View of Duurstede Castle from the North-west

In 1646 and 1647, Roelant Roghman drew a large series of views of Dutch castles. It was an ambitious project for a twenty-year-old, consisting of circa two hundred and fifty drawings. Some two hundred and twenty drawings from the series are known today. Of these, the Rijksprentenkabinet owns forty-nine sheets. Similar large holdings are kept in the Teylers Museum, Haarlem, and in a private collection. The rest is widely dispersed. The whole series was extensively studied by Henri van der Wyck, J.W. Niemeijer and Wouter Kloek in their two-volume publication (1989-90).¹⁰

Roghman depicted about a hundred and fifty different castles in the provinces of Holland, Utrecht and Gelderland, many of them from more than one vantage. The drawings served as ‘castle portraits’, impressive views of intact buildings or ruins, often on large sheets of paper and sometimes also on more than one sheet joined together. Besides close-up renderings, there are distant views, showing the buildings hidden among trees or giving an overview of a castle next to an adjacent village. There are even panoramic views, such as inv. no. RP-T-1898-A-3658 (Schagen), for which the artist climbed a nearby church tower.

The idea behind the series was to illustrate places of historical value, mostly castles acknowledged at one time or another as a ridderhofstad.¹¹ This might explain why Roghman drew not only inhabited, intact buildings, but also ruins. The resulting series is an accurate reflection of the state of historic sites at the time,¹² some of which were no longer extant;¹³ the drawings have occasionally even been used as models for restoration of surviving structures. The series – unique in scope and size, and masterly and surprisingly mature in its draughtsmanship – is unrivalled in the history of castle drawing and a highlight of Dutch seventeenth-century art.

Paper, watermarks, technique, signatures
In all likelihood, Roghman drew the castles in situ.¹⁴ He started each sketch with black chalk or graphite. As can be seen by discarded sketches, such as that on the verso of inv. no. RP-T-1898-A-3716 (Oosterwijk), he began by establishing a base line for the architecture by means of a horizontal line that separated the foreground from the middle ground.

The drawings were executed on several different sizes of paper. There are large, rather coarse sheets of circa 450 x 600 mm, with a watermark of a bunch of grapes below a fleur-de-lis.¹⁵ Roghman probably used these sheets during his first campaign in 1646.¹⁶ These large sheets were folded vertically at the centre, which sometimes hindered the consistent application of ink, as can be seen in inv. no. RP-T-1879-A-80. Smaller (unfolded) sheets were made by cutting larger sheets of circa 500 x 700 mm in half. These bear a watermark of a fleur-de-lis surmounted by a crown.¹⁷ Some of the Rijksprentenkabinet’s castle drawings are done on paper with different watermarks,¹⁸ and one castle drawing (Oud Heusden; inv. no. RP-T-1900-A-4344) is executed on blue paper.

Roghman clearly had a command of the rules of both one- and two-point perspective, apparently worked without perspectival aids and dispensed with a vanishing point and formal orthogonal lines.¹⁹ There are instances of abandoned versions, the presence of pentimenti and redrawn motifs that did not fit on the sheet – all of which reveal his working method (e.g. inv. no. RP-T-1888-A-1774). There is at least one case, the View of Wulven bij Jutfaas in the Teylers Museum, Haarlem (inv. no. O++ 021),²⁰ which is a second, amended version of a drawing with corrections in the Rijksmuseum (inv. no. RP-T-1899-A-4118). Generally, these minor corrections were covered by wash, with strong effects of light and shade. Roghman usually used grey wash, which he apparently added while still on the spot.²¹ Sometimes, his washes have a blueish tint, and there are also drawings with brown wash. In one case, he used charcoal soaked in linseed oil to apply accents, as can be seen by the oily residue visible on the verso of the Rijksmuseum’s view of Swieten (inv. no. RP-T-1879-A-80).

After finishing the drawing, Roghman often inscribed it in black chalk with his monogram, sometimes a date, and sometimes also a crossbar with a north arrow, documenting the castle’s orientation.²² Later, Roghman went over his monograms in brown ink, transforming them into full signatures,²³ and often also adding a date. These dates, 1646 and 1647, appear to be trustworthy, the series being done over a relatively short period of time.²⁴ Evidently, Roghman drew the castles in two different campaigns, interrupted by a pause during the winter months. An itinerary reconstructed by Kloek assumes that the artist started in 1646 in Leiden,²⁵ making excursions into its surroundings before visiting Utrecht, the Vecht region and the Sticht. In 1647, Roghman set out on a large round trip from Amsterdam to Medemblik, Schagen, Kennemerland, Rijnland, Delfland, then from Rotterdam to Dordrecht and its surroundings, to Rhoon, Geervliet, Oostvoorne, visiting the great river regions with Heusden, Altena, Buren, Duurstede, Amerongen and Utrecht, and then along the Vecht back to Amsterdam. This itinerary meant that Roghman apparently visited many castles twice.²⁶ Differences in the application of wash, and the intensity of the shades of grey, may be explained by the risks of working out-of-doors, with impending darkness or showers causing breaks in drawing sessions. In all likelihood, the seventeenth-century inscriptions in black chalk classified by Niemeijer as ‘Hand A’ were written by Roghman himself.²⁷ It is logical that the artist would record the name of the location in the same material as he used for the drawing.

History of the series, commission and early provenance
The first known reference to the series is found in 1708, in an advertisement for the sale of the library of Hillebrand Bentes the younger (1677-1708) – the first certain owner of the series. It is described as comprising ‘some 250 drawings’ of castles in the provinces of Holland and Utrecht.²⁸ Probably on the occasion of that sale, an inventory of the convolute was written by the Rotterdam antiquary Cornelis van Alkemade (1654-1737). The manuscript ‘Register van Adelijke Huijsen, Kastelen &c. gelegen in Holland, Uitregt, Gelderland &c. alle getekent door Roeland Roghman meest in de jaaren 1646 & 1647’, generally referred to as the ‘List Bentes’, is preserved in the Rijksmuseum (inv. no. MS 3684).²⁹ Though the list is in an arbitrary order, it provides valuable information on the series, which, at that time, comprised two hundred and forty-five castle drawings.³⁰ Hillebrand Bentes the younger could have inherited the drawings from his grandfather Hillebrand Bentes the elder (1591-1652), who was a contemporary of Roghman. Dudok van Heel proposed that the grandfather might actually have commissioned the series.³¹ There was also a connection between the Bentes family and the printer and publisher Joan Blaeu (1596-1673), who acted as guardian to nine-year-old Albert Bentes (1643-1701) after the death of his father, Hillebrand the elder. Blaeu, who had previously employed Roelant’s father, the engraver Hendrick Lambertsz Roghman (1602-1647/57),³² may well have chosen the young Roelant for this prestigious commission.

Other potential patrons have been suggested, including Cornelis Bicker van Swieten (1592-1654), whose castle was depicted four times (one of which may have been used as a title-page for the series) and who was related by marriage to the Bentes family;³³ Cornelis van Aerssen van Sommelsdijck (1600-62), the immensely rich owner of Spijk, which was represented in the series five times (more than any other castle);³⁴ and Adriaan Pauw (1585-1653), owner of Heemstede, of which there are three views.³⁵ Another previously proposed candidate, Laurens Baeck (1570-1642),³⁶ maternal grandfather of Hillebrand Bentes the younger, can be excluded since he died before Roghman started the series.

The next mention of the castle drawings is found in the Schatkamer der Nederlandsche oudheden (1711) by Ludolf Smids (1649-1720). Twenty-seven of its sixty engraved illustrations by Jacobus Schijnvoet (1685?-1733) were after castle drawings by Roghman. Though Smids’ information on individual drawings tends to be imprecise and sometimes incorrect, it contains further information about the provenance of the series. According to the foreword, the drawings were then in the collection of Christiaan van Hoek (1643-1715),³⁷ owner of the country estate Ouderhoek (near Loenen),³⁸ and the ‘friend and maecenas’ to whom Smids dedicated the work. Van Hoek must have bought the whole castle series at the Bentes sale of 1708,³⁹ in order to make it available for Smids’ studies.⁴⁰ The series next passed to Christiaan’s son Anthony van Hoek (1674-1719), who inherited Ouderhoek by 1702.⁴¹ In a poem by Claas Bruijn (1671-1732), written there on 16 August 1717, Anthony van Hoek is praised as the owner of the castle drawings (‘[…] of konst in Rogman op wilt zoeken, een schat by u op `t schoonst´ bewaart’).⁴² Probably while in the collection of one of the Van Hoeks, the drawings were given an engraved label that was pasted at upper centre, just below the border.⁴³ These labels not only bore the name of the relevant castle but also a number, both written in pen. The twenty surviving labels – including on a pair of drawings in the Städel Museum, Frankfurt-am-Main (inv. nos. 2839 Z [Loenersloot] and 2836 Z [Sterkenburg])⁴⁴ – suggest that this numbering system started with castles from the Utrecht region, continuing to Zuid-Holland and ending with Noord-Holland.⁴⁵ As in most other public collections, these labels, and thus the earlier ink numbering system, have been removed for conservation reasons from the drawings in the Rijksmuseum (though the earlier presence of such a label is evident on inv. no. RP-T-1879-A-81, where a paper repair replaces the excised label).

In 1718, the series was mentioned by Houbraken, without reference to its then owner (which could have been the Bentes heirs or one of the Van Hoeks); according to his description, they were kept in an album.⁴⁶ Another eighteenth-century source is the five-page handwritten list by Abraham de Haen (1707-1748) bound into a copy of Smids’ Schatkamer, now also kept in the Rijksmuseum (inv. no. 1896-4047).⁴⁷ This ‘Register der Teekeningen, Casteelen Adelijke huijzen etc.a in de provincie van Holland & Utrecht alle naar ’t Leven getekend door Roeland Rochman in den Jaare 1646, 1647-etc.’ is known as the ‘List De Haen’. It is not clear whether De Haen had access to Roghman’s drawings or if he simply copied another written source (such as the List Bentes).⁴⁸ However, here the castles’ names are in alphabetical order. In some cases, De Haen specified local references, alluded to then recent literature and sometimes also to the condition of the respective building – twenty-two castles are described as in ‘ruin’.

Anthony van Hoek never married and left Ouderhoek and the drawings to his cousin Jan [or Jean] de Wolff (1681-1735).⁴⁹ Circa 1730, Louis-Philippe Serrurier (1706-1751), an amateur draughtsman and acquaintance of De Wolff who lived nearby across the River Vecht, copied a number of castle drawings by Roghman. He referred to these as ‘Rogman van de Wolf’.⁵⁰ Whether they remained at Ouderhoek is unknown.⁵¹ They next turn up at the sale in 1800 of the collection of Cornelis Ploos van Amstel (1726-1798). There is no information as to when, where or from whom Ploos purchased the series.⁵² One and a half years after his death (20 December 1798), at the sale of his huge collection of drawings on 3 March 1800 sqq., two hundred and forty-one castle drawings by Roghman were offered in Albums KK 1-6.⁵³ The convolute that was listed in random order in the List Bentes was then systematically arranged according to regional location. The first album held forty-five drawings from the Utrecht area; the second album forty-five drawings of castles in the Sticht; the third album included thirty-five castle drawings from Rhijnland, Delfsland, Schieland, etc.; Albums 4 and 5 contained thirty-five and thirty-nine unspecified drawings (‘Huizen enz., als boven’); and the last, Album KK 6, contained forty-three castle drawings from Noordholland and Kennemerland.⁵⁴ Each album also included a title-page. In the first album, there were two such title-pages, plus a portrait drawing of Roghman by Jan Stolker (1724-1785), which is preserved in the Special Collections, Universiteit Leiden (inv. no. PK-T-AW-1415).⁵⁵ In that portrait, Roghman is portrayed seated among portfolios of drawings of convents and holding up a view of an Italianate landscape with ruins, thus it was probably not made for the castle series. Nevertheless, if Stolker made the portrait on behalf of Ploos, as he did in other cases,⁵⁶ Ploos must have acquired the series before Stolker’s death in 1785.

At Ploos van Amstel’s sale, the series went en bloc for fl. 2,000 to Cornelis Sebille Roos (1754-1820),⁵⁷ who apparently eventually dismembered the albums. Only six weeks after the Ploos sale, on 16 April 1800 sqq., some eighty of Roghman’s castle drawings were included in the sale of Jacob Cats (1741-1799), Steven Goblé (1749-1799) and others; in the sale catalogue, these lots are annotated with the letter ‘P’ as the consigner, which stood for Roos who apparently took the opportunity to bring a part of the freshly acquired series onto the market.⁵⁸ This convolute, offered in Albums A and B, was the largest intact portion of the castle series after its dismemberment.⁵⁹ Thanks to the often explicit titles, it was possible to identify many of the castle drawings in later sales, though one must be wary of instances in which the same castle was the subject of multiple drawings.

Verso inscriptions and later provenance
The drawings are extensively annotated on the versos. Of these inscriptions, only the earlier ones were systematically studied by Niemeijer, who discerned five different hands:⁶⁰ ‘Hand A’ and ‘Hand B’ of seventeenth-century origin (the first likely to be the artist himself); ‘Hand C’ of eighteenth-century origin; ‘Hand D’ of eighteenth- or nineteenth-century origin; and ‘Hand E’ of nineteenth-century origin. The post-seventeenth-century inscriptions are thought to transcribe or replace earlier inscriptions that had become illegible or were lost by trimming.⁶¹ ‘Hand B’ might be associated with one of the Bentes owners, and ‘Hand C’ could be that of either Christiaan or Anthony van Hoek.⁶² ‘Hand D’ and ‘Hand E’ for the moment remain unidentified.⁶³

Niemeijer did not assess later annotations on the castle drawings’ versos, such as numbers and price codes. Some of these can now be deciphered, providing valuable additional provenance information and a potential identification for ‘Hand E’ as the dealer Jan Coenrad Pruyssenaar (1748-1814). For instance, graphite inscriptions such as ‘Lr A N 11&12. 2 p ° = 150 – 170’, which appears on the verso of the Rijksmuseum’s view of Abcoude (inv. no. RP-T-1911-69), refers to two items in the 1800 Cats sale, where in Album A two drawings of Abcoude are listed as no. 11 (‘het Slot te Abcoude’) and no. 12 (‘het zelfde van een andere zijde’); these were bought together by the dealer Pruyssenaar for fl. 15.⁶⁴ The inscription can thus be decoded as ‘L[ette]r A, N[os.] 11&12. 2 pieces = 15:0:- guilders – 17:0:- guilders (with commission)’. The other item from that pair of views of Abcoude, now in the Frits Lugt Collection, Fondation Custodia, Paris (inv. no. 5108),⁶⁵ bears a near identical code, which seems to begin with an allusion to the artist: ‘R[…] Lr A. No 11 & 12. 2 p’ 150-170’.⁶⁶ A similar code is found on the verso of inv. no. RP-T-1900-A-4342 (Broekhuizen), ‘L.r A n°. 9.10 ƒ.520.100[?] 160’, which can also be linked to one of Pruyssenaar’s purchases at the Cats sale: in Album A (‘Letter A’), no. 9, a drawing of Broekhuizen is mentioned that went for fl. 50 to the dealer, together with no. 10, the Rijksmusem’s view of Groenewoude (inv. no. RP-T-1899-A-4104), which bears only a fragmentary annotation alluding to the same sale and price (9 ƒ 5). The last example of this type of code among the Rijksprentenkabinet’s castle drawings by Roghman is on inv. no. RP-T-1900-A-4340 (Vliet), which featured in Album A, no. 26 in the Cats sale and went to Pruyssenaar for fl. 5:15:-. Its inscription, partially effaced, reads, ‘A/ […] 26. . 60 . . -.[...] g’. Although such codes do not appear on all drawings that passed through the hands of Pruyssenaar (and might at some point have been erased or trimmed),⁶⁷ they feature exclusively on drawings inscribed by ‘Hand E’.⁶⁸ According to Kloek, ‘Hand E’ (thus Pruyssenaar?) was also reponsible for what was previously described as an ‘atypical’ signature on inv. no. RP-T-1891-A-2467 (Huis ter Does), ‘R. Roghman jnvent’, which probably replaced a lost or trimmed original signature.⁶⁹

Another type of inscription frequently seen on the Rijksmuseum’s castle drawings is an ‘N’ in graphite followed by a number, apparently written by the same early nineteenth-century hand on inv. nos. RP-T-1888-A-1772 (N 4), RP-T-1888-A-1780) (N 4), RP-T-1898-A-3661 (N 4), RP-T-1888-A-1769 (N 5), RP-T-1891-A-2419 (N 7) and RP-T-1898-A-3716 (N 11).⁷⁰

Another numbering system that is closely related to the Pruyssenaar code, if not by the same hand, consists of an ‘N’ with a double underlined superscript ‘o’, followed by a number and occasionally the letter ‘Z’.⁷¹ Still another type is the pen-and-ink numbering system found on inv. nos. RP-T-1899-A-4122 (Amerongen: n°. 106), RP-T-1899-A-4087 (Broekhuizen: n°105) and RP-T-1899-A-4104 (Groenewoude: n° 100).⁷² These are all from the collection of Cornelis Laurens de Leur (1850-1916) of Utrecht, which was apparently devoted mainly to castle drawings, but it probably predates his ownership, since the same system does not appear to have been applied to other drawings in the museum with the same provenance.

Annemarie Stefes, 2018

References
S.J. Fockema Andreae et al., Kastelen, ridderhofsteden en buitenplaatsen in Rijnland, Leiden 1952, pp. 56-57; T. Laurentius et al., Cornelis Ploos van Amstel (1726-1798): Kunstverzamelaar en prentuitgever, Assen 1980, pp. 16, 21; B. Broos, Rembrandt en tekenaars uit zijn omgeving, coll. cat. Amsterdam 1981 (Oude tekeningen in het bezit van de gemeentemusea van Amsterdam, waaronder de collectie Fodor, vol. 3), pp. 179-80, under no. 52; W. Sumowski, Drawings of the Rembrandt School, 10 vols., New York 1979-92, X (1992), pp. 4990-91; H.W.M. van der Wyck and J.W. Niemeijer, De kasteeltekeningen van Roelant Roghman I, Alphen aan den Rijn 1989, pp. VII-XV, 1-18; W. Kloek and J.W. Niemeijer, De kasteeltekeningen van Roelant Roghman II, Alphen aan den Rijn 1990, pp. 113-34; B. Olde Meierink, Kastelen en ridderhofsteden in Utrecht, Utrecht 1995, p. 82; H.L. Janssen et al. (eds.), 1000 jaar kastelen in Nederland: Functie en vorm door de eeuwen heen, Utrecht 1996, pp. 122, 142; E. Munning Schmidt and A.J.A.M. Lisman, Plaatsen aan de Vecht en de Angstel: Historische beschrijvingen en afbeeldingen van kastelen, buitenplaatsen, stads- en dorpsgezichten aan de Vecht en de Angstel – van Zuilen tot Muiden, Alphen 1997, p. 132; W. Beelaerts van Blokland and C. Dumas, De kasteeltekeningen van Abraham Rademaker, Zwolle 2006, pp. 53, 58, 64-66, 81; W. Beelaerts van Blokland and C. Dumas, De kasteeltekeningen van Abraham Rademaker. Aanvullingen en correcties, The Hague 2016, pp. 6, 35