Two Cocus Cabinets – Hidden in Plain Sight
A little while ago I found in a provincial saleroom a cabinet which has not only led to a fascinating restoration project but which will also play a part in research into the origins of English veneered cabinetwork.
Illustrated in Adam Bowett’s invaluable book on late 17th and early 18th century furniture is a beautiful cocuswood cabinet-on-stand (1), which may be dated to around the 1660’s – the period which has generally been taken as marking the beginning of veneered work in England, following the return of Charles II and his court from exile on the Continent.
The cabinet in Dr Bowett’s book, which is held by the National Trust in Powis Castle, may be considered amongst the first of its kind. The cabinet which I was lucky enough to spot is its sister-piece.
This cabinet had very clearly “had a life”. As well as lacking a stand below the base drawer, the pine fronts of the external drawers had been patched where the locks had been removed, with pieces of mahogany veneer inserted. The internal drawers had been fitted with reeded knobs and the whole piece had been liberally ebonised. There were the inevitable shrinkage cracks and splits, veneer patches and tears and some missing pieces of moulding.
From the stand drawer upwards, however, the cabinet was intact in all its essentials.
It is identical in form to the Powis Castle cabinet, to the extent that the case for a common maker is compelling (2). That opened up the fascinating opportunity – perhaps unique for a piece of this period – to restore the cabinet with a high degree of certainty as to the intentions of the original maker.
The Powis Cabinet
Having acquired the cabinet, the next step was to visit its sister in the splendour of Powis Castle. The National Trust’s curators were very helpful and allowed me excellent access to their cabinet on two occasions, on the second of which Jim Broughton was also able to be present.
The Trust has records of the cabinet being in the Castle for at least a hundred years, but none before that date. Its condition is certainly consistent with life in a grand house. The exterior is somewhat faded, but the oak linings of the internal drawers are as fresh as if they had only rarely been opened, and the interior veneers are a glorious golden brown.
Like the Richmond cabinet (as I will refer to mine, to distinguish it from its Welsh sister) the Powis cabinet is oyster veneered in cocus onto a pine carcase, with finely cut mouldings of solid cocus run along the grain, both on the exterior of the cabinet and on the drawer edges (3). With minor differences, the overall construction and design of the two cabinets are essentially identical; more particular features in common include a tromp l’oeil double drawer in the interior, a slide (more probably for display purposes than for writing), also veneered in cocus, which is concealed in the moulding above the base drawer, and the locations of the secret drawers.
One of the more unusual features of the Powis cabinet is an almost embryonic apron below the base drawer, made from oak veneered with cocus. It stands on five bobbin-turned legs of solid cocus, with straight stretchers and feet also of cocus. Marks were clearly visible on the underside of the Richmond cabinet where the apron and the five legs had been, which would eventually help to guide the restoration process.
The Powis cabinet has had some restoration, probably in the late nineteenth or early twentieth century. All the metalwork appears replaced with the possible exception of the internal drawer handles. The stretchers have been re-veneered in what looks like rosewood and may well be replaced in their entirety.
There is a mirrored perspective in the central recess, typical of continental cabinets of the period (4) and possibly contemporary, but probably not part of the original construction of the piece; it seems slightly too large for the space and the internal door is a replacement, perhaps because the original door had to be removed when the perspective was fitted (5).
Antique Cabinet Restoration
Having had the advantage of thoroughly examining my cabinet, I was able to point the curators at Powis Castle to a secret drawer inside the central recess, in which we found a small leg bone. I understand that it has since been identified as of animal origin, possibly a dog, but it is anyone’s guess as to when it was first put there, or for what purpose!
Overall Approach to Restoration
The most important decision in restoring the cabinet was, of course, to choose a restorer. It needed someone who would take a rigorous line even when it exceeded normal commercial practice. I have been very lucky that Jim Broughton (www.broughtonrestorations.co.uk), whose high standards I was already familiar with, has matched my enthusiasm throughout.
The only acceptable approach to restoring the Richmond cabinet was to apply as faithfully as possible whatever we could learn from Powis. We were fortunate in that the two pieces were almost entirely mutually complementary – what was missing or replaced on one, mostly was present on the other.
As mentioned above, the whole of the stand below the base drawer needed to be constructed, for which Powis provided a complete guide in all respects except probably the stretchers (6). We took the view that their straight form reflected the originals, which would have been veneered in cocus, probably along the grain.
Jim and I were both clear that we wanted to adopt a fully transparent approach to the restoration, with the process fully documented. Where anything needed replacing, the most appropriate materials would be used, if possible contemporary with the piece. This would cause more difficulties than we had at first envisaged!
Materials for Restoration
We were obviously going to need quantities of cocuswood for the five legs, the missing mouldings and to replace missing veneer. I soon discovered that no-one in the UK stocked it, or was able to source any, or even knew anyone that might.
Cocus (Brya ebenus) is a very dense (7), somewhat brittle tropical hardwood which became prized by flute and bagpipe makers during the nineteenth century. With nothing done to ensure the sustainability of use, stocks were exhausted almost to extinction.
A search of instrument makers’ forums told me that cocus was regarded as completely unavailable, except possibly in small quantities and at exorbitant cost. We were going to need five or possibly six large logs (one per leg plus enough for mouldings and veneering).
Settling for a substitute would have compromised the integrity of the whole restoration – but when I was offered two logs at a cost of US$5500 plus the shipping and import duties, I must admit to having second thoughts.
After a lot of searching, we eventually found a wood yard in the Netherlands with a few logs of cocus for sale – apparently the only supply in Europe. They were priced by the kilo and although not cheap, the cost was manageable. My wife very kindly agreed to share the driving on a rather lengthy day trip to the docks of Amsterdam. In total, we purchased 107 kilos (in five logs), which together with the fuel, two Channel crossings and a Belgian speeding ticket made for an expensive day!
The only other wood we would need was pine for the stretchers and to repair the external drawer fronts and oak for the apron. This presented no problem, as Jim had sufficient period materials available.
That left only the metalwork. Fortunately, the Richmond cabinet had good brass hinges, slightly splayed and with five small holes for nails on each leaf, and one good door bolt from which the other could be cast. Its two iron door locks were also intact. However, we needed four matching drawer locks and a fifth for the base drawer, as well as various handles and escutcheons.
Jim took on making the locks, as the chances of finding matching period locks seemed very low. Google was again our friend, enabling us to locate what claims to be possibly the world’s sole supplier of genuine wrought iron (8).
A cabinet in Jim’s possession has escutcheons in pleasing early patterns, conveniently in two sizes from which casts could be taken, the larger for the external and internal doors and the base drawer, with the smaller being used for the four external cabinet drawers.
The internal drawer handles on the Powis cabinet are the standard English bifurcated scroll drops (9). However, there were no signs that wire fixings were ever used on the Richmond cabinet. We decided that the right option was to use simple round brass knobs. Good reproductions are easily available commercially.
All the brass work would be gilded to match the hinges of the Richmond cabinet, on which the original gilding was clearly visible.
The Restoration Process
Jim began the restoration process with an in-depth examination of the cabinet, removing all the detachable parts, vacuuming a large amount of accumulated dust and dirt and carefully cleaning off the ebonising to which the cabinet had been subjected, having first softened it with a little methylated spirits.
Because cocus is so dense and also oily, the ebonising had formed a surface layer rather than sinking in, as would have happened with a more open-grained wood. It had concealed a number of replacements and repairs to the veneers and mouldings. All of the replacement pieces being of mahogany of considerable age, with only one set of holes for replaced drawer handles, Jim’s view was that the cabinet has been worked on only once before, almost certainly in the late 18th century, when the chest was constructed for it to sit on.
Having removed all of these later replacements, Jim stabilised any loose veneer or mouldings, flushing out perished glue and debris with a syringe and hot water, with hot glue made from animal hide then syringed behind the veneer. Each piece of veneer then needed clamping with a block of wood which was covered in paper to stop it sticking to the veneer; the next day it could be removed and any surplus glue and the paper cleaned off with water, with the whole process then repeated on the next patch.
Shrinkage to the pine timbers had caused more serious issues in some parts of the cabinet, such as the slide and the doors. The slide had some rather botched repairs, with slivers of wood and filler inserted in an attempt to conceal the cracks. Rather than continue with this approach, Jim decided to dismantle and re-build the slide, which required the lengths of veneer that formed the banding at each side to be removed with the use of a hot air gun and water.
They came off easily and cleanly, the glue being somewhat perished and the veneer being very thick. Jim removed the side cleats using the same technique, and then re-joined the two boards using a shooting board and jointing plane, before re-assembling the slide (now very slightly smaller, but not visibly so once it was in place in the cabinet) and replacing the veneers.
Both the right-hand external door and the internal door were similarly disassembled, with a fillet of wood inserted into the crack and the pieces of veneer replaced.
All five external drawers had been rather butchered at the front. The locks had been removed and a patch inserted into the drawer front, covered with two pieces of mahogany veneer. Even with the veneers and the lock replaced, an unsightly patch would still have been left on the interior of the drawer front.
Jim resolved this by removing the strip of cocus veneer which covered the top of the drawer front, reducing the thickness of the drawer front by ⅛”, then adding a veneer of period pine of equal depth which was taken from a breaker. With the strip of veneer at the top reinstalled, the patch was entirely invisible from the inside of the drawer.
With locks in place and new pieces of cocus veneer around the lock escutcheons covering the holes left by the later handles, the drawers would again look as originally intended.
The legs and feet were turned on a lathe, a process which Jim saw as completely standard but which caused me some anxiety nonetheless, as we couldn’t know whether the cocus logs were of sufficient quality until then. Despite some parts suffering from excess sap and rot, the wood stood up to the test and the legs looked very beautiful with their bobbin turnings and elegant feet.
Veneering the stretchers, which were made from period deal with half-lapped joints, and the apron which for additional stability was cut from oak, was a different test of Jim’s skills. The grain of cocuswood is uneven and easily tears out, making it tricky to cut long strips of veneer.
He started by cutting a log in half down the length, then planing it as flat as possible, finishing it off with a scraper plane and a razor-sharp cabinet scraper. Veneers were then cut from the band saw ⅛” thick and glued with the planed side down. Once they were fixed into place the saw marks could be removed, again with a scraper plane and cabinet scraper. This gave a beautiful finish with greens and yellows much like polished stone, which then rapidly began to oxidise, changing to a warm nutty colour.
Jim found some interesting knife marks under the veneers. A single, true knife cut can be seen, corresponding to the edge of each piece of veneer. This might be the result of adjoining pieces being overlaid and then cut through. But with such thick, brittle and dense pieces of cocuswood veneer, it is not clear that this technique could have worked, even if the cuts were made with a shoulder knife of the type used for marquetry (which of course would imply the requisite knowledge and experience by the cabinet maker). Another possible explanation may be that a knife was used to clean up any glue or rough edges, before putting the next piece in place.
Scratching the mouldings was also made more problematic by the variable grain. Jim made up a scratch stock by grinding a hack saw blade to the appropriate profile, trapping the blade in an L shaped section of timber, and scraping it along the grain to produce mouldings which matched those originals that had survived. Because the wood is so dense, it was hard work cutting the mouldings; but for the same reason, intricate mouldings can be cut from wood like cocus which might be problematic with a more open-grained timber such as walnut.
A moulding was needed for the centre of the cabinet to cover the door opening, which had widened due to shrinkage in the doors. There was no pattern to follow as both Powis and Richmond had the wrong mouldings on the door, so Jim based its profile on the other carcase mouldings.
Once all the cabinet work was done, finishing was all that remained. Previous poor preparation work had left quite a few scratch marks which were scraped out before the whole surface was burnished with 400 grit paper. This brought it up so well, it was almost a finish in itself.
The new veneers and mouldings had oxidised over two or three months and reached almost the right colour, but excessive colour variation on some pieces needed addressing. The timber was too oily for a water stain to penetrate, and a modern naphtha stain would just have sat on the surface. Jim’s preference was to use wax alone as far as possible, and where necessary to promote more oxidisation with nitric acid. This removed some of the oil, allowing a thin coat of water black to penetrate until the right colour was achieved.
After sealing the surface with a thin coat of shellac, Jim built it up with wax to get a depth to the finish. Like other aspects of the restoration, cocus presented a very different finishing challenge compared with other woods, and a fair bit of trial and error was required. The end result was a beautiful, deep lustre, with plenty of colour variation and figuring in the grain, and the occasional lighter flashes from pieces of sapwood.
That left the locks, which would be cut from two pieces of puddled iron sheet. First, Jim made brass templates for each component, based on a seventeenth iron drawer lock, with the backplate and the positioning of the bolt adjusted to fit the drawer front and the cabinet. Having made one complete lock first to be sure it fitted, he then cut all the components, each with its own tenon, then filed and fettled it to produce a good
interference fit even without peening. Getting the spring into the back-sprung bolt was the hardest part, requiring quite a bit of experimentation with annealing and tempering the metal.
The door bolts, drawer knobs and drawer and door escutcheons were gilded to match the original gilding visible on the internal and external door hinges. Only then was the finish built up with layers of wax and lacquer, to create depth and variation, allowing the contours of the metalwork to show through.
After many final touches from Jim, it left the justifiably proud restorer, and the cabinet’s equally proud owner, with nothing more to do than to organise a final photo session!
The Cabinets’ Origins
From The Age of Walnut onwards (10), the literature has gone back and forth on whether or not cabinets of this class should be regarded as English.
The use of tropical hardwoods for veneering was widespread in many centres on the Continent by the early and mid- 17th century. There are undoubted similarities between the Powis and Richmond cabinets and the less extravagant contemporary French, Flemish, and Dutch pieces, some of which are oyster-veneered (sometimes in star patterns) and largely flat-fronted, with raised drawer mouldings and on stands with base drawers (11).
Specific features in common include the slide, the double drawer simulating two drawers, and the small apron. Another point which Dr Bowett notes as typically French (12) are the small mitred joints used in the internal drawer bases of the Powis and Richmond cabinets. None of these are commonly found in the form of English cabinets which became established in the latter decades of the century.
However, the materials employed – oak drawer linings and cocus veneered onto a pine carcase – and the surviving metalwork are all suggestive of an English origin (13).
The basic construction of the cabinet carcase is consistent with English work of the period: a pine box section, the four slabs held together with large, heavy dovetails – easily visible where mouldings were removed during restoration – with internal divides, also of pine, housed in ploughs in the carcase sides and slotted in from the back. Note, however, that some Continental cabinets had similar construction techniques, albeit with the carcase as likely to be of oak, and the divides of other softwoods such as poplar or lime, as of pine (14).
Likewise, the external drawers are of customary English construction and materials. The surviving metalwork also conforms to contemporary English patterns, and the bobbin- turned legs of the stand are far from unusual on English furniture of the period.
Restoration of the Richmond cabinet offered us the possibly unique opportunity to examine its construction, including what there was to be seen underneath the veneers. Further research is now being carried out on this class of early veneered cabinets, including the Powis and Richmond cabinets, with a view to publication in due course.
For their encouragement and advice at various points in the process I should like to thank Adam Bowett, Millicent Creech, Michael Durkee and Peter Hall, as well as Margaret Gray at Powis Castle and Gwen Thomas at Ham House together with their staff, for their kind co-operation and assistance.
1. Adam Bowett, English Furniture 1660-1714 From Charles II to Queen Anne, (Woodbridge, 2002), plates 2.10-2.15, pp.41-43.
2. Dr Bowett has been kind enough to endorse this view.
3. The half-round moulding on the main door of the Powis cabinet, visible in one of the illustrations in Dr Bowett’s book, is a later addition.
4. See, for example, Monique Riccardi-Cubitt, The Art of the Cabinet, (London, 1992) p.59.
5. The Richmond cabinet has no perspective; the central recess is oyster veneered which conforms to the rest of the cabinet, as does the internal door.
6. With the cabinet in situ, we were unable to ascertain whether any part of the stretchers were original.
7. Cocuswood is rated as 3720 on the Janka hardness scale, giving it a place somewhere in the mid-thirties of the world’s 125 hardest woods. Its specific gravity of 1.19 means that it will sink in water.
9. As mentioned above, we couldn’t be sure whether these handles were original or not.
10. Percy MacQuoid, A History of English Furniture –The Age of Walnut, (London, 1904-8), p.110.
11. Illustrations of such cabinets can be found in Reinier Baarsen, 17th Century Cabinets, (Amsterdam, 2000); see also Agnès Bos, Meubles et panneaux en ébène, (Paris, 2007), p. 10 (Fig. 3) and p. 21 (Fig. 10).
12. Bowett op. cit., p.41.
13. Mouldings similar to those on these cabinets can also be seen on English clocks of the mid-17th century. On display at the British Museum is a longcase clock by Joseph Knibb and a table clock by Henry Jones of London, both dated to the early or mid-1670’s, both with finely cut mouldings run along the grain.
14. See, for example, Wilfried de Kesel and Greet Dhont, Flemish 17th Century Lacquer Cabinets, (Oostkamp, 2012), pp. 86-7.