Category Archives: Tools and Techniques

Roubo and Watin: The Sweet Scent of Early Modern Varnishes

By Érika Wicky

Fig. 1: Nécessaire, wood and leather, ca 1760, 8,4 x 6 cm (closed), Grasse. Courtesy of Musée international de la parfumerie.

If the decorative arts of the eighteenth century shine so brightly, it is thanks to the innovation and mastering of techniques such as gilding and varnishing.[i] While the latter may have given surfaces the shiny appearance that was particularly desirable at the time, both were above all essential to protect wood, especially against insects. Varnishes were used not only for furniture, but also for small painted objects such as snuffboxes, fans, and boîtes à mouches or fly boxes (Fig. 1). Varnishes’ appealing visual characteristics and protective properties were marred, however, by what was generally perceived as their unpleasant smell. The recipes for some of these varnishes have been passed down to us, and with them, the possibility to retrieve their odors. Some of these recipes continue to be in use, such as the varnish described by French cabinetmaker André-Jacob Roubo in L’Art du menuisier-ébéniste in 1774 (Fig. 2), whose use remains frequent in restoration[ii] so that the same recipe is regularly remade at the Center for Research and Restoration of Museums of France. While current olfactory accounts of these products are much more nuanced, they do persist, which is a sign that olfactory sensitivity has certainly evolved, but also that the smell of the varnish remains a major part of the experience of its manufacture and use. What are the substances that have given varnishes such a characteristic olfactive identity through the centuries? And how has the meaning attributed to these smells evolved?

Fig. 2 : André J. Roubo, L’Art du menuisier-ébéniste, Board 11 (Paris: 1774). Courtesy of WikiMedia Commons.

There are a number of recipes for varnishes found in eighteenth-century texts, and all essentially contain aromatic plant resins. For instance, the varnish used by Jean Félix Watin (1773) was prepared as follows: “Two ounces of mastic in tears and half a pound of sandarac in a pint of wine spirit, when the materials are well dissolved together, incorporate four ounces of Venice turpentine.”[iii] Similarly, Roubo’s (1774) was “composed of one pint or two pounds of wine spirit, five ounces of sandarac, two ounces of mastic in tears, one ounce of gum elemi and one ounce of oil of aspic.”[iv] Essentially, the process involved Venice turpentine (the name given to the resin of Melese[v]), mastic in tears, sandarac and other more precious resins such as copal, or of lesser quality such as galipot, which were generally dissolved in alcohol (wine spirit), and then heated. This gave off such a strong and unpleasant odor that the practice was forbidden within city walls.[vi] Indeed, this smell could be significant during the process of preparation, and Jean Riffault, for example, recommended that cooking copal should be stopped when it gave off an empyreumatic or burnt smell [vii] (Fig. 3).

Fig. 3: Hendrik van Reede tot Drakestein, India Copal (Vateria indica L.), colored line. Courtesy of the Wellcome Collection.

Although the smell of varnish released during firing is still considered strong and dangerous today, our current perceptions seem to accord less significance overall to the smell of varnishes. For example, the sensory-aware description of a varnish recreation at the V&A mentions that the smell of lavender dominated that of turpentine.[viii] That the smell of turpentine could be camouflaged was well known in the eighteenth century, when aspic oil was often adulterated with turpentine, which was much less expensive. To ensure the quality of the aspic oil, Watin recommended approaching a soaked cloth to a fire, which dissipates the lavender oil quickly and allows the detection of the odor of any turpentine that may have been added.[ix] Since most of these raw materials were subject to adulteration, the artisans could, in fact, evaluate their quality thanks to their sense of smell. For example, according to Baudeau’s Encyclopédie méthodique (1783-4), an unpleasant smell from gum elemi (which is usually pleasant) can reveal that it was substituted with galipot.[x] The olfactory consensus here seems to rest on the fact that the artisan knew enough about the agreeable smell of gum elemi not to be tricked. Similarly, olfactory references in professional treatises, such as Roret’s guide, often rely on experience, invoking “characteristic” or “particular” odors rather than defining them by analogy, like when elemi gum is described as smelling like fennel.[xi]

Far from being a mere nuisance, the smell of the raw materials thus made it possible to identify them, in the same way that wood, whose different species were characterized by their smell, was sometimes considered when choosing materials. Odour could be so defining that, as Cheng He has shown, the concept of lacquer was long associated with a range of odorous resins rather than with a finished object.[xii] For example, objects decorated with Martin varnish [xiii] were simultaneously prized in the eighteenth century and dreaded for their bad odor. They were distinguished in particular by the use of copal, a plant resin whose use as an incense in Mexico is often mentioned in technical treatises.[xiv]

The manufacture of varnishes thus intersects with other types of olfactory knowledge. The resins used for the varnishes are found in many other recipes, in particular for perfumes. Louis Peyron’s 1986 compilation of recipes for perfumes to be burned, which had been recommended as a way to defeat the plague in the 1720s, reveals that the components of varnishes were all found in these recipes and that their strong odor was believed to have the prophylactic virtue of keeping miasmas away.[xv] The connection between varnish and perfume can be further explored in eighteenth-century perfumery treatises: in 1761, the royal perfumer proposed a “varnish” for the complexion made with sandarac spirit, whose protective and preserving virtues aligned with those of wood varnishes.[xvi]

The negative connotations associated with the smell of varnish seem to have faded with time; tear mastic was used in perfumery into the nineteenth century, [xvii] while, according to perfumer Lucille Lefranc Gallo, gum elemi is now experiencing renewed interest in perfumery, alongside a range of spicy raw materials. The scent of varnishes produced according to eighteenth-century recipes, such as Roubo’s or Watin’s, takes us back to a time when smells had specific connotations and were part of a network of knowledge provided by olfaction, particularly with regard to the identification of materials. It is not only the sensitivity that has evolved, but also the entire olfactory culture.


[i] My warm thanks to Marc-André Paulin (C2RMF), Lucille Lefranc Gallo and Ersy Contogouris for their advice and comments. This text forms part of the work conducted for the Marie Sklodowska-Curie research project PaintOdor (845788).

[ii] Nathalie Balcar, Frédéric Leblanc et Marc-André Paulin, “La protection de surface pour les meubles en marqueterie de métal du musée du Louvre : étude d’un vernis, entre formulations anciennes et expérimentations actuelles,” Technè, 49 (2020).

[iii] Jean Félix Watin, L’art du peintre, doreur, vernisseur (Paris: 1773), 229. Watin claimed to have discovered the secret of an odorless varnish, but he does not give its recipe in his book.

[iv] André J. Roubo, L’Art du menuisier-ébéniste (Paris: 1774).

[v] Pierre Chomel, Abrégé de l’histoire des plantes usuelles (Paris: 1712), 190.

[vi] Watin, 224.

[vii] Jean Riffault, Manuel théorique et pratique du peintre en bâtimens, du doreur et du vernisseur (Paris: Roret, 1825), 254.

[viii] “This time, despite both ingredients being very strong smelling, the scent of the lavender oil completely overpowered that of the turpentine.” Simona Valeriani, “Recreating Sixteenth-Century Varnishes on the History of Design Course,” June 23, 2017,

[ix] Watin, 202.

[x] Nicolas Baudeau, Encyclopédie méthodique (Liège: 1783-1784), 63.

[xi] Riffault, 229.

[xii] Cheng He, “Understanding the Fragrance of Lacquer in Early Modern Europe,” University of Toronto Art Journal 9.1 (2021): 68-76.

[xiii] Anne-Solenn Le Hô, Céline Daher, Ludovic Bellot-Gurlet, Yannick Vandenberghe, Jean Bleton, Myrtho Bonnin, Léa Drieu, Juliette Langlois, Céline Paris, Marc-André Paulin, “French lacquers of the 18th century and vernis Martin,” ICOM-CC 17th Triennial Conference, September 2014, Melbourne, Australia. hal-01279161

[xiv] Chomel, 512.

[xv] Louis Peyron, Odeurs, Parfums et parfumeurs lors des grandes épidémies méridionales de peste Arles 1721, Talk presented at the Arles Académie on March 23, 1986.

[xvi] Le parfumeur royal (Paris: 1761), 115.

[xvii] Auguste Debay, Nouveau manuel du parfumeur-chimiste (Paris: E Dentu, 1856), 37.




Cherries Galore in a Cesspit

By Merit Hondelink

As an archaeobotanist, an archaeologist specialised in studying plant remains found in archaeological excavations, I aim to reconstruct and interpret the relationships between humans and plants in the past. Archaeological plant remains, also known as subfossil plant remains, help us to reconstruct the former landscape and inform us how humans exploited it and even transformed the vegetation. Archaeobotanists do not necessarily study one time period, nor a specific region or topic. They can study plant remains from the Palaeolithic or the 20th century, and everything in between. They can focus on one specific site, work across the country or continent, and even work worldwide. They can delve into topics such as natural vegetation, forestation, domestication, trade, food consumption and much more. The one thing that all of this has in common is the link between humans and plants. But most archaeobotanists do specialize, most notably in the plant parts they study, such as fruits, seeds, pollen, wood or phytoliths. And most archaeobotanists have a beloved time period, favourite region or topic that they find most intriguing. In my case my research focuses on early modern Dutch urban food consumption.

I study what people ate in early modern Dutch cities, and how this changed through time. The best way to study what people ate in the past, is to look at their excrement and kitchen refuse, both of which can be found in the archaeologists treasure trove: the cesspit. These latrines were used to empty one’s bowels, but also served as a place to discard kitchen refuse and household waste. The content of a cesspit consists of organic remains from plants and animals, inorganic (culinary) material culture such as earthenware, glassware and ceramics, but also wooden cups and plates, as well as (decorative) objects, personal belongings and much, much more.

Figure 1: A selection of faunal and floral items found in a late medieval cesspit sample from Groningen. Photo: Dirk Fennema.

The content of an archaeobotanical cesspit sample consists of, among others, floral remains in different shapes and sizes (Figure 1). The items are sorted with the use of a microscope (Figure 2) and identified on a species level (and sometimes even on the level of species variety) by using a reference collection (Figure 3). The Groningen Institute of Archaeology offers a wonderful digital, open access, reference collection, see

Figure 2: A peek through the microscope. Visible is a fragment of text and different seeds and fruits, taken from an early modern Delft cesspit sample. Photo: Merit Hondelink.

When the content of a cesspit sample is analysed, sorted and identified, the interpretation begins. What can these plant remains tell us about past human-plant relationships? Most plant species are interpreted in a standardized way: wild plants inform us about the vegetation composition, make-up of soils and hydrology, whilst agricultural weeds in particular inform us about the crops grown and their local, regional, international or even global provenance. Wild but poisonous or toxic plants inform us about potential medicinal applications. A majority of plant species found in cesspits are classified as economic plants, grown as a food crop or cultivated for other useful purposes, such as fibres for textiles or seeds for oil. Identifying edible plants helps us better understand what plants people used for food and which parts people consumed. It also helps us better understand how food was prepared in the past, as preparation marks can be left behind on seeds and fruits.

Some preparation marks are easier to identify than others: nuts need to be cracked to get to the seed; apple seeds may be sliced when cutting up an apple, cereals can be ground, resulting into fragmented bran. But sometimes the archaeobotanist finds fragmented plant parts that, at a first glance, do not make sense.

Figure 3: A small selection of the tubes from the archaeobotany reference collection housed at the Groningen Institute of Archaeology (GIA) at the University of Groningen. Photo via GIA.

I have come across dozens and sometimes hundreds (or even more) cherry stones and plum stones in a single cesspit sample. No surprise there, cherries and plums were grown in local orchards, sold in the market and consumed with gusto. Most of these stones will have been discarded in the cesspit as a result from eating the fruits and spitting out the stones, or after de-pitting the fruits for dinner preparation. Only a small percentage is assumed to have been accidentally swallowed and secreted as excrement. Still, archaeobotanists find many fragments of cherry and plum stones (Figure 4). This is something that raises questions when you think about it. Why would these sturdy fruit stones be fragmented? A more pressing question when you are aware that the Rosaceae family, among others also including almond, peach, and even apple, contains – to varying degrees – hydrocyanic acid, also known as hydrogen cyanide and sometimes called prussic acid. The seed coat and fruit wall protects the consumer from digesting this acid, which can be poisonous when consumed. So why would someone break the stones of these fruits?

Figure 4: Two fragments of cherry stones found in an early modern cesspit in Vlissingen. Photo: Merit Hondelink.

To test the assumption that cherry stones were fragmented intentionally, and not through, for instance, pressure, an experiment was devised. Cherries were bought at the farmer’s market and taken to a physics lab to measure the pressure required to fragment the stones. After a number of tests, the calculated force to fragment a cherry stone averaged 23,9 kg or 239 Newton (Graph 1). This makes it more plausible that the stones were intentionally fragmented, as opposed to – for instance – fragmentation due to soil pressure.

Graph 1: Force needed to fragment a cherry stone. On the vertical axis the force (N), on the horizontal axis the elongation (μm). The point where the line falls is the moment the cherry stone breaks (max. force – max. elongation).

Consulting early modern cookbooks provided me with a list of recipes requiring the cook to de-stone cherries for the preparation of jams, sauces, syrups and tarts. Delicious experiments ensued, but I did not manage to fragment cherry stones whilst cutting and de-stoning, pressing through a cloth or colander, or by just baking the fruit with stones in a tart in the oven. Working a batch of cherries with a mortar and pestle did the job, though. But than you would have to pick the fragmented stones from the mushy cherries: not ideal at all. Picking up the eighteenth century encyclopaedia compiled by Noël Chomel gave me the hint I needed. In the Dutch version of his Dictionnaire œconomique (Algemeen huishoudelijk-, natuur-, zedekundig- en konst- woordenboek), he mentions different recipes for preparing cherries. Two recipes for cherry liquor instruct the reader to fragment the cherry stones by using a mortar and pestle (Figure 5). The fragmented fruits, including the stones and (I assume) the seeds are added to the brandy (Dutch: brandewijn) and, after closing the bottle, the mixture is put in the sun to infuse. Adding spices such as cinnamon, cloves and sugar is optional, according to the author.

Figure 5: How to make a pleasant cherry liquor (Noël Chomel, 1778).

So, it is plausible that the fragmented cherry stones found in early modern cesspits are the result of the domestic production of cherry liquor. Other fruits, such as plums and peaches, are also used to make a fruity liquor according to Chomel’s encyclopedia. However, what happens to the acid contained in the seeds? That requires further research. It might be that the prescribed infusing in the sunlight helps denature the acid into harmless molecules, leaving only the (bitter) taste behind. This line of research will be undertaken come summer with the aid of a brewer and some chemical analysis. In the meantime, a cherry and cinnamon flavoured lemonade is my poison of choice. Bottoms up!

A Roman Vegetarian Substitute for Fish Sauce

By Edith Evans

Roman cookery has been one of my research interests since the 1980s; I’ve accumulated a large repertoire of ancient recipes and usually do at least one live demonstration a year.  Most of the recipes include garum or liquamen – fish sauce – as a taste enhancer, providing salt and umami. Whilst finding fish sauce is fairly easy nowadays in Britain (the Romans used the same techniques to make it as the modern Thai and Vietnamese), using it at demonstrations disappoints vegetarians who would otherwise like to sample the plant-based dishes.

I found the answer to this problem in a Late Antique agricultural treatise:

Liquamen from pears: Ritually pure liquamen (liquamen castimoniale) from pears is made like this: Very ripe pears are trodden with salt that has not been crushed. When their flesh has broken down, store it either in small casks or in earthenware vessels lined with pitch. When it is hung up [to drain] after the third month without being pressed on, the flesh of the pears discharges a liquid with a delicious taste but a pastel colour. To counter this, mix in a proportion of dark-coloured wine when you salt the pears.
– Palladius: Opus Agriculturae 3.25.12

Liquamen castimoniale must have been required for people observing certain religious strictures (castimoniale means ‘to do with religious ceremonies’). Why would ordinary liquamen have been thought unsuitable? Was it the fish? (Pliny the Elder writes of a special fish sauce for Jews (Natural History 31.95) that he calls garum castimoniarum, although he’s obviously got the wrong end of the stick when it comes to Jewish food laws because he says it’s made using fish without scales). Alternatively, was it because liquamen was the product of fermentation? Fermentation was often considered a form of decomposition, which might have led it to be regarded as ritually unclean.

This has a bearing on how we interpret the recipe. Although Palladius tells us the ingredients to use (whole pears and salt, plus optional red wine) he does not give any information about the relative proportions. This leaves us with two possible techniques. Either you use a high proportion of salt and effectively create a brine utilising the juice of the pears, or you use a low proportion and promote a lactic fermentation by incubating the mix a suitable temperature (although Palladius doesn’t mention this). When used to flavour food, the product of the first method adds a strong taste of salt but no umami. The second would add some umami but also acidity, but a much lower amount of salt. However, if the problem was the fermentation itself, the second method would have been as unacceptable as standard fish sauce.

I’ve had a go at the lactic fermentation method, using 2% of the weight of the pears in salt, but when I tried it, the mix went mouldy before fermentation had a chance to take hold.  I’ve had much more success with the first method and have repeated it enough times to get a consistent product. The best pears to use are juicy varieties with very tannic skins, like Williams (also known as Bartlett) and Comice. I mash up the pears – stalks, skins, cores and all – mix them with 25% – 50% of their weight in coarse sea salt (I don’t bother with the wine), and leave them at the back of the fridge in a glass jar with the lid only lightly screwed on. At the end of two months (unlike us, the Romans counted inclusively), the pulp has started to separate out. The heavier elements form a pale layer at the bottom of the jar, whilst the top part of the mixture is more liquid and is a pale pinkish-brown. When drained through a nylon sieve, the colour of the resulting liquid is a very pale version of the colour of fish sauce. 

I’ve tried various proportions of salt, and found that, if you use 50%, you seem to get more liquid, probably because the mixture doesn’t draw in moisture from the air to the same extent.  But a smaller percentage of salt allows more of the delightful pear flavour comes through – I find it much more difficult to detect in the 50% version. Stored in a clean bottle it will keep for months without refrigeration.

Figure 1: The pear liqumen is in the flask with dark blue trim

I’ve only had a problem once, when spots of mould had appeared on the surface of a batch six months after I’d made it. As I was due to give a Roman cookery demonstration in a few days’ time I had to quickly rustle up something I could use, so I cored and cut up a pear, boiled it with 25% salt and a little water, removed the peel and pulped the flesh in the blender. It was much too pale, but the taste was the same and I decided it would be a useful method for someone who couldn’t wait two months – in fact that’s what I recommend for my Roman Cookery School videos ( and

Say Ohm: Japanese Electric Bread and the Joy of Panko

By Nathan Hopson

In 1998, the New York Times introduced readers to an exotic new ingredient described as “a light, airy variety [of breadcrumb] worlds away from the acrid, herb-flecked, additive-laden bread crumbs in the supermarket,” with a texture “more like crushed cornflakes or potato chips” than its plebeian brethren (Fabricant 1998). That ingredient was panko, which has since become a staple for American home and professional cooks alike. In 2007, panko accounted for only 3% of US breadcrumb sales, for instance. Five years later, one in six American households regularly stocked panko in the pantry. Panko caught on because it is crunchier (and stays crunchy under restaurant heat lamps), absorbs less oil, and adds more volume than traditional breadcrumbs (Nassauer 2013).

Why are these Japanese breadcrumbs different? How did they get to be that way? The story told by American manufacturers such as LA-based Upper Crust Enterprises―an ironic name given that the secret to panko is crust-free bread―is that “Japanese soldiers during World War II discovered [that] crustless bread made for better breadcrumbs as they cooked it with electricity from tank batteries, not wanting to draw the enemy’s attention with smoke from a fire”(Nassauer 2013). Upper Crust’s president, Gary Kawaguchi, affirmed this account in a recent interview.

This is a cool story. Turns out, the truth is just as cool.

Japanese inventors had tinkered with electric cooking prototypes since at least the 1920s. Then in 1933, the Imperial Japanese Army (IJA) commissioned a “field kitchen that can prepare both rice and bread”(Aoki 2019, 11). Cost was no object and time was of the essence. As Katarzyna Cwiertka has noted, the military generally advocated bread, but there was a special urgency in light of the logistical difficulties of supplying rice to new front lines in Siberia and Manchuria. In 1937, paymaster captain Akutsu Shōzō’s design became the “Type 97” field kitchen, first deployed with the IJA’s First Independent Mixed Brigade that year (Uchida 2020, 2–4). The 97’s cooker was an insulated wooden box with electrode plates attached to the base and four sides of the interior. The highly efficient cooking process Akutsu used goes by several names, including ohmic and Joule heating. It is a form of electroconductive heating that passes electric current through foods to heat them rapidly and uniformly, quickly producing a light, yeasty, crust-free bread.

Figure 1: Type 97 field kitchen interior structure. Courtesy of JACAR.

After the war, companies such as Sony began selling rice cookers and bread machines that adapted these wartime technologies, and DIY home bread makers were showcased in magazines and newspapers. Influential women’s magazine Shufu no tomo and the inaugural issue of boys’ DIY magazine Shōnen kōsaku both featured instructions for bread makers derived from Akutsu’s design in 1946, reflecting the popularity of electric power in light of consumer fuel shortages and, conversely, excess generating capacity with military factories shut down (Uchida and Aoki 2019, 484).

In the 1960s, the new postwar frozen food industry hungered for high-quality breadcrumbs. Wheat had poured into Japan after 1945, the result of food aid; the use of bread and other wheat products in Japan’s school lunch program; and endless marketing promotions. Although ambitious American visions to recenter the national diet on wheat were soon abandoned, US agricultural imports and food technologies remained critical to Japan’s changing postwar food systems. Improved and upscaled food processing equipment met a market awash in cheap wheat, enthusiastic consumers (about half of whom owned electric refrigerators by the mid-1960s), and improved logistics. Frozen foods were among the shiny new things of postwar Japan’s shiny new “bright life,” and the mass use of frozen foods to cater the 1964 Olympiad and 1970 World’s Fair made them even more attractive symbols of Japan reborn.

These factors spurred rapid growth in breadcrumb demand, which was met in large part by the industrial-scale use of ohmic heating to create “electric breads” that were airy and uniform, and fried up crisply and uniformly when made into panko (Uchida and Aoki 2019, 485).

Sources Cited

Aoki Takashi. 2019. “Denkyokushiki chōri no hatsumei kara panko e tsuzuku rekishi oyobi saigen jikken.” Science Journal of Kanagawa University, no. 30 (June): 9–16.

Fabricant, Florence. 1998. “From Japan, the Secret of Crunchy Coating.” New York Times, December 1998.

Nassauer, Sarah. 2013. “Panko Tries to Find a Place in Every Pantry.” Dow Jones Institutional News, March 7, 2013.

Uchida Takashi. 2020. “Suihan o kigen to shi panko seizō ni tsuzuku denki pan no rekishi (1): Rikugun suiji jidōsha to Kōseishiki denki suihanki to Takara ohachi.” Tōkyō Yakka Daigaku kenkyū kiyō, no. 23 (March): 1–14.

Uchida Takashi, and Aoki Takashi. 2019. “Suihan, denki pan, pan seizō ni itaru Nihon no denkyokushiki chōri no rekishi: Rikugun ‘suiji jidōsha’ o kigen to suru denki pan jikken.” Nihon Yakugaku Kyōiku Gakkai ronbunshū, no. 43: 483–86.

This post is part five in an ongoing series by Hopson on the history of nutrition in modern Japan. You can read his previous post here.