¹ For Dughet, see the catalogue raisonné by Boisclair, M.-N., Gaspard Dughet 1615–1675 (Paris, 1986)Google Scholar; and French, A. and Boisclair, M.-N., Gaspard Dughet called Gaspar Poussin 1615–1675. A French Landscape Painter in Seventeenth Century Rome and his Influence on British Art (exhibition catalogue, The Iveagh Bequest, Kenwood) (London, 1980)Google Scholar.

² Baldinucci, 1728, pp. 473–4 in Boisclair, Gaspard Dughet (above, n. 1), 21.

³ London, National Gallery, no. NG98. Boisclair, Gaspard Dughet (above, n. 1), 276–7, cat. 348. I am grateful to Humphrey Wine for bringing the question of the identity of the site to my attention. The identification of the site proposed in this article, based on an earlier draft, is accepted by Wine in his entry on this painting in Wine, H., The Seventeenth Century French Paintings (National Gallery Catalogues) (London, 2001), 156–9Google Scholar.

⁴ For the provenance, see Wine, The Seventeenth Century French Paintings (above, n. 3), 154.

⁵ London, National Gallery, no. NG68. Boisclair, Gaspard Dughet (above, n. 1), 276, cat. 347. The two works are not easy to identify in the Corsini inventories; see Wine, The Seventeenth Century French Paintings (above, n. 3), 154.

⁶ It was identified as a view of ‘L'Arricia, near Rome’ when it was offered with its pendant for sale from the Ottley collection in London, 24–5 January 1801, lots 32 (NG68) and 33 (NG98), as ‘Gaspar Poussin’. They were apparently not sold on this occasion and were offered again at Christie's, London, 16 May 1801, lot 20, as ‘Gaspar Poussin’ ‘Landscape with Figures — the Entrance to L'Arriccia, near Rome’ (NG98) and lot 21 ‘The Companion to the Former, & Woody Scene, a Shepherd leading his Flock; — out of the Picture — striking’ (NG68), when it was bought in. Wine has suggested that this pair might be identifiable with a pair in the sale of Sir William Young, London, Peter Coxe, 20 May 1802, lot 87, bought by Sir Abraham Hume. If these are indeed the same works it may be significant that each is identified as being a ‘View of Tivoli’. However, it seems odd that the identification should shift from Albano to Tivoli and back again in so short a space of time. In the sale of Sir Scrope Bernard Morland, London, Christie's, 25 April 1818, lot 125, they were titled ‘A View of Larici — a clear and beautiful chef d'oeuvre, in his finest manner’ and ‘A Road through a Wood, with a flock of Sheep’. They were bought by the Reverend Holwell Carr, and were bequeathed to the National Gallery in 1831 when NG98 was called ‘Larici’. An engraving of NG98 by Alfred Smith made at this time identifies it as ‘A View of La Riccia’. See Wine, The Seventeenth Century French Paintings (above, n. 3), 154–5, 158.

The identification of the pendant (NG68) as a ‘Road near Albano’ is a more recent tradition than the identification of NG98 as a view of Ariccia. Wine noted that the earliest printed reference seems to have been in The National Gallery of Pictures by the Great Masters published in 1838(?), when the anonymous author (probably John Landseer) wrote that Holwell Carr had vinscribed beneath it ‘A View near Albano’ and that, ‘having resided for some considerable time in Italy, Mr Carr probably knew the very spot which is here represented’. It is impossible to confirm or refute this identification on topographical grounds, but in view of the arguments presented here about its pendant, this identification should be treated with reserve. Wine rejected the identification with Albano, noting that Ottley, who had acquired it in Rome, made no reference to Albano, and accordingly titled it Landscape with a Shepherd and his Flock, following Ottley (Wine, The Seventeenth Century French Paintings (above, n. 3), 152). One wonders whether the title might not have been suggested by a desire to give it a title with a certain topographical symmetry with its pendant (Albano and Ariccia are adjacent towns). Boisclair referred to another painting attributed to Dughet identified as being a view in the region of Albano, in the J.J.B. Owen sale, Christie's, London, 1868, with a provenance to the Palazzo Falconieri that she had not located (Boisclair, Gaspard Dughet (above, n. 1), 276; Redford, G., Art Sales. A History of Sale of Pictures and other Works of Art, 2 vols (London, 1888), II, 280Google Scholar).

⁷ In the earlier catalogue of French paintings in the National Gallery Davies titled it simply ‘Ariccia’ noting that ‘the ancient Aricia stood in the hollow at the right; the modern town is on a hill to the left’. Davies, M., National Gallery Catalogues, French School (London, 1957), 86–7, no. 98Google Scholar.

⁸ For example, in a View of Ariccia by Alexandre Desgoffe exhibited in The Lure of Rome (exhibition catalogue, Hazlitt, Gooden and Fox, 31 October–27 November 1979) (London, 1979)Google Scholar, cat. 63, we see a roughly similar arrangement of forms: the town, seen from the direction of Albano, is on a hill to the left of centre, with the valley in the middleground curving around behind with further hills to the right and beyond. But the Chigi palace and the dome of Santa Maria dell'Assunzione are clearly identifiable.

⁹ Boisclair, Gaspard Dughet (above, n. 1), 277 wrote: ‘L. Salerno indique que le no 98 ne représente pas Ariccia, mais n'en identifie ni la localité ni le château’. Salerno, L., Pittori di paesaggio del Seicento a Roma, 3 vols (Rome, 1977–1980), II, 526Google Scholar, called the subject of this and the Hazlitt gallery variant a ‘village’ (villaggio), which by implication is not Ariccia, normally considered to be a town, although the plates are captioned ‘View of Ariccia’. However footnote 21, which refers to the ex-Hazlitt Gallery variant fig. 86.10 (Boisclair, Gaspard Dughet (above, n. 1), cat. 380), is missing or incomplete.

¹⁰ For example, Boisclair, Gaspard Dughet (above, n. 1), cat. 348.

¹¹ For example Boisclair, Gaspard Dughet (above, n. 1), cat. 186 (Ashmolean Museum, Oxford); possibly cat. 227 (whereabouts unknown), but very freely treated; cat. 281 (Florence, Palazzo Pitti, Galleria Palatina) (the captions of the plates are wrong, so that it appears as fig. 319, cat. 280); cat. 380 (Rome, private collection); cat. 411 (whereabouts unknown).

¹² Braunschweig, Herzog-Anton-Ulrich Museum, inv. 550; Andrews, K., Adam Elsheimer. Paintings–Drawings–Prints (London, 1977), cat. 18, 148–9Google Scholar.

¹³ Röthlisberger, M., Claude Lorrain, the Drawings, 2 vols (Berkeley/Los Angeles, 1968), 189Google Scholar, cat. 431.

¹⁴ Lorrain, Claude, View of the Porta S. Angelo, Tivoli, from the Temple of the Sibyl, British Museum, no. Oo. 6–80Google Scholar, H. 8 21.4 × 31.4 cm. Pen, brown wash, frame line. R 429 pp. 188–9. The first extant page from the Tivoli book. Röthlisberger dated it c. 1640, but Whiteley, J.J.L., Claude Lorrain. Drawings from the Collections of the British Museum and the Ashmolean Museum (London, 1998), 130–1Google Scholar, cat. 74, argued for a later date in the 1650s.

¹⁵ Röthlisberger, Claude Lorrain, the Drawings (above, n. 13), 188–9, cat. 429.

¹⁶ Baltimore, Walters Art Museum, inv. 37.1175. Briganti, G., Gaspar van Wittel (second edition) (Milan, 1996), 220Google Scholar, cat. 244. See also the drawing of the area in Naples, Museo Nazionale di San Martino, inv. 1136/2 (Briganti, cat. D. 242 and related painting cat. 248; E.P. Bowron and Rishel, J.J., Art in Rome in the Eighteenth Century (London/Philadelphia, 2000), 558–9, cat. 407Google Scholar).

¹⁷ In Dughet's own views towards the head of the valley, the gateway motif is rarely visible, except perhaps in Boisclair, Gaspard Dughet (above, n. 1), cat. 414 (whereabouts unknown). In the painting in Puerto Rico (Musée Ponce; in Boisclair, Gaspard Dughet (above, n. 1), cat. 203, fig. 245) he shows the roadway, the rocky outcrop and the small cascade with behind it a cubical tower-like building that is perhaps a conflation of the gateway and the building with the re-entrant angle. Cf. also Boisclair, Gaspard Dughet (above, n. 1), cat. 389 (Rome, Galleria Nazionale d'Arte Antica).

¹⁸ Honing, Adriaen, View of Tivoli from the Weir of the Cascata Grande, towards the Temple of the Sibyl and Monte Catillo. London, Witt CollectionGoogle Scholar. On Honing, see Röthlisberger, M., ‘Adriaen Honing alias Lossenbruy’, Münchner Jahrbuch der Bildenden Kunst 14 (1963), 121–8Google Scholar, figs 3–5, 8–10, and Röthlisberger, M., ‘À la recherche d'Adriaen Honing’, National Gallery of Canada Bulletin 4 (1964), 14–21Google Scholar. Honing (Dordrecht 1644–Rome 1683) was established in Rome from 1667 to 1683, and his one dated drawing, dated 1673, is a view of Tivoli (Adriaen Honing, Landscape at Tivoli, Cabinet Graphique, University of Leyden, Weleker collection, no. 545), and his other Tivoli drawings may date from about the same time.

¹⁹ Cf. the letters quoted in Marocco, G., Monumenti dello Stato Pontificio e relazione topografica di ogni paese XII (Rome, 1836), 36, 41Google Scholar etc.

²⁰ Massimo, F., Relazione storica del traforo nel Monte Catillo in Tivoli per l'inalveazione del fiume Aniene compilata da Monsignor Don Francesco Massimo, 2 vols (Rome, 1838)Google Scholar.

²¹ Marocco, Monumenti dello Stato Pontificio (above, n. 19), 27. Carlo Fea, on the other hand, identified the remains of a weir upstream of the pre-1826 weir that he considered to be ancient. Fea, C., Considerazioni storiche, fisiche, geologiche, idrauliche, architettoniche, economiche, critiche (Rome, 1828)Google Scholar, ‘Considerazioni storiche, fisiche, geologiche, idrauliche, architettoniche, economiche, critiche sul disastro accaduto in Tivoli il dì 16. Novembre 1826’, 1–126, see p. 37.

²² A letter of Cardinal Cesi of 29 October 1597 listing tax rates for water usage (Fea, Considerazioni storiche (above, n. 21), 70) refers to the following industries using the water: molini ad oglio (olive oil mills), cartiere (paper mills), valchiere,ferriere (iron foundries), ramiere (copper foundries), polviere (powder magazines, or probably mills; that is, salnitriere), and concie di corami (tanneries), although the Villa d'Este was the major consumer. He reproduced (tav. I, C) a plan of the river upstream of the weir based on a plan accompanying an expertise by Matti a de’ Rossi of 1680 showing the various channels leading off the left bank. See also Massimo, Relazione storica (above, n. 20), II, tav. X, ‘Piant a generale della città di Tivoli’.

²³ Marocco, Monumenti dello Stato Pontificio (above, n. 19), 28–9.

²⁴ Fea, Considerazioni storiche (above, n. 21), 21–2 was scathing about the quality of the construction of the weir.

²⁵ Marocco, Monumenti dello Stato Pontificio (above, n. 19), 32. Fea, Considerazioni storiche (above, n. 21), 19, gave a detailed account of an antique channel beneath the canale della stipa, which he believed continued to function until 1578, when it was abandoned in favour of a new channel built above it, referred to in 1578 as ‘aquedotto sotto Porta Sant'Angelo’. In 1589 there was a reference to repairs to ‘il condotto sotto la strada Cornuta’. He published the contracts and other documents for the work of 1592 (63–5).

²⁶ Bernini's name is associated with the canale della stipa because it was repaired in 1680–2 under the direction of Monsignor Pietro Bernini, Gianlorenzo Bernini's son, who was secretary of the responsible body, the Sacra Congregazione delle Acque. Pietro Bernini contacted his father who, being too old to deal with it himself, referred the matter to Mattia de' Rossi, who had a reputation for hydraulic work. De' Rossi's work is extensively documented: Fea, Considerazioni storiche (above, n. 21), 66–9. There was also a tradition that Luigi Bernini, Bernini's brother, was involved, and Fea (p. 65) cited documents referring to a visit by ‘il sig. Luigi Bernini architetto celebre’ in 1669.

²⁷ The shrine is also visible in a well-known drawing by Claude Lorrain in Madrid (Röthlisberger, Claude Lorrain, the Drawings (above, n. 13), cat. 489), which shows a similar view but from further upstream and which does not extend as far as the gateway.

²⁸ Fea, Considerazioni storiche (above, n. 21), tav. III, D, shows the broken wall.

²⁹ Massimo, Relazione storica (above, n. 20), tav. V: ‘Pianta d'un tratto del Fiume Aniene entro la Città di Tivoli da servire per dimostrazione dei nuovi progetti presentati dalla Commissione speciale nominata con Dispacci dell E∼mo Cardinale Prefetto della S. Congregazione del B. Governo dei 31 Luglio, e 15 Agosto 1829’. Fea, Considerazioni storiche (above, n. 21), tav. II, also provided a useful plan of this area.

³⁰ Fea, Considerazioni storiche (above, n. 21), tav. II, caption no. 33.

³¹ Nibby, A., Viaggio antiquario ne' contorni di Roma (Rome, 1819), I, 165Google Scholar.

³² For example in a painting by Pierre-Athanase Chauvin. The Falls of Tivoli with the Temple of the Sibyl, oil on canvas, 96.3 × 74 cm. Exhibited in The Lure of Rome (above, n. 8), cat. 13.

³³ Marocco, Monumenti dello Stato Pontificio (above, n. 19), 7–130, ‘Cronaca delle diverse vicende del fiume Aniene in Tivoli fino alia deviazione de medesimo me' cunicoli del monte Catillo’; esp. p. 16.

³⁴ Marocco, Monumenti dello Stato Pontificio (above, n. 20), 36–8. Fea, Considerazioni storiche (above, n. 21), 66, published the documents referring to the visit of the three experts (nos. 10–12). One (no. 10, p. 65) is a permesso to clean the channels: ‘ … et pro expurgatione alveorum veterum, nempe della stipa, et alterius nuncupat. del CHIAVICONE sotto le salnitrerie; … ’. Another (no. 11, p. 66) refers to the visit on 5 April 1671 of the three experts:‘ … quali avendo visto tutto quello si poteva vedere, lodarono l'opera, che si faceva nel CHIAVICONE, e vollero, che in un medesimo tempo si dasse principio ad espurgare l'alveo della stipa; credendo, e persuadendosi detti sig. architetti, che per di là si potesse divertire buona parte delle acque del fiume Teverone; e così anco parte per li acquedotti, acciò l'edificij potessero lavorare comodament e senza servirsi del CHIAVICONE; quale ricerca grossissima spesa …’. While these documents make it clear that the chiavicone and the stipa were two separate channels, the document s associated with the work by de' Rossi in 1680–2 (pp. 66–71) seem to use the terms ‘chiavicone’ and ‘stipa’ interchangeably for the canale della stipa. On Fea at Tivoli, see R.T. Ridley, The Pope's Archaeologist. The Life and Times of Carlo Fea (Rome, 2000), 297–302.

³⁵ Fea, Considerazioni storiche (above, n. 21), in his tav. II, I, indicated with dotted lines two channels in this area, one that he identified as the canale di Vopisco, an antique channel feeding the villa of Manlius Vopiscus (no. 39 in the captions to this plate on p. 36), and the other nearby as being modern (no. 40). (The plate is reproduced in Ridley, The Pope's Archaeologist (above, n. 34), 298 fig. 7.) The former begins near some Roman remains slightly upstream of the weir and runs beneath the salnitriera; the modern channel is not shown throughout its length but likewise runs under the salnitriera and comes out in a similar place close to the rocky outcrop. Fea seems to identify the chiavicone with part of the canale di Vopisco: in a note to the document (no. 11) cited in n. 34 above he explained the chiavicone as being ‘Nel canale di Vopisco dalla parte di sotto verso la caduta, che si è trovato qualche poco sterrato’. No doubt the chiavicone supplied water to the salnitriera, which manufactured saltpetre, the main ingredient of gunpowder. Encyclopedia Britannica (11th edition) (1911), XXII, 93, s.v. ‘saltpetre’. Gunpowder was produced in the polveriera, which was located nearby outside the Porta Sant'Angelo from sometime between 1597 and 1602, until an explosion in 1663 led to it being abandoned and rebuilt on another site outside the Porta del Colle, where it was operational from 1675. See C.P. Scavizzi, ‘Sulla polveriera di Tivoli tra XVI e XIX secolo’, Rivista Storica del Lazio 5 (7) (1997), 3–32; and C.F. Giuliani, ‘Un progetto inedito per la polveriera di Tivoli: il Ritiro di S. Ignazio e gli Orti Teobaldi a Porta Scura’, Atti e Memorie della Società Tiburtina di Storia d'Arte 74 (2001), 7–24.

³⁶ See, for example, attributed to Jan Asselijn, View of Tivoli with the Villa of Maecenas, 40.1 × 50.5 cm. Kassel, Staatliche Kunstsammlungen, inv. nr. GS 5110. Oehler, L., Niederländische Zeichnungen des 16. bis 18. Jahrhunderts (Fridingen, 1979)Google Scholar, cat. 9/10, p. 32. Photo: Museum, neg. G3 004.