Abbreviation

1 The chief monograph on the Farnesina is C. L. Frommel, Die Farnesina und Peruzzis architektonisches Frühwerk (Berlin, 1961; reviewed by Ackerman, J. S., Art Bulletin, XLIV (1962), 243 ff Google Scholar). See also Frommel, , Der römische Palastbau des Hochrenaissance (Tübingen, 1973) (hereafter Frommel) I, 101; II, 149 ff.Google Scholar; III, 60 ff. and Coffin, D. R., The Villa in the Life of Renaissance Rome (Princeton, 1979), pp. 87 ffGoogle Scholar. On Chigi’s patronage see Ray, S., ‘Opere per Agostino Chigi’ in Frommel, C. L., Ray, S., and Tafuri, M., Raffaello Architetto (Rome, 1984), pp. 119–24Google Scholar; Rowland, I., ‘“Render Unto Caesar the Things Which are Caesar’s”: Humanism and the Arts in the Patronage of Agostini Chigi’, Renaissance Quarterly, 39/4, pp. 673–730 Google Scholar; and Quinlan-McGrath, M., ‘A Proposal for the Foundation date of the Villa Farnesina’, Journal of the Warburg and Courtauld Institutes, 49 (1986), 245–50.CrossRefGoogle Scholar

2 Vasari, G., Le Vite dei piu eccellenti Pittori, Scultori e Architettori, ed. Milanesi, (Florence 1878–85), IV, 593.Google Scholar

3 The chief monograph on Palazzo Massimo is Wurm, H., Der Palazzo Massimo alle Colonne (Berlin, 1965)Google Scholar. See also Frommel I, 133 ff, II, 233 ff, III, 92 ff; and idem, ‘Palazzo Massimo alle Colonne’, Atti, pp. 241–62.Google Scholar

4 Portoghesi, P., Rome of the Renaissance (London, 1972), p. 178 ff Google Scholar; Frommel, 1, 136; and Ray, S., ‘Peruzzi architetto e la coerenza di una ricerca’, Atti, p. 49.Google Scholar

5 Frommel, C. L., Die Farnesina, p. 65 ff Google Scholar. He gives the overall width and length as 25.81 m (≃115.7 p) and 36.41 m (≃163.5 p). However, the centre lines of the corner pilasters were in fact the critical limits, see Appendix 1. On the use of √2 in the Renaissance see Saalman, H., ‘Early Renaissance Theory and Practice in Antonio Filarete’s Trattato di Architettura ’, Art Bulletin, 41 (1959), 89–106 CrossRefGoogle Scholar; Thoenes, C., ‘Proportionsstudien an Bramantes Zentralbau- Entwürfen’, Römisches Jahrbuch für Kunstgeschichte, 15 (1975), 37–57 Google Scholar; and M. Wilson Jones, The Tempietto and the Roots of Co-incidence, publication forthcoming (hereafter Wilson Jones). For its use in Antiquity and the Middle Ages, see n. 47 below; Jacobson, D. M., ‘Hadrianic Architecture and Geometry’, Americanjournal of Archaeology, 90 (1986), 69–85 CrossRefGoogle Scholar; Shelby, L. R., ‘Medieval Masons’ Templates’, Journal of the Society of Architectural Historians, XXX (1971), 141–54Google Scholar; Mojon, L., Der Munsterbaumeister Matthäus Ensinger (Bern, 1967)Google Scholar; Bucher, F. ‘Medieval Architectu ral Design Methods 800–1560’, Gesta, XI (1972), 37–51 CrossRefGoogle Scholar; and Kidson, P., Systems of measurement and proportion in early medieval architecture, PhD, University of London, 1956.Google Scholar

6 For measurements of the Farnesina see Appendix 1. The following room proportions seem to have taken precedence in laying out the plan:

7 See Appendix 1. The height today measures about 17.685 m, some 15 cm short of 80 p, because the ground has risen above its original level. There are five existing steps from the ground to the floor, of which the bottom step is 5–7 cm shorter than the rest. When these were made the ground was therefore at least 5–7 cm lower. The sixteenth-century drawing (Fig. 3) shows a total of seven (not so steep) steps and a substantial base moulding. The profile of this moulding may can be seen by the descending service stairs near the present entrance: its vertical face measures 24 cm, some 16 cm more than is visible elsewhere. The original height of the building can therefore be estimated as between 7 and 16 cm taller, that is about 17.75 m and 17.85 m, or 79.59 p and 80.04 p.

8 See Appendix 1. Presumably the widths of room 1, 2, 3, 6, and 7 were intended as 30 palmi. The Sala delle Prospettive measures approximately 40 by 70 palmi, the dimensions specified by Peruzzi in UA565v/Wurm 8 when the room was enlarged some ten years after its initial completion. The wall nearest the stairs was repositioned and the opposite wall reduced in thickness (hence the unhappy additional band on the ceiling above]. The original size of the room was probably 40 by 60 palmi (2:3).

9 Vitruvius, v, 6.

10 For measurements see Appendix 1. This argument was first advanced in my dissertation, The Famesina: Villa, Palazzo and Theatre, Polytechnic of Central London, 1982.

11 It was common Renaissance practice for cornices to be sized to suit the whole building (e. g. Palazzo Medici in Florence), as opposed to that of the top storey alone (e.g. the Cancelleria in Rome).

12 In Palladio’s version the most marked modification is the increased height of the pedestals of the upper order. On the development of this design see Magagnato, L., ‘The Genesis of the Teatro Olympico’ JWCI, 14 (1951), 209–20Google Scholar; and Forster, K. W., ‘Stagecraft and Statecraft: The architectural integration of Public Life in Theatrical Spectacle in Scamozzi’s Theater at Sabbioneta’, Oppositions, 9 (1977), 63–87 Google Scholar. Peruzzi’s use of Vitruvius’ formula for the scaenae jrons may find a precedent in Bramante’s design of the Belvedere courtyard elevation. The relative sizes of the columns/pilasters, pedestals and entablatures closely match Vitruvius’ recommended 4:3 diminution for the orders of— suitably — both the theatre and the Forum (v, 1, 3, see M. Wilson Jones, The Famesina, Fig. 27).

13 To quote Portoghesi, p. 93. See also Frommel, , Die Famesina, p. 114 ff Google Scholar; Coffin p. 107 ff.; Cruciani, F., Teatronel Rinascimento 1450–1550 (Rome, 1983), p. 349 ff.Google Scholar; and Rowland, , ‘Render Unto Caesar …’, p. 712 ff.Google Scholar

14 Frommel, p. 85 ff; Ackerman, J. S., ‘Sources of the Renaissance Villa’, Acts of the XXth International Congress on the History of Art (Princeton, 1963).Google Scholar

15 Lawrenson, T. E. and Purkis, H., ‘Les éditions illustrées de Térence dans l’histoire du théatre’, in Le lieu théâtral a la Renaissance, ed.Jacqot, J. (Paris, 1964), Figs 4, 7 and 10Google Scholar; and Magagnato, Fig. 38, a, b, and c.

16 Originally the Famesina was entered through the loggia of Cupid and Psyche, and the courtyard in turn entered directly off the Via Lungara. On the villa ‘Le Volte’ see Scoppola, F., ‘Villa Chigi alle Volte Alte’, in Rilievi di Fabbriche senesi di attribuzione peruzziana (Siena, 1981), pp. 361–433 Google Scholar; and Fiore, F. P., ‘La Villa Chigi a “Le Volte” eil linguaggio architettonico peruzziano nella tradizione di Francesco di Giorgio’, Atti, pp. 133–67.Google Scholar

17 Cruciani, p. 349 ff; Chastel, A., ‘Cortile et théâtre’ in Le lieu théâtral, pp. 41–47 Google Scholar; Ruffini, F., Teatro prima del Teatro (Rome, 1983)Google Scholar. Peruzzi’s skill in designing theatrical settings is well known, see Krautheimer, R., ‘The Tragic and Comic scene of the Renaissance’, Gazette des Beaux Arts, 33 (1948), 327–46Google Scholar; Cruciani, p. 440 ff, 510 ff, and 572 ff.; idem., ‘Gli allestimenti scenici di Baldassarre Peruzzi’ in Bolletino del Centro Intemazionale di Studi dell’ Architettura: A. Palladio, XVI 1974 (1976)Google Scholar. On the theatrical interests of his contemporaries see Frommel, C. L., ‘Raffaello e il teatro alia corte di Leone X’, in the same volume, pp. 173–88 Google Scholar; and Cerutti Fusco, A., ‘Teatro all’antica e teatro Vitruviano nell’interpretazione di Antonio da Sangallo il giovane’, Antonio da Sangallo ilgiovane; La vita e l’opera, Atti del XXII Congresso di Storia dell’Architettura (Rome, 1986), pp. 455–69.Google Scholar

18 See Appendix 1. Adjustments of this sort (like the additional height given to the crowning entablature compared to the formula for the scaenae frons) were, of course, sanctioned by Vitruvius (v, 6, 7, and VI, 2).

19 Two contemporary poems in honour of Chigi give an insight into intentions regarding the garden: Gallo, E., De Viridarie Augustini Chisii (Rome, 1511)Google Scholar; and Palladio, B., Suburbanum Augustini Chisii (Rome, 1512)Google Scholar. See Coffin, p. 88 ff.; and Rowland, I. ‘Some Panegyrics to Agostino Chigi’, JWCI, 47 (1984), 194–200 Google Scholar. On Peruzzi’s involvement in garden design see Barsali, I. B., ‘Il Peruzzi architetto di giardini’, Atti, pp. 103–32.Google Scholar

20 Saxl, F., The Villa Famesina (London, 1957)Google Scholar; Rowland, I. ‘The Birth Date of Agostino Chigi — Documentary Proof, JWCI, 47 (1984), 192–93Google Scholar; and Quinlan-McGrath.

21 Coffin, p. 108.

22 L. B. Alberti, Momus o del Principe, as translated by Tafuri, M., ‘Discordant Harmony from Alberti to Zuccari’, Architectural Design, 49, no. 5–6 (1979), 36Google Scholar. See also Ponte, G., Leon Battista Alberti: Umanista e scrittore (Genoa, 1981), p. 79 ff Google Scholar.

23 Marcucci, L, ‘Regesto cronico e critico’, Studi e Documenti di Architettura, 8 (1978), 11–184.Google Scholar

24 The principal passages addressing mathematical order in design are Vitruvius, I, 1, 8–9;1, 2:1, 3, 2:II, Intro., 5; III, 1; IV, 3; IV, 4; IV, 6; II, 8; V, 6, 7 and VI, 2 (see also Knell, H., Vitruvs Architekturtheorie (Darmstadt, 1985)Google Scholar). For Crticisms of Vitruvius’ methodology, see Alberti, , De Re Aedificatoria, VI, 1 Google Scholar (and Krautheimer, R., ‘Alberti and Vitruvius’, Studies in Western Art (Princeton, 1963), II, 42–52 Google Scholar); Bertani, G. B., Gli oscuri e difficilipassi dell’opera ionica di Vitruvio (Mantua, 1558)Google Scholar; Gros, P., ‘Structures et limites de la compilation vitruvienne dans les livres III et IV du De Architectural, Latomus, 34/2 (1975), 986–1009 Google Scholar; and Coulton, J. J., ‘Towards understanding Greek temple design: general considerations’, Annual of the British School at Athens, 70 (1975), 69 ff.CrossRefGoogle Scholar

25 Particular interest was aroused by Pliny’s description of the tomb of Lars Porsena (NH, XXXVI, 19), the elements of which supposedly measured 50, 75, 100, 150, and 300 pedes (other passages which cite round dimensions include NH, XXXIV, 41, 43, and 45, and XXXVI, 95–97). See UA 634v/Wurm 465 for a possible reconstruction of Porsena’s tomb by Peruzzi, and for that by Antonio da Sangallo, Borsi, S., ‘Disegni dell’antico di Antonio da Sangallo: Le antichità etrusche’, in Antonio da Sangallo il giovane: La vita e l’opera (Rome, 1986), p. 440 ff Google Scholar., and p. 112, n. 27 of the same volume.

26 Kings, 1, vi and vii. On the possible connections between the dimensions of ecclesiastical buildings and biblical numerology, see, for example, Sunderland, E. R., ‘Symbolic Numbers and Romanesque Church Plans’, JSAH, XVIII (1959), 94–103 Google Scholar; Battisti, E., ‘Il significato simbolico della Capella Sistina’, Commentari, VIII (1957), 96–100 Google Scholar; and Rykwert, J. and Tavernor, R., ‘Sant’ Andrea, Mantua’, Architects’ Journal, May 1986, pp. 36–57 Google Scholar. The overall dimensions of the interior of Alberti’s masterpiece are exactly double the number in terms of Mantuan braccia (40 wide by 120 long by 60 tall) as those of the Temple of Solomon in terms of cubits (20 by 60 by 30).

27 Wittkower, R., Architectural Principles in the Age of Humanism (London, 1962; first pub. 1949)Google Scholar; Hersey, G. L. Pythagorean Palaces: Magic and Architecture in the Italian Renaissance (Ithaca and London, 1976)Google Scholar; Naredi-Rainier, P. v., Architektur und Harmonic: Zahl, Maβ und Proportion in der abenländischen Baukunst (Cologne, 1982), with bibliography, p. 232 ff.Google Scholar

28 Alberti, , De Re Aedificatoria, IX, 5 Google Scholar. See Gadol, J., Leon Battista Alberti (Chicago and London, 1969), p. 108 ff Google Scholar; Aiken, J. A., ‘Leon Battista Alberti’s System of Human Proportions’, JWC, 43 (1980), 68–96 Google ScholarPubMed; and Tavernor, R., Concinnitas in the theory and practice of L. B. Alberti, Cambridge University PhD, 1985.Google Scholar

29 Vitruvius, III, I; Wittkower, p. 89 ff. Hersey, ch. 1; Tavernor, pp. 23–24 and p. 107; Naredi-Rainier, p. 30 ff.

30 Wilson Jones, M., ‘Principles of design in Roman architecture: the setting out of centralised buildings’, Papers of the British School at Rome (1989), Appendix II and n. 64 and 84 Google Scholar; and Coulton.

31 The following is a comparative table of the overall sizes of some ancient Roman centralized monuments in metres and in nominal values of ancient Roman feet and cubits (M. Wilson Jones, ‘Principles of Design’, Appendix I).

32 Note, for example, the 1:1 relationship between the Pantheon’s height and diameter, the 1:2:3 relationship between the three rings of S. Stefano Rotondo, and the 2:1 relationship between the diameter of the stylobate and that of the interior (which sometimes equals the column height) in many peripteral temples such as the so-called Temple of the Sibyl at Tivoli. See also Rolland, H. ‘Le Mausolée de Glanum’, XXIe supplèment à Gallia (Paris, 969)Google Scholar, Amy, R. and Gros, P., La Maison Carrée à Nimes (Paris, 1979)Google Scholar; Kurrent, T., ‘Preferred numbers in modular sizes of Emona, Diocletian’s Palace and Mogorjelo’, Arheoloski Vestnik (1980), pp. 113–31Google Scholar;Jacobson; Rasch, J.J., Das Maxentius-Mausoleum an der Via Appia in Rom (Mainz, 1984)Google Scholar; Almagro-Gorbea, M., ‘Der Junotempel in Gabii und Vitruv’, in Vitruv-Kolloquium, Schriften des Deutschen Archaölogen-verbendes, Band VIII, 198, p. 163 ff.Google Scholar; Davies, P., Hemsoll, D. and Wilson Jones, M., ‘The Pantheon: triumph of Rome or triumph of compromise?’, Art History, 10 (1987), 133–53CrossRefGoogle Scholar; and Wilson Jones, M., ‘The design of the Corinthian order’ in Journal of Roman Archaeology, 1 (1988).Google Scholar

33 Symmetria is a concept central to classical ideas of order ( Panofsky, E., ‘History of the Theory of Human Proportions’, in Meaning and the Visual Arts (New York, 1970), pp. 83–138, especially n. 19Google Scholar; and Pollitt, J. J., The Ancient View of Greek Art: criticism, history and terminology (New Haven, 1974), pp. 14–22 Google Scholar). At the start of Book III Vitruvius asserts that symmetria and proportio are indispensable to good architecture, just as they are to Nature’s design of the well-formed man: ‘Namque non potest aedis ulla sine symmetria atque proportionae rationem habere compositionis, nisi uti ad hominis benefigurati membrorum habuerit exactam rationem’. The well-formed man is the perfect illustration of symmetria, which is manifest in the simplicity and coherence of the underlying measures, ratios, numbers, and shapes — as the following matrix shows:

One of the salient characteristics of Vitruvian Man is the divisibility of the parts into the whole — the overall dimension of 6 ft. Similarly, the Basilica at Fanum, the only building for which Vitruvius quotes dimensions, which is 120 ft long (VI, 5, 6–10); the tomb of Lars Porsena, which is 300 ft long (see n. 25 above); and the specification for the Arsenal at Piraeus, which is 400 ft long ( Jeppeson, K., Paradeigmata (Aarhus, 1958)Google Scholar). There are parallels in many surviving free-standing ancient buildings, see the works cited in n. 30 and 32 above.

34 Frankl, P., ‘The Secret of the Medieval Masons’, Art Bulletin, 27 (1945), 46 ff.Google Scholar; Ackerman, J. S., ‘ Ars sine scientia nihil est’: Gothic Theory of Architecture at the Cathedral of Milan’, Art Bulletin (1949), p. 88 ff Google Scholar; Toker, F., ‘Gothic Architecture by Remote Control: An illustrated building contract of 1340’, Art Bulletin, 67 (1985), pp. 67–95 CrossRefGoogle Scholar; and the works cited in n. 5.

35 Burns, H., ‘Raffaello e “quell’ antiqua architectura’”, in Raffaello Architetto, ed. Frommel, C. L., Ray, S., Tafuri, M. (Milan, 1984), p. 398, n. 48Google Scholar. See also Scaglia, G., Il ‘Vitruvio Magliabechiano’ di Franceso di Giorgio Martini (Florence, 1985)Google Scholar. Franceso di Giorgio’s patchy and awkward translation of Vitruvius is a fascicule accompanying the Codex II, I,141 in the Biblioteca Nazionale, Florence ( Martini, Francesco di Giorgio, Trattati di architectura, ingegneria e arte militare, ed. Maltese, C. and Maltese Degrassi, L. (Milan, 1967), 1, 36ff.Google Scholar). Other Quattrocento translations werejust as flawed, see Scaglia, G., ‘A translation of Vitruvius and Copies of Late Antique Drawings in BuonaccorsoGhiberti’s Zibaldone’, Transactions of the American Philosophical Society, 69/1 (1979), 3–30.Google Scholar

36 See Quednau, R., ‘ Aemulatio veterum. Lo studio e la recenzione dell’antichita in Peruzzi e Raffaello’, Atti, p. 413 ff.Google Scholar, especially n. 62, for Peruzzi’s repetition of Vitruvian terminology. On Peruzzi’s imperfect understanding of Latin see Adams, R. N., Baldassare Peruzzi: Architect to the Republic of Siena 1527–1535, PhD, New York University (Ann Arbor edition) (1977), p. 6 Google Scholar. A translation of Vitruvius was not available in print before Cesariano’s edition of 1521, although Peruzzi may have read Fabio Calvo’s version a few years earlier (see Fontana, V. and Morachiello, P., Vitruvio e Raffaello. Il “De Architectura” di Vitruvio nella traduzione inedita di Fabio Calvo Ravennate (Rome, 1975)Google Scholar).

37 On Peruzzi’s debt to Franceso di Giorgio see Burns, H., ‘Progetti di Francesco di Giorgio per i conventi di San Bernardino e Santa Chiara di Urbino’, in Studi Bramanteschi (Milan, 1970), p. 293 ff.Google Scholar; Marani, P. C., ‘A reworking by Baldassare Peruzzi of Francesco di Giorgio’s Plan of a Villa’, JSAH, 46 (1982), 182–89 Google Scholar; Frommel, C. L., ‘Baldassare Peruzzi pittore e architetto’, Atti, p. 26 ff.Google Scholar; and F. P. Fiore, ‘La Villa Chigi …’. On mathematical aspects of Francesco di Giorgio’s ideas see Millon, H., ‘The Architectural Theory of Francesco di Giorgio’, Art Bulletin 40 (1958), 36–44 CrossRefGoogle Scholar; Betts, R. J., The Architectural Theories of Francesco di Giorgio, Diss. Princeton University, 1971 Google Scholar; and Lowic, L., ‘Francesco di Giorgio on the Design of Churches: The Use and Significance of Mathematics in the Trattato’, Architectura, 4 (1982), 151–63.Google Scholar

38 Saalman, H., ‘Brunelleschi, Capital Studies’, Art Bulletin, 40 (1958), 115 ff.CrossRefGoogle Scholar; Burns, H., ‘Quattrocento architecture and the Antique: some problems’, Classical influences on European Culture A.D. 500–1500, ed. Bolgar, R. R. (Cambridge, 1971), pp. 269–87Google Scholar; and Buddensieg, T., ‘Criticism and Praise of the Pantheon in the Middle Ages and the Renaissance’, in the same volume, pp. 259–67Google Scholar.

39 Ibid., pp. 259–67; Ericsson, C. H., ‘Roman Architecture expressed in the sketches by Francesco di Giorgio Martini’, Commentationes Humanorum Literatum, 66 (Helsinki, 1980)Google Scholar; Vasori, O., I monumenti antichi in Italia nei disegni degli Uffizi (Rome, 1981), p. 12 ff.Google Scholar; and Borsi, S., Giuliano da Sangallo: I disegni di architettura e dell’antico (Rome, 1985)Google Scholar; for example (p.94), the column shafts of the Pantheon are in fact 8 rather than 7 diameters tall.

40 Golzio, V., Raffaello nei documenti (Vatican City, 1936), p. 85.Google Scholar

41 The principal known models for alternating fluting are the Temple of Apollo Sosianus and the Parthian Arch of Augustus. Modillions of this type are found in the main order of the Hadrianeum and several other coeval monuments, see Strong, D. E., ‘Late Hadrianic Architectural Ornament in Rome’, Papers of the British School at Rome, 21 (1953), pp. 118–51.CrossRefGoogle Scholar

42 UA 631r/Wurm 459 gives the height of the shafts of the Temple of Castor as br or 12.410 m; the actual average height is 12.405 m (M. Wilson Jones, ‘The Design of the Corinthian Order’, Appendix 1, 3). See also the subsequent notes below and Vasori, pp. 39–76; Burns, H., ‘A Peruzzi drawing in Ferrara’, Mitteilungen des Kunsthistorischen institutes in Florenz, 12 (1966), 245–70 (hereafter Burns)Google Scholar, and R. Quednau, ‘Aemulatio veterum …’.

43 On Peruzzi’s reliance on the drawings of others for details of ancient monuments in Verona, see Burns, p. 252 ff. His hasty survey of the Mausoleum of Maxentius on the Via Appia (UA 488/Wurm 407 and 408), for example, gives accurate values for parts that could be measured easily, but was otherwise less precise (such cases are indicated by asterisks):

The error in measuring the porticoes may have resulted from multiplying the number of bays by the measured width of one of them, perhaps one that was unrepresentative. In the tomb itself Peruzzi seems to have measured the annular corridor and the niches of the central octagonal pier, but estimated the width of the pier itself and of the whole space (which would be understandable had he been without an assistant or the time to calculate the width of the pillar from that of its facets). His guess displays his preference for round dimensions and simple ratios: the presumed overall width of 112 p is compatible with a pier width of 48 p, or times the width of the surrounding corridor (32 + 48 + 32 = 112). (In fact the actual dimensions are simply related to one another, the central pier being as wide as the annular corridor (24 + 32 + 24 = 80 pedes of 0.2956 m, roughly equivalent to 32 + 42⅔ + 32 = 106⅔ p)).

44 Zervas, D. F., ‘The Florentine braccio dapanna ’, Architectura, 9 (1979), 6–11.Google Scholar

45 Wilson Jones; Tavernor, passim; and Coffin, D. R., ‘Pope Marcellus II and architecture’, Architectura, 9 (1979), 12.Google Scholar

46 UA 1166r and UA 1165. 10 p equals approximately 5 ancient cubits.

47 UA 625r/Wurm 145. The width of the Curia measures 17.75m, or 60 pedes of 0.2958 m ( Bartoli, A., Curia Senatus: lo scavo e il restauro (Rome, 1963)Google Scholar). The length of the interior, the external width of the front faҫade, and the interior and the external heights are approximately equal to each other and to √2 X 60 P (respectively 25.63, ≃24.85 m, >23.4 and ≃25.4 m). The tomb chamber of the pyramid of Gaius Cestius, I4by20by 14 pedes (4.10 by 5.89 by 4.12 m) is another example of a rectangular space set out using √2. Although Vitruvius mentioned √2 when discussing the proportions of rectangular spaces (VI, 3, 3), in practice it was more usual for this ratio to be used in the setting out square or circular plans/architectural elements, seeJacobson; and Wilson Jones, ‘Principles of Design’.

48 UA 419v/Wurm 59. The diameter of the building’s drum is about 29.55 m, or 100 pedes of 0.2955 m ( Fellmann, R., Das Grab des Lucius Munatius Plancus bei Gaeta (Basle, 1957)Google Scholar; and n. 32 above).

49 On the Renaissance palmo see Thoenes, C., ‘Proportionsstudien an Bramantes Zentralbau-Entwürfen’, Römisches Jahrbuch für Kunstgeschichte, 15 (1975), 57.Google Scholar On ancient mensuration see Fernie, p. 384 ff. and 397 (bibliography). Theancientpes varied generally between 0.294 and 0.297 m (cf. n. 31 above), and either 0.2955 or 0.2960 would seem to be acceptable as nominal values.

50 UA 632r/Wurm 463. Apart from induction from surveyed measurements (a method fraught with possibilities for error), the length of the ancient units could have been ascertained from surviving measuring rods, such as those in Palazzo Conservatori (Coffin, cited n. 44). A unique piece of evidence recorded by Peruzzi (UA 405v/Wurm 48) was a series of ancient horizontal levels on a road cutting near Terracina, labelled at intervals of 10 pedes.

51 UA 477r/Wurm 469; the width of the base moulding of the drum is given as br, equivalent to 29.64 m, in actuality this is 29.63 m or 100 pedes of 0.2963 m.

52 Peruzzi rarely made explicit distinctions between ancient and contemporary units. One exception is the note ‘misurato con pe antiquo ptito in xvi digiti e ogni digito in grane 4’ on UA 413r/Wurm 74, which shows details of an antique architrave. Generally it is impossible to distinguish between the use of antique or contemporary units because of the margins for error both in modern surveys and those of Peruzzi, and because it may not be clear to what limits he was measuring, see n. 54. The situation is further confused because Peruzzi occasionally used ancient units when he was apparently using contemporary ones. His measurements of the columns of the Temple of Mars Ultor (UA 632v/Wurm 460; cf. Serlio, , l quattro libri d’architectura di Sebastiano Serlio Bolognese, (Venice, 1966) III, f. 39)Google Scholar are inaccurate in terms of the ‘braccjo fiorentjno’ with which it was supposedly measured, but consistently and more characteristically accurate if this braccio equalled 2 pedes. (The column height is given as br29 44 m51 — if the braccio was the typical Florentine unit of 0.584 m this yields 17.43 m. but if the braccio was 2 pedes, or 0.592 m this yields 17.67 m, which compares favourably with the actual height of 17.74 m, 60 pedes).

53 UA 108/Wurm 117 and 118 (according to Krautheimer, R., Corbett, S., Frazer, A., Corpus Basilicamm Christianarum Romae, v (1977), 232, the additions are in another hand, but in fact the change seems to be only in the ink and writing implementGoogle Scholar). Serlio wrote that both Peruzzi and Bramante used ancient units in their respective designs for St Peter’s and the Tempietto (III, f. 64V, 65V and 67V), but it is not clear whether this alludes to a difference of length in comparison to equivalent contemporary units. That differences in length were known is implicit in a letter written by Tolomei, Claudio in 1542, in which he mentions ‘true measurements according to the length of the Roman foot, with a note on its proportion in relation to the measurements of our own times’ (Portoghesi, p. 185).Google Scholar

54 UA 462r/Wurm 471. The diameter of the Pantheon measured between opposing pilaster/column faces is about 43.41 m ± 7 cm, or approximately 194.32 contemporary palmi of 0.2234 m. Alternatively, he could have been measuring between opposing column bases, a distance of about 43 m or 194 aantique palmi of 0.2216 m.

55 UA 574v/Wurm 426 (see Burns, n. 10). The Pantheon measures about 44.52 m to opposing column centres, equivalent to about Renaissance palmi or 201 antique palmi/ pedes/ cubits, see P. Davies, D. Hemsoll and M. Wilson Jones, p. 140 ff. and Fig. 9. The braccio used here measures about 0.667 m, like no other that Peruzzi used. Why he made use of the unorthodox braccio is unclear, but it raises the possibility that, as was said of Brunelleschi, he wished to keep some of his ideas to himself. See A. di Tuccio Manetti, The Life of Brunelleschi,Intro, and commentary by H. Saalman (Pennsylvania State University, 1970), p. 51 (with Saalman’s qualifications, p. 29).

56 Wurm, , Der Palazzo …, p. 191 ff.Google Scholar; Lotz, W., ‘Die ovalen Kirchenräume des Cinquecento’, Römische Jahrbuch des Kunstgeschichte, 7 (1955), 7–100 Google Scholar; and Licht, M., L’edificio apianta centrale. Lo svilippo del disegno architettonic o Rinascimento (Florence, 1984), especially p. 123 ff. and 134 ff.Google Scholar

57 Portoghesi’s observations (p. 95) are a rare exception.

58 UA 593r/Wurm 144; UA 625r/Wurm 145; UA 616r/Wurm 263; and UA 596r/Wurm 323, for example.

59 UA 472r/Wurm 291 (tower, or tribune?), diameters of 10, 20, and 100 p (10 canne); UA 475r/Wurm 292 (tower), diameter of 100 p (10 canne); UA 505r/Wurm II (project for S. Giovanni dei Fiorentini?) 90 and 50 p widths of main spaces, 20 by 10 p niches; UA 492v/Wurm 206 (church), widths of 40 and 20 units; UA 344r/ Wurm 229 (project for S. Domenico), 20 br radius of curving walls.

60 UA 510r/Wurm 9. Scaling off UA 505r/Wurm II, an alternative project for S. Giovanni, on the basis of the 90 and 50 p specified widths suggests that the external width was here also intended as 200 p. See also UA 346r/Wurm 257 (plan for the villa at Belcaro): the courtyard is 36 by 18 br (2:1) and the Triclinio is 18 by 12 br(3:2); UA 513r/Wurm 345 (church for ‘Via del Campidoglio’): the portico is 20 p wide, the walls 20 p thick, the niches 32 p ( X 20 p) wide, the dome 96 p wide (3 x 32 p), and hence the width to the opposing piers must be approximately 100 p (5 X 20 p); Vienna, Nationalbibliothek, Codex 10935 b.136r/Wurm 262 (villa), only the central rooms are specified as ‘quadrata’, other rooms include those that are 45 by 30 p (3:2), 45 by 27 p (5:3), 33 by (5:4), and 62 by 21 p (approx. 3:1); UA 356r/Wurm 327 (project ‘B’ for Palazzo Ricci at Montepulciano), only the courtyard is specified as ‘inquadrato’, but one of the rooms is also square as its sides are the same width (‘p(er) ogni faccia b(raccia) VIII’), the main room is 16⅔ by 10 br (5:3), and the two antecamere to one side are 8 by 6⅓ br (5:4) (similar attempts to reconcile the constraints of the existing site with satisfactory proportions can be seen in the other surviving projects for Palazzo Ricci, UA 355r/Wurm 325, UA 357r/Wurm 329 and UA 358r/ Wurm 333); UA 495r/Wurm 344: scaling dots indicate an internal width of 60 units, hence the sides of the hexagon measure 30 units; UA 474r/Wurm 293, (addition produces a diameter of 60 br).

61 Vitruvius, IV, 6, recommends height to width (measured at the bottom) ratios for apertures as follows: Doric and Attic, 24:II; Ionic, 5:2. Although Alberti (De Re Aedificatoria, I, 13) preferred the double square proportion some later writers returned Vitruvian ideals. Although Serlio repeated Vitruvius’ recommendations, the additional alternatives he supplied were all double squares (IV, I, f. 21 ff.). Palladio also copied Vitruvius’ proportions for doors, but recommended windows to be double square (I Quattro Libri deli’Architettura, I, 25). The following is a selection of buildings with doors which have 2:1 openings.

Sources: a: Desgodetz, A., Les Edifices antiques de Rome (Paris, 1682), ch. 1.Google Scholar b: Amy and Gros, II, pls 32 and 37. c: Stegmann, C. v. and Geymuller, H. v., Die Architektur in der Toscana, 12 vols (Munich 1885–1907), vols 1 and 2Google Scholar. d: Tavernor, p. 390 ff. e: Author, f: Letarouilly, 1, pls. 1–3 and II, pls. 115–39.

For the doors in the Farnesina see Appendix I, and for those of Palazzo Massimo see Appendix 2. The doors off the main salone of Peruzzi’s Palazzo Pollini in Siena also have an opening ratio of 2:1,see Sani, C. and Franchina, L., ‘Palazzo Celsi-Pollini’, in Rilievi di Fabbriche senesi, p. 288 ff. (no dimensions supplied).Google Scholar

62 See n. 58 and 59 above for examples of Peruzzi’s use of the braccio and the palmo when designing. For examples of his use of the piede, see UA 625r/Wurm 145, UA38ov/Wurm 347, UA 501/Wurm 349 and 350, and UA 546r/Wurm 310.

63 The Tempietto’s internal width (measured to the faces of opposing pilasters) is 20 p (4.46 m or 20.02 p), its external width (measured to the faces of opposing pilasters) is 20 P (5.92 m or 19.94 P), and, furthermore, the interaxial diameter of the colonnade is 20 column diameters (6.94 m or 20.06 column diameters).

64 For the history and documentation regarding Palazzo Massimo see Wurm, , Der Palazzo …, ch. 1 and 2 Google Scholar; Frommel, II, 233—50. Legal agreement to divide the family’s property was reached in February 1532 (Frommel, II, p. 234).

65 UA 368r/Wurm 317. See Wurm, H., Der Palazzo … p. 24 ff.Google Scholar; Frommel, II, 238–39 and 243 ff.; idem, ‘Palazzo Massimo …’, p. 245 ff. Google Scholar

66 Ray, S., ‘Peruzzi architetto …’, Atti, p. 50 ff.Google Scholar

67 Adding together written and scaled dimensions for the economical alternative produces an overall depth in the region of 154 p. The extra 6 p was achieved both by the curvature at the front and by rotating and squaring up the rear wall (see H. Wurm, Der Palazzo…, Figs 5, 6, and 8 for superimposed plans).

68 Measuring tangentially (i.e. following the curve of the portico), the part dimensions of the faҫade give a total of . The intention that many of the individual elements be integral numbers of palmi presumably determined the shortfall from 120 p of ⅔p. The central bay is wider than the typical, in accordance with numerous antique precedents, but Peruzzi could not have made it even wider and yet maintain regular bugnati on the upper floors.

69 The dimensions of the part open to the sky are 48 by 40 p/36 by 30 P (6:5). The following is a comparative table of the sizes of some notable courtyards, either the part open to the air (marked with an asterisk), or that of the overall space including arcades:

Sources: c, e, and fare cited in n. 61 above, g: Tönnesmann, A., ‘Der Palazzo Gondi in Florenz’, Römische Studien der Bibliotheca Hertziana (Rome, 1983), Taf. 27 Google Scholar; h: Giinter, H., ‘Bramantes Hofprojekt um den Tempietto und seine darstellung in Serlios dritten buch’, Studi Bramanteschi (Rome, 1970), p. 483 ff.Google Scholar

70 In the Quattrocento the dimensions of internal spaces are generally measured to wall surfaces. For example, the nave of S. Maria del Calcinaio is 20 br wide from wall to wall, and only some br to the pilaster faces. By the beginning of the Cinquecento the increased importance of the architectural orders becomes manifest in the tendency for dimensions to be fixed by the position of the pilaster faces (see Wilson Jones, and n. 63 above).

71 As the faҫade is 90 p long, the bottom storey is proportioned 3:1. The overall height, about 70 P or p, is almost exactly the hypothetical overall width of the faҫade (to the ends, or pilaster edges) of 93 p (see n. 68 above). Given that the height of the faҫade is in fact between 69 and 70 P, an alternative reading is possible, that this was also intended as 3 times the height of the lower order (excluding the balconies) of 23 P (23 X 3 = 69).

72 UA 531r/Wurm 319. See UA 530r/Wurm 318 for an earlier version.

73 See Howard, D. and Longair, M., ‘Harmonic Proportion and Palladia’s Quattro Libri’, JSAH, 41 (1982), 116–43Google Scholar. The following is a comparative table of the proportions advocated by the most influential architectural treatises.

74 See n. 33 above.

75 Alberti, IX, 10 (as translated in Chastel, A., The Studios and Styles of the Renaissance (London, 1965), p. 28 Google Scholar).

76 Frommel, C. L., ‘Palazzo Massimo …’, p. 252 ff.Google Scholar; Wurm, H., Der Palazzo … p. 89 ff.Google Scholar

77 M. Wilson Jones, ‘The design of the Corinthian order’, section 5; Serlio (IV, I, f. 17 and IV, 8, f. 46) was also aware of the parity between the plinth and the abacus, and that between the astragal and the lower part of the shaft.

78 M. Wilson Jones, ‘The design of the Corinthian capital’, forthcoming.

79 The dimensions given by Peruzzi in his most elaborated design for a Corinthian column (Chatsworth, Duke of Devonshire’s collection, Inv. n. 40/Wurm 158) do not adhere exactly to the 6:5 rule. However it is not impossible that he had this proportion in mind before rounding off dimensions for the sake of convenience: the column height (H) is 6⅚ br and that of the shaft (h) is 5⅔ br; H/h is therefore 1.2058 or + 0.48%.

81 Cellini, B., ‘Discorso di M. Benevuto Cellini dell’ Architettura’, in Tassi, F., Ricordi, prose, e poesie di Benevuto Cellini (Florence, 1829), vol. III, 369.Google Scholar

82 Gros, P., Aurea Templa (Rome, 1976), pl. LVI.Google Scholar Of course, Peruzzi may have seen such a pattern in a building (or mosaic?) now lost.

83 Raffaello Architetto, especially pp. 136 ff., 197 ff., 311 ff., 381 ff.

84 Frommel, , ‘Palazzo Massimo …’, p. 252 Google Scholar. Cf. idem et al, Le Palais Farnèse (Rome, 1981), I, 127 ff Google Scholar

85 Frommel, 1, 93, 103, 105, 108, and 111.

86 Pagliara, P. N., ‘Palazzo Alberini’, in Raffaello Architetto, p. 171.Google Scholar

87 The following is a comparative table of the average bay widths and the heights of windows (piano nobile) of the facades of selected palaces in Rome:

88 P. N. Pagliara, cited n. 86 above.

89 Both Peruzzi and his patrons must have agreed with Altieri’s complaints of the 1520s against the papal policy, begun under Sixtus IV, of blocking up the once common street porticoes. See Altieri, M. A., Li nuptiale di Marco Antonio Altieri, ed. Narducci, (1873), p. 15 Google Scholar. The vestibule is an element that is common to both the Farnesina and Palazzo Massimo, see Ray, S., ‘Peruzzi architetto …’, p. 56 Google Scholar. See also Alberti’s comments on the decorum of porticoes (De Re Aedificatoria, IX, 4).

90 See n. 87.

91 The type of surround for the upper windows, with an architrave which steps out at the corners, was an antique motif visible, for example, in the so-called Temple of the Sibyl at Tivoli and the Temple of Hercules at Cori. A cruder version of the lower surrounds – straight at top and bottom and only curvilinear on the sides – was used by Giulio Romano in Palazzo Maccarani in about 1520, see Frommel, C. L., ‘Palazzo Massimo …’, p. 249 and Fig. 7.Google Scholar

92 This cycle itself was painted after Peruzzi’s death, probably by Daniele della Volterra. See Wurm, H., Der Palazzo … p. 265 ff.Google Scholar; Stricchia Santoro, F., ‘Daniele di Volterra’, Paragone, 213 (1967), 8 ff.Google Scholar

93 Codex Escurialensis, fol. 14. See Dacos, N., La découverte de la Domus Aurea et la formation des grotesques a la Renaissance (London, 1969), pl. XXXI Google Scholar. Other types of curvilinear frames decorated the Volta gialla in the Domus Aurea, see pl. XXIX. For parallels in the grotteschi of Peruzzi and his assistants see Frommel, C. L., Baldassare Peruzzi als Maler und Zeichner (Vienna, 1967), taf. XLIIIa, XLVb, LXXXIII and LXXXIVGoogle Scholar. This interpretation of Peruzzi’s intentions in using this type of window surround conflicts with that of Frommel (‘Palazzo Massimo …’, p. 249), who regards them as allusions to a supposed Ionic style in the upper parts of Palazzo Massimo’s faҫade. Frommel argues that the Ionic style is suggested by the main portal below, and the windows of the first floor (both having pulvinated friezes and scroll brackets in keeping with Vitruvius’ recommendations). This argument applies much better in the case of Palazzo Maccarani, where the partly curvilinear surrounds seem to mimic the plan of Ionic capitals (cf. the plans of Ionic capitals in Antonio da Sangallo’s drawing, Fig. 4 (UA 1166r)), and, where they are indeed part of an Ionic storey complete with pilasters and trabeation. In Palazzo Massimo the presence of pulvinated friezes in the windows and doors may have reflected another theme — all such examples are found in walls running laterally (i.e. in the same sense as the curving street facade), while all those with flat friezes are found in walls running longitudinally.

94 Davies, P. and Hemsoll, D., ‘Entasis and diminution in the design of Renaissance pilasters’, Actes de la Centre d’Etudes Superieures de la Renaissance (Tours, 1986 (1988)).Google Scholar

95 Ibid.

96 In the place where I measured, the top part of the back wall of the portico leans in with respect to the lower part by about 6 mm.

97 In antiquity the majority of columns do not widen before diminishing towards the upper part. However, there are some examples which do, notably those of the Pantheon’s interior, see Stevens, G., ‘The entasis of Roman columns’, Memoirs of the American Academy in Rome, 4 (1924), 145.CrossRefGoogle Scholar

98 This marginal increase in size is maintained in all of the pilasters in the courtyard without exception. The increase is never less than 5 mm and never more than 11 mm.

99 Bibliotheca Communale di Siena, Taccuino S IV, 7, facsimile ed., Sovicille (1981), p. 118 (fol. 60).

100 Marconi, P., ‘Le facciate della Farnesina Chigi e del Palazzo Massimo alle Colonne. Osservazioni sulle techniche e problemi di conservazione e restauro, Atti, p. 717 Google Scholar. For a review of the current restorations see Art e Dossier, 18 (Florence, November 1987).