant if the metal were exposed, which it would not be if embedded in the concrete of the vault, or for that matter if protected by the plaster or mortar of the vault decoration. Finally, the evidence of the attested use of large, iron, supports in the hot rooms and the barrel vaults linking the caldarium piers requires a strong counterargument of practical advantage for their replacement by bronze or copper in the caldarium, and I cannot see that this exists.

⁴² De Pachtere, op. cit., 401.

⁴³ Although mechanis is another hapax legomenon I do not think that it is at all problematic. Mechanicus appears twice, both in the life of Alexander Severus; on the first occasion (Alex. Sev. 22, 4) ‘mechanica opera … instituit’ appears in a list of his good deeds at Rome, and on the second (Alex. Sev. 44, 4) ‘mechanicis’ appear among a list of professionals including ‘mathematicis’ and ‘architectis’. The mechanici, and hence the docti mechanis, would thus seem to be the engineers and specifically the civil engineers. The form mechana is clearly derived from the Greek μηχανή, and used in place of the Latin machina which does not occur in the Historia Augusta.

⁴⁴ This paper is an offshoot of a much broader continuing investigation into the construction of the Baths of Caracalla in Rome. Most of the material was collected during various periods of residence at the British School at Rome from 1982–1985 and I would like to thank Dr. David Whitehouse, Dr. Graeme Barker and all the staff at the School for their assistance during my frequent visits there. I am grateful to Dottssa. I. Jacopi of the Soprintendenza Archeologica di Roma for permission to work on the Baths, and to her colleagues at the site for their cooperation. My examination of the high vaults of some of the rooms would not have been possible without the help of Prof. Lucos Cozza and his associates from the Servizio Giardino of the Comune di Roma. Among the many people on whom I have tested these ideas I would particularly like to thank Amanda Claridge, Dr. Jim Coulton, Dr. John Patterson, Prof. Peter Wiseman and my colleagues from the Department of Classics at the University of Adelaide for their advice, encouragement and occasional disbelief.