Chippendale’s Harewood Writing Table

Reproduction door panel

 Marqueteur Jack Metcalfe describes a unique project, which reveals, for the very first time, how Thomas Chippendale applied marquetry to his furniture.

Thomas Chippendale’s Library Writing Table was made for Edwin Lascelles of Harewood House c1771.(Near Leeds, Yorkshire, England)  It formed part of the largest and most prestigious commission of Chippendale’s career, to furnish the house from top to bottom. The work was to last for 30 years.  The library writing table epitomises the neo-classical theme, full of style and grace, so synonymous of that period. The table remained at Harewood until 1965, when it was sold at Christies, at auction, for £43,050, then a world record price for a piece of furniture, and purchased by the Leeds City Arts & Galleries. It is now on permanent display to visitors to Temple Newsam House, Leeds.

 

The Chippendale Millennium Exhibition was held at Harewood House in the year 2000, and the writing table made a brief return to its original home. It was considered appropriate at that time to take the opportunity to carry out some much needed restoration to the door panels, where surface veneers were lifting, due to movement of the oak substrates.  This work (carried out by Ian Fraser – Conservator, Leeds Museums & Galleries & Dr Adam Bowett – Furniture Historian) permitted access to the veneers and a chance to see, for the first time in 230 years, what colours the veneer were after lifting and turning them.   The entire marquetry, it was found, consisted of holly (Ilex aquifolium) set into Indian rosewood (Dalbergia latifolia), but more surprising, to discover an array of dyed greens and reds, together with the natural white holly, which made up the neo-classic masterpiece. The central panels were surrounded with a West Indian satinwood (Fagara flava) narrow banding & green stringers, and finally bordered with South American tulipwood (Dalbergia flutescens) cross banding.  A school in Otley, Leeds (ironically the birthplace of Chippendale) produced a computerised simulation of the panel, by scanning in the colours and matching them to the appropriate parts of the design. It was like painting-in-wood electronically and the result was amazing. 

In 2003, the BBC asked if one panel could be reproduced in wood to form part of a documentary on the life and work of Thomas Chippendale. I was approached to perform the marquetry work. Due to short timescales, it was not possible to dye the holly to the exact colours found on Ian’s original restoration work, therefore forcing me to use veneers from my existing stock. While the reproduction did serve to illustrate the stark difference between the coloured veneers, compared to the present mid brown colours caused by over 200 years exposure to the ultra-violet rays, the work did not satisfy the now burning desire to make a ‘precise’ reproduction.  Ian, Adam and I, decided to undertake further investigations into the original work. More veneers were selectively lifted from each of the four identical door panels. Each discovery digitally photographed, until an exact ‘map’ was re-created. As a Marqueteur, I was also keen to establish the method of cutting and laying the veneers prior to assembly. Many clues were laid out before me on the original doors and drawer fronts. The acanthus leaves, for instance, clearly shows the kerf (gap) of the fret saw blade, (see bottom leaf at Fig 2) proving that they, at least, were cut as a packet. However, the laurel leaves, which formed the swags, clearly indicate that the grain direction of each leaf changes as the shape of the swag changes, proving that the three petals making up the leaves were cut out separately, pushed together to remove the kerf, then sawn into the rosewood background. Two-part sawing demands great accuracy, and the tight joints of the laurel swags pictured at (Fig 2) illustrates the level of skill they had, which will also be needed by myself, in order to replicate the technique.

 

                               
Fig 2
 The primary source for dying the veneers came from Peter Weber’s Cabinet-maker’s guide to the Whole Art of Staining, Varnishing and Beautifying Wood (Second Edition 1809).  Three shades of green was produced, two to depict the acanthus (dark green) and laurel leaves (mid green), and a third, but lighter shade for the stringers. These colours were obtained by combining Barberry yellow with Indigo blue, and an addition of natural mordents such as Alum, or aqua fortis (nitric acid) sealed the dyes. Similarly, three different reds were needed. The central vase consisted of a burgundy colour, which was produced with Brazilwood chips. Another, pinkie red, was achieved using Dragons blood, a resin obtained from berries of the Dragon tree. Finally, a bright red colour was needed for the berries around the laurel leaves and in and around the small compass roses and plumages.  Mexican cochineal produced the desired red, and, after many trials to get the colour to penetrate the veneer thickness, Ian found that tin chloride finally provided the right result.

The Indian rosewood used for the background veneer was degreased using acetone, not a substance available in the 18th Century, but Fullers Earth (I am indebted to Michael Haines (BAFRA) for this information) would have most likely provided the same result. It was necessary to degrease the rosewood to prevent the black oil, inherent in the wood, from bleeding into the lighter holly during the polishing process.

 

 

 

The construction of some of the marquetry closely followed the methods and tools used to construct the original. I fret sawed the acanthus leaves using the packet method. Fig 3 illustrates the packet, where the left and right-hand-sides of the design are folded - book like, with the green veneers trapped in ‘the sandwich’. A fret-sawing-donkey with the horizontal saw frame (Chevrolet) was used to saw out the packet. On completion, the packet is opened out, like a book, and the resulting fret work is assembled, as seen at Fig 4.  You can see that the two sides of the design form a mirror image of each other.

I have to confess that the remaining fret work was performed on a modern electric ‘Hegner’ fret saw machine, mainly because my arms could not cope
with the physical demands of a marqueteur’s donkey! However, the exercise did serve to make me realise the high skill and dexterity achieved, not to
mention the muscles 18th Century cutters developed with constant sawing in the arm-stretching horizontal position.  

The remaining components were fret sawn using the two-part sawing method, where the components are sawn out in isolation, then the resulting ‘window’
was sawn in the background veneer to accept the component.  Great accuracy in sawing the background is vital to achieve tight joints.  Bright red berries
were inserted, using a hollowed out punch, made specially to produce the correct diameter (approx 1.25mm) berries. The berries are constant in size
and shape, suggesting that a tool which produces repeat copies would have been used. Sand shading was applied extensively, as in the original work,
to create the all important three-dimensional effect.

The veneers were glued to the substrate with PVA. I felt this to be the best choice, mainly for well proven reasons - PVA is not affected by temperature
and humidity change, and accepts polishes without any short or long term affects. It is a user-friendly proven product used extensively by today’s
manufacturing industry. On the other hand, animal glue has the advantage of being reversible, but sadly it is affected by centrally heated buildings.
The need to reverse the glues on these panels should not arise. To safeguard the coloured veneers from fading, the room at Temple Newsam where
the panels will be displayed, is installed with UV filters.  

 

After making up the 7mm wide satinwood cross banding, with its green stringers, I cut tulipwood, across the grain to form the cross banding border.
Two raised oak pilasters veneered with tupilwood will eventually hold the brass work.

 

The assembled veneers were sanded flat. It was at this stage that the burgundy dyed holly, which we dyed with Brazilwood, became blotchy.
It was evident that although the dye had struck through the veneer thickness, it had done so unevenly. 
Ian immediately pointed out that the same veneers are also blotchy on Chippendale’s original panels.  How right he was, Chippendale’s artisans had,
as we first suspected, used Brazilwood to obtain the burgundy colour, but had also experienced the same blotchy state when flattening the veneers.
The evidence is clearly visible as different shades of brown across the effected veneers. Since he didn’t try to alter it, neither shall we.
The unplanned event only serves to prove how close we are in our reproduction.

The inside of the door panels and drawer fronts (not exposed to daylight) on the original table is a deep mahogany colour, suggesting that the Indian rosewood (normally a brown/purple colour) was stained prior to polishing.  Martin Speak, Head of Leeds College Art & Design stained the rosewood after masking off the marquetry with shellac sanding sealer. The stain used was an aniline water-based type, selected because it is the very opposite material to the spirit-based shellac, forcing the latter to repel the stain where it was not wanted.  A final decoration of pen work is to be added to most of the marquetry, using a fine open-nibbed pen and indelible black Indian ink. Two cabinet-making students (with artistic skills) from the Leeds College - Alison Monks and Simon Brock, will apply the ink work.  Finally, Martin will apply French polish (sandarac) to complete the polishing. 

The two photographs above show the new panel on the left, compared to the original panel on the right.

The original panel is fitted with gilded brasswork which was not used on the new one. The degree of fading after more than two centuries of exposure to daylight can be clearly seen.

The project, now nearing completion, will produce two door panels, which will stand alongside the writing table and show visitors how the neo-classical showpiece would have looked when first constructed. This project, which started 5 years ago, has been performed in the spirit of conservation and education, and by all players, on a purely voluntary basis.