1 The pioneer history of the building is Connors, Joseph, Borromini and the Roman Oratory, Style and Society(Cambridge, MA, and London, 1980)Google Scholar.

2 See n. 67.

3 Published by Sebastiano Giannini as Francesco Borromini, Opus architectonicum(Rome, 1725; facsimile reprint Farnborough, 1964)Google Scholar. See now the English version with commentary: Downes, Kerry, tr. and ed., Borromini's Book, the ‘Full Relation of the Building’ of the Roman Oratory (Wetherby, 2009).Google Scholar For the manuscript text, see Borromini, Francesco, Opus architectonicum, ed. Connors, Joseph (Milan, 1998)Google Scholar, where both Spada's manuscript amendments and Giannini's 1725 changes and lapses are shown. Images of the oratory complex are not easy to find, but Downes, Borromini's Book contains reductions of Giannini's 67 plates together with some 40 photographs of the building.

4 Virgilio Spada and his cardinal brother Bernardo continued to employ Maruscelli elsewhere, not without finding occasional faults both structural and aesthetic ( Connors, , Borromini and the Roman Oratory, p. 111)Google Scholar.

5 As in n. 3.

6 Downes, , Borromini's Book, p. 52.Google Scholar Questo libro fu fatto da me in nome del Cavre Borromino.

7 Connors ( Borromini, , Opus architectonicum, ed. Connors, , pp. xxiii, xxv—xxviiGoogle Scholar) sees in the text the mind of a trained theologian, believing that Borromini was not intellectual enough to compose it. We should remember, however, that he had a library of about a thousand books.

8 della Rocchetta, Giovanni Incisa, ‘Un dialogo del P. Virgilio Spada sulla fabbrica dei Filippini’, in Archivio della Società Romana di Storia Patria, 90 (1967), pp. 165–211 Google Scholar, hereafter cited as Dialogo. The statement that he had his rooms under the clock tower in the past tense (Io ebbi due stanze,p. 201) suggests a date after his appointment to the Ospedale di S. Spirito in March 1661.

9 See Downes, , Borromini's Book, pp. 505-09.Google Scholar

10 della Rocchetta, G. Incisa and Connors, Joseph, ‘Documenti sul complesso borrominiano alia Vallicella (1617- 1800)’, in Archivio della Società Romana di Storia Patria, 104 (1981), pp. 159–326 Google Scholar, hereafter cited as Decreti. Some other important documents between 1623 and 1644, including some accounts, were calendared in Pollak, Oskar, Die Kunsttätigkeit unter Urban VIII. Kirchliche Bauten … und Paläste (Vienna, 1928)Google Scholar, hereafter cited according to given register numbers.

11 Regarding the oval shape of the recreation room: ‘The Provost had read out […] the paper made by His Reverence and endorsed by the architect, as to why it proves impossible to make the room on a rectangular plan’ (Decreti, 146: 13 January 1640). A good enough reason for the shape would be that the refectory beneath it — a year into construction — was also oval.

12 15 April 1638 concerning the oratory's lower floor level (Decreti, 108), and (probably) 23 November 1641 with outside consultants (Decreti, 174; see below at n. 98).

13 Connors, , Borromini and the Roman Oratory, pp. 75, 77 and 48Google Scholar(where the example is Spada's paper). Since this paper is lost, there is nothing to say about its content, structure or style.

14 Baldinucci, Filippo, Vita del Cavalier Gio. Lorenzo Bernino (Florence, 1682)Google Scholar; Bernini, Domenico, Vita del Cavalier Gio. Lorenzo Bernini (Rome, 1713)Google Scholar. For Borromini's subsequent critical fortune, see Blunt, Anthony, Borromini (1979), pp. 218-22.Google Scholar For a comprehensive synthesis of the disagreements between Borromini and Bernini in the context of their lives and work, see Burbaum, Sabine, Die Rivalität zwischen Francesco Borromini und Gianlorenzo Bernini (Oberhausen, 1999).Google Scholar A full account of the acrimonious affair of the St Peter's facade is given by McPhee, Sarah, Bernini and the Bell Towers, Architecture and Politics at the Vatican (New Haven and London, 2002).Google Scholar For Bernini's early critical fortune, see Delbeke, Maarten, Levy, Evonne and Ostrow, Steven F., eds., Bernini's Biographies; Critical Essays (University Park, 2006)Google Scholar which contains seven references to Borromini. Previously unconsidered material about relations between the two architects is in Downes, Borromini's Book.

15 Dialogo,p. 181 Google Scholar.

16 Downes, , Borromini's Book, p. 63 Google Scholar; Opus, p. 7 Google Scholar. Italics in text are mine.

17 Connors, , Borromini and the Roman Oratory, pp. 20, 24 (italics mine)Google Scholar; also p. 192 leading to a quite unnecessary date amendment.

18 Downes, , Borromini's Book, p. 63 Google Scholar; Opus, p. 8 Google Scholar.

19 non averlo formato perfettamente(ibid., p. 83; Opus,p. 14).

20 Downes, , Borromini's Book, p. 93 Google Scholar; Opus,p. 17. Maruscelli's plans for all floors were made in 1627, soon modified after a report by Spada, and finally adopted in 1629 ( Connors, , Borromini and the Roman Oratory, pp. 185-92Google Scholar). The modifications included eliminating the south loggia of the second court (see Fig. 2) and reducing the height of the west and north ranges of living rooms from four floors to three by taking out the mezzanine. One reason for this was to make the courts lighter and more airy. Borromini managed to achieve the same improvement of light and air while reinstating the mezzanine, but how soon he did this is unrecorded. It was certainly earlier than 1640, the first time it is mentioned in the Decreti(see n. 24).

21 One minor design change was accepted and minuted, but never carried out: on 19 July 1638 the oratory portal to the street was to have helical Salomonic columns (Decreti, 113).

22 Sopra la volta dell’ oratorio nuovo si faccino stanze per Padri (Decreti,112: 17 July 1638).

23 Essendo desiderio di molti.

24 ‘After a paper was presented to determine the room heights of the new building, expressing the views of various parties, the last one was approved and the solution was approved by twenty-two votes to one’ (Decreti, 152: 2 June 1640). Connors, , Borromini and the Roman Oratory, pp. 45–46 Google Scholar, merely refers to ‘the decision of the patron in 1640 to add an extra storey to the building.’ But this decision is not the subject of Decreto 152.

25 See n. 20.

26 Connors, , Borromini and the Roman Oratory, p. 214 Google Scholar, distinguishing the two phases of the sheet (1636-37,1637) from a later one (see below at n. 32; see further in section ‘A complete plan’ and also n. 84).

27 Decreti, 127: 12 January 1639.

28 Ibid.; Decreti, 125: 29 December 1638; Decreti, 129: 26 January 1639. The impetus for this work came from the legacy of Father Saluzzi, who had died on 30 November 1638.

29 Connors, , Borromini and the Roman Oratory, p. 222.Google Scholar

30 Ibid., p. 220. The italics are mine.

31 See n. 74; after n. 90, and Fig. 17.

32 Connors, , Borromini and the Roman Oratory, p. 218 Google Scholar, cat. 40; pp. 214-16, cat. 39, suggesting (p. 216) that exchanging sites for refectory and kitchen areas was ‘possibly out of a desire to provide the refectory with greater privacy and the library with adequate light.’ Maybe, but he must have seen very early that neither was well sited. See also n. 26.

33 al disegno del tutto ( Downes, , Borromini's Book, p. 63 Google Scholar; Opus, p. 8 Google Scholar).

34 Alli benigni Lettori ( Downes, , Borromini's Book, p. 55 Google Scholar).

35 For a detailed discussion of these three sources of inspiration and support, see Blunt, Anthony, Borromini (London, 1979), pp. 27–51.Google Scholar

36 Quite often he does not name this triad, but his implication is clear. This number does not include every feature of the service block: many an eating establishment has foundered through bad planning behind the scenes, and the relevant chapters go into exhaustive detail.

37 It did occur, however, to Leo Steinberg — although likewise without naming Vitruvius — when writing of ‘Borromini's implication of forms in multiple functions; his solutions tend to be points of convergence for many necessities. It is a constant theme of the Opus architectonicum — the recitation of several problems and the unique form by which all are simultaneously solved’ ( Steinberg, Leo, Borromini's S. Carlo alle Quattro Fontane (New York and London, 1977), p. 363 Google Scholar).

38 Connors, , Borromini and the Roman Oratory, p. 212.Google Scholar

39 Opus, p. 5 Google Scholar.

40 Connors, , Borromini and the Roman Oratory, p. 28 Google Scholar, quoting Decreto 134: non si innovò cosa alcuna.

41 Decreti, 134:11 May 1639.

42 Opus, p. 5 Google Scholar.

43 English version here quoted from Wittkower, Rudolf, Studies in the Italian Baroque (London, 1975), pp. 153-76.Google Scholar

44 Downes, , Borromini's Book, p. 281.Google Scholar

45 Ibid., pp. 273, 284.

46 Wittkower, , Studies in the Italian Baroque, p. 162.Google Scholar

47 Parts of the volume concerning San Carlino were printed in Pollak, Die Kunsttätigkeit unter Urban VIII, reg. 225; for the prior's account of the church, see now Montijano García, Juan María, San Carlo alle Quattro Fontane di Francesco Borromini nella ‘Relatione della fabrica’ di fra Juan de San Buenaventura (Milan, 1999).Google Scholar A more recent example of the miswriting of history is in Raspe, Martin, ‘The Final Problem: Borromini's Failed Publication Project and his Suicide’, in Annali di architettura, 13 (2001), pp. 121-36Google Scholar. Filling out his story with picturesque embroidery from the unreliable Pascoli, Lione (Vite de' Pittori, Scultori ed Architetti (Rome, 1730), 1, esp. pp. 202- 03)Google Scholar Raspe writes (pp. 132-33) that on the last visit of Borromini's nephew, Bernardo, they quarrelled and he turned the nephew away, misquoting Bernardo's own account as ‘licenziò il detto nipote’ (he dismissed the said nephew) and then altered his will to prevent the latter from receiving a capital sum. But the source actually reads si licenziò il detto nipote, the verb is reflexive: the said nephew took his leave: no quarrel, no story. It was as prudent then as it is today to tie up capital left to a young person who nevertheless would benefit from the interest. Serious scholars have been advising caution in the use of Pascoli for the last 250 years.

48 See below, ‘Reactions and a third plan’.

49 23 November, 2 December 1641 (Decreti, 174, 175). See below at n. 98.

50 Connors, Borromini and the Roman Oratory, cat. 66. See again n. 98.

51 Chapter 19: Downes, , Borromini's Book, p. 105 Google Scholar; Opus,p. 21.

52 Connors, , Borromini and the Roman Oratory, p. 46.Google Scholar

53 Ibid., pp. 51, 77.

54 This story is told in all the early Neri sources as an instance of divine guidance.

55 Barbieri, Costanza et al., Santa Maria in Vallicella (Rome 1995), p. 24.Google Scholar

56 Alberti, Leon Battista, On the Art of Building in Ten Books, ed. and trans. Rykwert, Joseph et al. (Cambridge, MA, 1988), p. 156 (Book 6, chapter 2).Google Scholar

57 For the long tradition of progressive design over time, see Trachtenberg, Marvin, Building-in-Time (New Haven and London, 2010).Google Scholar This controversial book illuminates a subject often discussed but little studied. On post-Albertian architects’ (including Michelangelo's) attempts to ensure the continuity of their designs see Burns, Howard, ‘Building Against Time: Renaissance Strategies to Secure Large Churches Against Changes to their Design’, in Guillaume, Jean, ed., L'Église dans l'architecture de la Renaissance (Paris, 1995), pp. 107-25.Google Scholar

58 Decreti, 102 Google Scholar.

59 Pollak, Die Kunsttätigkeit unter Urban VIII, reg. 1802,1803,1805.

60 Connors, , Borromini and the Roman Oratory, p. 16.Google Scholar For Maruscelli's plan, see ibid., cat. 21.

61 Ibid., cat. 31-33. In the ‘Relation’ various designs were made ‘both by their own architect and by others’ ( Downes, , Borromini's Book, p. 67 Google Scholar; Opus, p. 9 Google Scholar). Cf. Saluzzi who ‘treated with this and that architect, but in vain’ (Dialogo, p. 181).

62 This is just the kind of thing Spada would enjoy. A later example is Borromini's visual metaphor of the curved five bays as the Church's welcoming arms ( Downes, , Borromini's Book, p. 75 Google Scholar; Opus, p. 11 Google Scholar). In a report to the Fabbrica of St Peter's, Bernini attributes to Alexander VII the same imagery, applied in 1657 to the oval colonnades flanking the basilica, as representing the welcome of ‘Mother Church’ to Catholics, heretics and unbelievers alike; he also made a quick sketch of the complex in the form of a human figure. Between August 1656 and May 1657, in connection with the colonnades, Alexander met Bernini several times and Spada three times — and on 4 February before a meeting with the architect. See Brauer, Heinrich and Wittkower, Rudolf, Die Zeichnungen des Gianlorenzo Bernini (Berlin, 1931), p. 70 Google Scholar and pl. 62b; Krautheimer, Richard and Jones, R.B.S., ‘The Diary of Alexander VII: Notes on Art, Artists and Buildings’, Römisches Jahrbuch für Kunstgeschichte, 15 (1975), pp. 199–233.Google Scholar Whether the idea was really the pope's or Bernini was flattering him with the ascription, Spada had known of it since at least 1647.

63 Why Maruscelli initially thickened the western bays, and also the south end of the west elevation, is a mystery, although it may have been for stability. What is beyond doubt is that, once established, these bays proved useful both structurally and visually.

64 See n. 59. There are two accounts for the removal of earth between 30 May and 20 June (Pollak, Die Kunsttätigkeit unter Urban VIII, reg. 1805); the foundations were ‘nearly finished’ to ground level on 26 July (Decreti, 104).

65 A Tad. Landi sc. 10.50 al S. Franco Castelli per misure e disegni per la fabrica di Mte Giordano (Pollak, Die Kunsttätigkeit unter Urban VIII, reg. 1787). Connors (Borromini and the Roman Oratory, cat. 38) identified one of Borromini's survey drawings, a measured plan of the south end of the site and church, dating it probably 1637.

66 Al S. Franco Castello Borromino nostro Architetto sc. 25. per la sua provisione dell'Anno presente 25.-. (Pollak, Die Kunsttätigkeit unter Urban VIII, reg.1641). Subsequent half-yearly payments (ibid., reg. 1642-56) confirm that he was paid not for eight months starting in May but for twelve starting in January.

67 A previous payment to Borromini, also of 10½ scudi on 15 December 1636, again through Landi, was for 'the chapel of St Philip in the sacristy’ (ibid., reg. 1699); the ‘chapel’ is the recess at the west end of the sacristy with a small altar and Algardi's statue of the Saint, not even complete at that date; see n. 76. Borromini made a couple of drawings for the cupboards and chests in the sacristy, probably no later than 1634 as they were not followed, and there is a design for a picture frame datable 1636 (Connors, Borromini and the Roman Oratory, cats 24(0), 24(p), 28). These imply earlier contact with Landi but with nobody else, but what is clear from the description is that these drawings had nothing to do with the Oratory hall or any larger project.

68 They are not shown on any plan of the church.

69 Connors, Borromini and the Roman Oratory, cats 23c, 23a.

70 ‘Maruscelli ignored ground levels’ (ibid., p. 85); Connors's fig. 13 reconstructing Maruscelli's south elevation diverges substantially from the information in the plan.

71 See below, n. 106.

72 Risoluto, i.e. a question was asked and answered: Decreti, 108. The complexity of steps might be a virtue in a Frank Lloyd Wright private mansion, but never at the public interface of a conventual house.

73 Downes, , Borromini's Book, p. 109 Google Scholar; Opus, p. 22 Google Scholar.

74 Vienna, Albertina, Az. Rom. 283, 36.7 x 53.7 cm. References to Albertina drawings here (all in the same class) are simply by the Arabic number. Figs 4-6 and 10 here are traced for the sake of clarity in reproduction; some minor breaks in wall planes are omitted from Fig. 6. The large-scale details of other drawings were extracted from high-definition images.

75 Downes, , Borromini's Book, pp. 69–71.Google Scholar

76 Larger sacristies usually have a small altar because sacred objects are handled there, not because mass is ever said there. Nor were masses normally said in the oratory hall. The stepped structure in the middle of the north side is neither a pulpit, as Connors identified it, nor the support for a chair, as he suggested in correspondence on the grounds that the Oratory preachers were seated. It is too small and steep, and the shaped top implies a pedestal for a standing (i.e. inanimate) figure: a statue of St Philip was eventually installed in the niche, with the pulpit opposite on the south wall.

77 Connors, , Borromini and the Roman Oratory, p. 207 Google Scholar, detected five risers; the question whether they really exist is left open in Downes, Borromini's Book. In 2011, Professor Connors kindly re-examined the original and reported: ‘It is the faintest part of the drawing. Looking closely, one can pretty clearly see two or three steps — at the entrance and at the exit into the loggia. Whether there are, or were, other steps is unclear even with close looking. Thus the drawing stands about halfway between being a witness for my reconstruction and yours.’ Two or three steps would solve no problem — and see n. 72. As built, the east corridor has six steps. The faint lines are not evenly spaced, and the simplest reading is that they are construction lines for window jambs ruled through the whole length of the wall, not part of the design. This was common practice; for early examples in Borromini, see Thelen, Heinrich, Francesco Borromini, die Handzeichnungen (Graz, 1967)Google Scholar, cats 25, 29, 47 (detail), 62, 77(bottom).

78 This is omitted from Fig. 6 to preserve the clarity of the initial concept.

79 Downes, , Borromini's Book, p. 63 Google Scholar; Opus,p. 8 Google Scholar.

80 Fatti molti disegni censurati sempre dal P. Virgilio Spada … finalmente ne feci uno a sodisfazione, di tutti col quale restorono superate tutte le dificoltà ( Downes, , Borromini's Book, p. 67 Google Scholar, Opus, p. 9 Google Scholar). Or rather, ‘to the satisfaction of all’: Opus has a misleading comma after sodisfazione absent in the manuscript, and none after tutti.

81 Dialogo,p. 181 Google Scholar.

82 Albertina, no. 285. Connors (Borromini and the Roman Oratory, cat. 39) was the first to date it rightly before no. 284, although the terms ‘first’ and ‘second’ are only part of the story.

83 No complete upper-floor plans by Borromini are known. His other careful copies are ibid., cats 37b-f (Albertina nos 278, 282, 281, 900 and 901).

84 Connors identifies three phases in this drawing: (1) the straight underlying copy from Maruscelli's plan; (2) the rest of the initial sheet; (3) another sheet pasted on later with a short-lived proposal of 1644 for a Pamphili mausoleum east of the church. The present discussion concerns only phase 2. Connors (Borromini and the Roman Oratory, cat. 39) saw no reason to doubt that this phase was all drawn in one episode.

85 See above at n. 32.

86 Connors does, however, refer in passing (ibid., p. 215) to the realignment of the sacristy windows in this drawing, but without remarking on its unfeasibility

87 Ibid., cat. 40.

88 In particular the ruled setting out of the same kitchen building as in no. 285.

89 Personal communication.

90 No conclusion can be drawn from this plan or the previous one as to Borromini's intention for the upper storey; see below. In support of his later dating, Professor Connors has argued (in correspondence, 2011) that Borromini made the freehand alterations to no. 284 in August 1638 to demonstrate to Spada that, if the latter wanted to move the library to the south range, it would have to be placed above the oratory hall, ‘vault below and vault above‘, and that this was of course impossible since the vault below was already complete in an inappropriate location. Quite apart from this is the fact that by that date Spada had accepted that such a construction was unnecessary and told the Congregation that there were no serious problems — Borromini could point to a similar library with one wall built over the crown of a vault below which had recently been completed in the Palazzo Barberini — a graphical demonstration of statics which did not require him to identify the subjects of the altarpiece and the flanking statues in niches (St Philip and St Cecilia). See also n. 102.

91 Connors, , Borromini and the Roman Oratory, p. 218.Google Scholar

92 See below at n. 112.

93 Rome, Vallicella archive, C. II. 8a, f. 12; not in Connors. The drawing was possibly produced by Maruscelli from memory after he had resigned in May /June 1637, but the early date matters more than the authorship; see Downes, , Borromini's Book, pp. 415-17Google Scholar; and Ceradini, Vittorio and Pugliano, Antonio in Architettura, Storia e Documenti, 1986/2, pp. 87–98.Google Scholar

94 Also in the Windsor drawing; see below.

95 This plurality did survive as a pipe-dream, even being engraved by Domenico Barrière around 1660 as a commission, but that image is a fantasy, with a set of never-built steps, with all the rejected skyline ornaments, and standing like part of a film set in an open field. An original impression is the British Library (K.134 g. 11, f. 33; Borromini, Opus architectonicum, ed. Connors, fig. 38). Giannini used the plate, slightly reworked, as plate V of the Opus(Fig. 19).

96 Connors, Borromini and the Roman Oratory, cat. 59. See n. 20.

97 One that does shows three bays of the second court loggia with giant pilasters. Borromini must have made this change of scale by early 1639 because the east arcade gives abutment to the refectory behind it, begun in January (ibid., cat. 57); the same change in the first court is implicit, for consistency. The left-hand pier and arch seem to descend four palmi lower; Connors read this as a ‘drastic measure’ inspired by Michelangelo's Palazzo dei Conservatori, lowering the loggia floor to oratory level. Not only drastic but the arches would be disproportionately tall — totally impractical, far from Michelangelesque and offensive to the eye. Moreover, having mounted six steps from the entrance one would have to descend again to the second court. Some elevations, like some plans, have an implied third dimension; but the piece of lowered floor is surely imagined inside the refectory, which really is at the lower level. In the middle and right-hand bays the bottom zone is occupied by plans of the giant order piers.

98 Connors, Borromini and the Roman Oratory, cat. 66 (Decreti, 174). Borromini recycled the unfinished recto, folding it three times to make the verso into what Heinrich Thelen dubbed a ‘pocket sheet’ (Taschenblatt)with memoranda on various parts of the works. One concerns an event on Friday 28 February, which must be 1642. Whether he brought it to the site meeting is an open question. This drawing in London's Victoria & Albert Museum (cat. E510-1937 - VA/128/19/4) is illustrated in Connors's catalogue but the original is now enshrined (described as a preliminary design for working out the ground plan of the Chiesa Nuova!) in a high-tech plan chest, labelled Developing, in the museum's room 128. It cannot under any circumstances be removed for inspection, so a quarter of the recto and all of the verso are hidden.

99 See Downes, , Borromini's Book, p. 83.Google Scholar The niches answer the arcades; the string courses continue the lines of mouldings on the arcades. Neither the plans in Opus nor that of Paul Letarouilly ( Édifices de Rome moderne, 1 (Paris, 1840), pl. 109 Google Scholar) place the windows accurately.

100 Payment, 2 November (Pollak, Die Kunsttätigkeit unter Urban VIII, reg. 1790).

101 Decreti, 313 Google Scholar.

102 The principal exception was the failure of the library's wall over the oratory vault; it fell to Arcucci to remove it and extend the library to the west end, which he did with discretion and skill.

103 See Downes, , Borromini's Book, p. 443.Google Scholar

104 Royal Collection, inv. 905594. For a detailed account of this drawing, see Downes, Borromini's Book, pp. 426, 431. Martin Raspe (see n. 47) sees the drawing as ‘late’ (after 1650) and made for engraving. He excludes it from ‘the planning process’ because it is not dimensioned, which is not unusual — and he entirely ignores the pentimenti which are indiscernible in his small illustration. He also finds the curve of the wall ‘much stronger than in the executed version’, an error that is not even original, but was made more than forty years ago (Paolo Portoghesi, Disegni di Francesco Borromini, exh. cat., Rome, Accademia Nazionale di S. Luca (1967), p. 16, cat. 39). The curve is identical with those seen in the Albertina drawings nos 284-86.

105 Connors (Borromini and the Roman Oratory, cat. 41), without mentioning the plinth, linked the drawing to the ‘permission’ of August 1638 to move the library, in the belief that until then the south front comprised only one giant order and an attic. In correspondence (2011) Professor Connors suggested that the bottom plinth was built but was buried in 1745 when the piazza was surfaced. However, it is not shown in the side elevation of 1725 (Borromini, Opus,pi. XXVII; Fig. 23 here) or in any of the images in his catalogue prior to that date, and careful study of prints, and photographs for over a century, shows no significant rise of the level between 1660 and the present (Fig. 24). Further, if it was to be buried Borromini would not have drawn it: it is a universal convention from the time of Raphael that an elevation starts at ground level, not below.

106 Connors (ibid., p. 33) writes of the Oratory facade that ‘It was not only smaller, but the fictive podium on which it stood was kept as low as the podium under the church, even at the cost of a general lowering of door thresholds and floor levels throughout the casa.’ This is precisely not the case with the Windsor drawing; moreover, the lower floor levels were the result of Maruscelli's ignorance of the site levels and the need to compensate for it.

107 Connors, Borromini and the Roman Oratory, fig. 14. The compound facades of S. Susanna and S. Giacomo degli Incurabili offer very poor precedents.

108 Downes, , Borromini's Book, p. 63 Google Scholar; Opus,p. 8 Google Scholar.

109 Downes, , Borromini's Book, p. 127 Google Scholar; Opus,p. 29 Google Scholar.

110 Downes, , Borromini's Book, p. 75 Google Scholar; Opus,p. 11 Google Scholar.

111 Downes, , Borromini's Book, p. 77 Google Scholar; Opus, p. 11 Google Scholar.

112 Downes, , Borromini's Book, p. 129 Google Scholar; Opus, p. 29 Google Scholar.