Historical Furniture Finishes – the Wax Polissoir
Yannick Chastang investigates 18th-century finishing techniques
The publication of the English translation of A.J. Roubo’s technical treatise on cabinetmaking (To Make as Perfectly as Possible published by Lost Art Press in 2013) has sparked considerable interest in 18th-century furniture techniques. His encyclopaedic volume has led to the rediscovery of many long forgotten techniques and recipes and woodworkers around the world are currently carrying out practical experiments to test Roubo’s practices. As part of this global inquiry, furniture finishes have attracted a lot of interest. However, our understanding of historical finishes is still clouded by myths and misunderstandings. Yannick Chastang has had a head start on many of his English speaking counterparts having had access to Roubo in its original French for the past 20 years. He has been researching and experimenting with original recipes, from Roubo and from other 17th- and 18th-century woodworkers with a particular interest in wax finishes. Wax finish has, for a long time, been a favoured option for both makers and conservators and its application using the little known ‘polissoir’ provides an insight into 18th-century finishing techniques.
Myths and realities
Our understanding of historical finishes is still very patchy and is based on a variety of incomplete historical sources. Material identification, using some of the latest scientific analyses, is slowly helping to fill the gaps in our knowledge but very few pieces of furniture, if any, have retained their original finish meaning an absence of viable samples for analysis. The effects of time and regular restoration have destroyed most original finishes. If, by chance, an original finish has survived it’s unlikely to be intact; regular dusting, cleaning and natural ageing and drying out of the materials (often referred to as cross linking) would, at best, have resulted in the loss of original transparency, colour and gloss; at worst, the finish will have perished entirely. Even on the highly prized musical instruments made by Stradivarius, many of the finish surfaces have either degraded or been replaced. Great mystique still surrounds the sound quality of Stradivarius’s instruments, which is supposedly linked to his unique choice of finish; essentially an oil-based resin varnish. Unlike musical instruments, furniture’s original finish was far less valued and was deemed to have the practical purpose of enhancing the wood and protecting the surface. Furniture finishes were regularly renewed, either because they had degraded or as part of a repair of the piece as a whole.
A lack of reliable evidence therefore has meant that the true nature of 18th-century furniture finish is still the subject of debate. Two schools of thought have emerged in recent years: one believing that most pieces of furniture were finished with wax, the other advocating the popularity of a resin-based hard finish (French polish). Both options were available during the 18th century and Roubo and his contemporaries accurately describe many of the available techniques. Ancient treatises suggest that a wax finish was favoured for ordinary furniture while resin-based finishes were favoured for high-end furniture. However, this is contradictory to many historical facts and sources of primary evidence, which prove that wax finishes were prevalent. Do we therefore have to question the extent to which Roubo and his contemporaries were well informed about true workshop practices? Was Roubo a cabinetmaker or an enclopaedis; a compiler of information rather than someone actively engaged in all the trades of which he records details?
Archival evidence, such as the bills of the royal cabinetmaker Jean-Henri Riesener, contradicts Roubo’s claim that high-end furniture normally had a resin-based finish.
In 1776, Riesener, with the help of two craftsmen, was at Versailles restoring the famous Bureau du Roi and a commode. Based on his bill, it is possible to establish that in less than two weeks, Riesener and
his helpers dismantled, scraped and sanded the marquetry, cleaned the bronzes and repolished the Bureau du Roi, and worked on the commode (avoir été avec deux ourviers….avoir raclé et repoly tout la marqueterie et netoyé la garniture de bronze). This was obviously an amazingly productive two weeks, the Bureau du Roi being one of the most complex pieces of furniture ever made and its dismantling alone would have taken days. As regards the finish, we know that the resin-based polish available at the time of Riesener (and Roubo) would not dry quickly enough to be applied in this short a time frame. Riesener’s bills for mahogany furniture delivered to the crown in 1786 also clearly mention ‘poli a la cire en dedans et en dehors’ (wax polish inside and out).
The study of historical recipes needs to encompass the available materials and their inherent limitations at the time. Today it is possible to apply a respectable French polish using shellac in only a few days, ‘French polish’ at the time of Roubo was much different and its application was greatly influenced by the materials to hand. It is in fact believed that French polish, as we know it today, is a 19th-century improvement on previously used polishing techniques. Most of the techniques described by Roubo, such as the use of pumice powder, and most of the materials described in Roubo appear to be those used in today’s French polish. However, Roubo and his contemporaries do not favour shellac but, instead, praise the natural resin sandarac, with its application by brushing. The reason for using sandarac instead of shellac is easily explained by the colour of shellac available at the time. Modern day alcohol used as a solvent in French polish differs greatly from its 18th-century equivalent, the so-called ‘esprit de vin’. Distillation processes have improved to a point that modern alcohol has a water content of less than 5%. 18th-century alcohol is estimated to have contained, at best, 65 to 70% alcohol and the rest water. The high water content of the esprit de vin would reduce evaporation rate and would have resulted in a cloudy finish. Additionally, the shellac available at the time of Roubo differs in colour and wax content. Archival evidence reveals that shellac was only available in a dark version, similar to our modern sticklac, seedlac or the more refined dark garnet shellac. Lightly coloured shellac (commercially known today as de-waxed or blonde shellac) is obtained via a difficult process of removing some of the natural wax content from the pure sticklac and by chemically bleaching the shellac.
As a result, Roubo only recommended shellac for dark woods such as rosewood (Dalbergia retusa). Applying a dark shellac finish to a colourful marquetry (on both European and English furniture) or to a light-coloured wood would have spoiled it. Instead of shellac, Roubo recommended using the best ‘white polish’ for light coloured marquetry. According to Roubo, the best white polish was not only suitable because of its almost translucent colour but he also states optimistically that the best white polish would stop the colours of the marquetry from ‘evaporating’ and stay colourful longer (‘saisir leur couleur, qui, ne pouvant plus s’evaporer, reste toujours dans le meme etat’). This is an amusing misunderstanding of the true reason for the discolouration of wood but it demonstrates that UV protective varnish was ‘invented’ during the 18th century (albeit not very successfully!). Sandarac-based finishes were not invented by Roubo and the best white varnish, as described by Roubo, finds its origin, almost word for word, in the treatise on ‘Japaning (sic) and Varnishing’ written in 1688 by Stalker and Parker.
On paper, a sandarac-based finish provides the perfect solution. Sandarac finish produces a clear and translucent finish; it beautifully enhances the colour of the wood and does not obscure marquetry. However, without exception, all experiments have demonstrated that sandarac finish is extremely slow to dry and is very difficult to work with. The author’s attempts to create a thick and glossy finish, similar to today’s French polish, were only successful if layers between applications were left to dry for several weeks. One must therefore conclude that if sandarac was used, it was limited to a couple of applications.
In view of such evidence, it appears that a wax finish must have been the most practical option and therefore the most ‘authentic’ finish used during the 18th century for the large majority of furniture produced. Darker furniture may have received an application of dark shellac while the more expensive colourful furniture could have been finished with a couple of layers of sandarac, but furniture almost certainly never had the thick glossy appearance commonly associated with French polish.
The ‘polissoir’ by Roubo
A wax finish does not necessarily denote a matt finish. The late 18th-century portrait paintings by the American John Singleton Copley often depict a sitter next to a table or a piece of furniture. In the majority of Copley’s paintings, the sitter’s image is reflected in the furniture’s top, suggesting that the furniture was finished to a high level of gloss. It is generally believed that the primitive nature of 18th-centrury materials would have precluded a wax finish, however the practice of achieving gloss on 18th-century furniture may have been much greater than first assumed. Wax recipes that use a combination of hard wax and soft wax were discussed in the earlier article ‘Is wax the only finish you need?’, F&C, issue 189, February 2012.
Anyone experimenting with wax will know that it is a difficult finish to apply; for a successful application you must first consider how much to apply and how hard or soft the wax should be. The secret of a good application is the thickness of the wax; too much and the surface remains tacky, not enough and the coverage is poor with application marks. The quality of the wax will also affect the end result as will the solvent to wax ratio. Again, wax recipes were shared in F&C, issue 189.
To ease application, Roubo recommended the use of what is called in French a ‘polissoir’, a wax burnisher used both for application and for burnishing the wax finish. Roubo describes the polissoir well, albeit briefly, on plate 296, figures 8 and 9 and on page 859, section III (1774 French edition). Unfortunately for English readers, the 2013 translation To Make as Perfectly as Possible by Donald C. Williams contains some errors and one of these relates to the polissoir. As a result, the wrong type of polissoir is now being made and marketed in the USA. This criticism is not meant to detract from the mammoth achievement of Don Williams and the authors of this translation, however to fully appreciate the potential of the polissoir it needs to be re-visited. In an interesting parallel Roubo in 1774 already knew of cabinetmakers who were not using the correct material to make the polissoir and wrote that other materials were ‘pas bon’ (not good). These materials often result in the marquetry lifting as they overheat the surface and the animal glue used to adhere the marquetry. Roubo described the ideal polissoir as ‘un faisceau de jonc ordinaire’, in translation a ‘sheaf or bundle of common rushes’. Rushes are also known as Juncus and are commonly found in wetlands. Rushes vary in size and, in my experience, the best results are achieved by using tall rushes found growing in small rivers (some compensation for a bad day fishing is the accumulation of rushes collected while wading in the river). Despite the brevity of information Roubo accorded to the description of the polissoir, he puts great emphasis on the use of the correct material and stresses the importance of using rushes rather than any other reed or straw on page 862. Wetland rushes are entirely different to the corn grass or broom straw used by Don Williams and in the commercially available American polissoir. Common rushes have a soft foam-like interior and a relatively thin external wall. Growing in wet sandy conditions, common rushes are also believed to contain other beneficial minerals.
Making your own polissoir is very straightforward, once you’ve identified which rushes have the correct natural properties. The polissoir can be made as follows: cut common rushes into 200mm lengths; dry them and fasten a handful of dried lengths together with cord. Once the bundle is securely bound, soak the polissoir in melted wax of any type (my favourite wax is 1/3 shellac or carnauba and 2⁄3 beeswax). Thanks to the porous nature of the rushes’ culms, the wax is absorbed along the bundle. Once the polissoir has cooled down and the wax hardened, saw the end off the bundle and then smooth the end with rough sandpaper. I make different sizes of polissoir as I like to have some with a small diameter that can better access recesses and corners on my furniture.
Using the wax-impregnated polissoir is very easy but also very physically demanding. The flat end of the polissoir is rubbed over the wood surface but you need to apply considerable pressure and energy. Sufficient heat must be generated to warm the wax and force it into the pores of the wood filling the grain with wax. If necessary, some extra wax can be added at the same time. Witnesses are often amazed by the quality of the burnish and the speed that the wood can be finished.
When I first learnt French polishing, I was taught that ‘you cannot have a good polish without a good grain filling’ and I still adhere to this advice. I therefore religiously sand the surface as thoroughly as possible with pumice powder so that the grain is filled prior to applying the French polish. However, the polissoir technique works virtually to perfection even on open grain timber and even if poorly prepared. The success of the polissoir is clearly down to the mineral content of the rushes. The rushes appear to be slightly abrasive and the combined wax plus abrasion clearly helps fill the grain. Inevitably, the polissoir will leave some application marks, however, those can be removed by polishing the surface with a clean cloth, if necessary adding a little more wax to the surface at the same time.
Wax surfaces finished in this way are durable and easy to maintain. Wax can easily be added at any time on the surface and burnished back to a fairly high gloss level, depending on the wax used. Although Don Williams apparently achieved great results, my own experience of a corn grass polissoir compared to a rush polissoir has demonstrated that the rush version is far superior. This reinforces my belief that, once again, Roubo (preferably in its original version) is the most reliable source for anyone studying 18th-century furniture traditions and can teach us much that is relevant today.