BnF Ms. Fr. 640, fol. 98r: Varnish for lutes
Primary Author: Caroline Marris
French transcription:
<title id=”p098r_a1″>Vernis pour les luts</title>
<ab id=”p098r_b1″>Ils prenent un peu de tormentine & huile de tormentine ou daspic & de<lb/>
lambre pulverise & passe fort subtillem{ent} & font co{mm}e celuy de mastic & y<lb/>
adjoustent pour le colorer un peu de sang de dragon pour le faire<lb/>
rougeastre Et aultres de la terra merita pour iaulne </ab>
English translation:
<title id=”p098r_a1″>Varnish for lutes</title>
<ab id=”p098r_b1″>They take a little turpentine and some turpentine oil, or spike lavender [oil], and some
amber pulverized and very finely sieved & proceed as with that of mastic, & they add some dragon’s blood to
color it and turn it red. And others some terra merita[633] to turn it yellow.</ab>
Twenty-first-century stringed instruments, often built using modern industrial techniques and not in an artisan’s workshop, are prized for their brilliance of tone and volume. In the early modern period, however, instruments were not only quieter and employed different strings and performance technique, but were invariably made in workshops by skilled craftsmen. The recipe for a “varnish for lutes” in the BnF Ms. Fr. 640 book of secrets is an interesting inclusion in a manuscript otherwise mostly preoccupied with the arts of metallurgy, painting, and medicine (though not, it should be noted, unique in this sense – the manuscript incorporates several stranger and more varied instructions for, among other things, the capturing of nightingales on fol. 105v or varnishing painted boxes and painting taffeta banners on fol. 67r). That the varnish initially described in the recipe above is for a red varnish is even more intriguing, as it is not a common color for instruments, and is involved with strong cultural connotations and technical secrets.1 In the late twentieth century, for example, the 1998 film The Red Violin began in the seventeenth century with the eponymous instrument being painted with a varnish turned red by the blood of the artisan’s, an act which carries its sins forward several hundred years. In technical terms, modern luthiers insist that the composition of a varnish can totally change the sound of an instrument – the German master Sourène Arakélian, for instance, wrote in 1981 that “The student violin of today, factory-made, with dry hard varnish, can only give out strident and sour sounds…The function of the varnish is not only to preserve the instrument against humidity and manual perspiration, but also to improve its tone”.2 Though Arakélian says the influence of varnish alone has been exaggerated and that its qualities work in tandem with those of the wood of the instrument itself to produce tone, color is still an important aesthetic decision to be made by the luthier, ending with those “which suit his taste.” Intriguingly, a violin left unvarnished is described by Arakélian as “white,” which color apparently has its own aural qualities.3
Our goals for this annotation were to better understand the context and skills necessary to the making or decorating of a musical instrument in our period, and also to open a discussion on the nature and experience of color in early modern artisanal practice. While our experiment protocols and use of certain ingredients clarified the former, the latter is still a topic very much up for debate.
Setting the Scene: Varnishes and the History of Musical Instruments
A European or Near-Eastern lute is a plucked stringed instrument with a deep, rounded back and wide neck. Its origins are classical, and archival sources indicate that it was established in Europe (Spain) no later than the ninth century, with the first known written music for the instrument appearing around 1500 (Fig. 1 – lute by Sixtus Rauchwolff, 1596).4 It shares features with several other instruments which have survived to the current day (albeit often in highly-specialized performance environments) including the oud, the theorbo, the archlute, the kobza, and so on. A fantastic variety of sizes, shapes, and string numbers is obvious in any collection of pre-‘high classical’ (pre-Mozartian or pre-early nineteenth-century) instruments; it is possible to find lutes of the period of 1600-1800 with anywhere from 6 to 12 strings in a variety of tunings.5 The lute was also a common subject of contemporary paintings, often as a vanitas object, or engravings to do with the instrument-maker’s craft. Titian, for example, included a lute-player in a painting of Venus dated 1565-70 (Fig. 2 – Titian, ‘Venus and the Lute Player’) ; and Albrecht Dürer engraved a ‘draughtsman’ at work on constructing a perspective drawing of a lute (Fig. 3 – Dürer, ‘The Draughtsman of the Lute’). Determining the social status of instrument-makers and musicians in the 16th century is often difficult beyond the biographical details of certain master composers, most of whom were artists working in a similar system of support to painters, being attached to courts or nobles (this system would persist for at least a hundred and fifty years after the probable date of our manuscript – J.S. Bach, for instance, b.1685-d.1750, spent most of his life in the service of individual or civic patrons). It is, however, possible to find evidence of the look and habits of lower- to middle-class jobbing musicians for hire in early modern engravings, especially in Germany (see Fig. 4 – Meckenem, ‘Harpist and Lute Player’ ; and Fig. 5 – Aldegrever, ‘Two Musicians Playing the Violin and the Lute’).
There are dozens of mentions of various ‘varnishes’ in BnF Ms. Fr. 640. The vast majority of these are varnishes meant for the surface of paintings, as in several listed on fol. 4, fol. 31, fol. 57, fol. 60, and so on.6 They can also be applied to objects other than paintings, such as “boxes covered with painted paper” (fol. 67), engraved iron (fol. 4), and sword hilts (fol. 96). They seem to function as both protection, in the case of their completely covering and ‘finishing’ an object, and as a conscious aesthetic choice on the part of the artisan. Varnishes are likewise common material in other contemporary sources, but neither Biringuccio nor Cennini, two of our other main comparative artisanal writings, mention the treatment or art of decorating musical instruments; Hugh Plat’s Jewell House (1594) instructs only how to make an “Oil or varnish made to dry speedily”. Likewise, general historical works on the development of musical instruments in the period, though often concerned with the physical form of stringed instruments including the viola da gamba family, the lyra da braccia, the violin, and the crwth, have traditionally left out the question of how a varnish might play into the production of music, either aesthetically or scientifically.7
More recently, however, there have appeared works which attempt to replicate the same process of historical reconstruction as we are engaged in with the Making and Knowing Project. Experiments have been written up and published, for example, on the stripping and examination of “the earliest surviving set of English string instruments” by a craftsman named William Baker.8 A reconstruction of a lute from scratch using contemporary illustrations and instructions likewise has much to say about the preparation and carving of wood, but unfortunately little to say about varnishes other than that egg wash might have been used to give a “blond” finish to an instrument.9 Elsewhere, there are brief mentions of the history of “cheap” varnishes being preferable to expensive ones in the making of violins in early eighteenth-century Italy, resulting in the conclusion that certain imported varnishes may have “damp[ened] out the upper harmonies” or that it was the use of a particular type of varnish which might have produced that ever-famous ‘Stradivarius’ sound.10 Despite these mentions, it is far more common to see varnishes for paintings described in art historical literature. These do, however, carry a common theme: that a varnish’s composition can fundamentally alter the aesthetic of a work of art, and, if one is to extend that power to musical instruments, an instrument’s sound as well.11
Our Recipe, fol. 98r
Many of the manuscript’s varnish recipes – including those meant for painting canvases or nearly any other object apart from lutes – have overlapping or complementary ingredients and techniques, such as heating the base liquid, and the use of mastic, amber, turpentine, or aspic oil. It was fairly easy to determine the base ingredients of our own recipe: turpentine oil or spike lavender oil, ground amber, and dragon’s blood resin (which could be obtained from a number of different plant genuses including Croton, Dracaena, Daemonorops, Calamus rotang or Pterocarpus). ‘Terra merita’ was more difficult to pin down, but appeared to be equated by modern art suppliers with turmeric powder. Our final initial step for our ingredients was to determine what type of wood we wanted to paint our varnishes onto. We searched the online collection of the Metropolitan Museum of Art and were able to pull up a list of 96 lutes held by the Museum which were a) of European origin and b) were built between 1600 and 1800; even a cursory examination of their details showed us that by far the most common wood out of which they were made was spruce.12 We therefore obtained spruce ‘patching boards’ from a site catering to guitar luthiers to work with.
In establishing our experiment protocol, we found that a recipe on fol. 71v was most helpful in laying out a set of steps with which to tackle fol. 98r. This recipe stated:
Add 4 [drams] of sandarac and finely pulverized mastic to a half lb of lavender spike oil. First of all, boil your oil in a pot on a stove, and then add the aforesaid gums little by little, stirring continually with a small stick split and quartered at the end.
Combining this with our initial recipe, we determined that we should heat our base liquid and ground amber together. Another varnish recipe on fol. 73v provided us with a more concrete ratio of ingredients to test (“two ounces of aspic oil and one ounce of sandarac”), i.e. two parts liquid to one part amber. Some recipes mentioned merely that this liquid combination should be ‘heated,’ while others mentioned ‘boiling’ in translation (for example, ‘chaufer’ in fol. 4r_b1a, ‘Aspic oil varnish,’ vs. ‘bouille’ in fol. 4v_c1a, ‘For an excellent black varnish’). Given that our original recipe did not mention specific shades of color we decided to take our heated mixtures off the heat and stir in however much coloring agent (turmeric powder or powdered dragon’s blood resin) we felt gave the varnishes ‘enough’ color.13 Given that we had two liquid bases and two coloring agents to try, we ended up running our experiment four times:
- Turpentine oil and turmeric powder → yellow
- Spike lavender oil and turmeric powder → yellow
- Turpentine oil and dragon’s blood → red
- Spike lavender oil and dragon’s blood → red
Our first step was to grind our solid ingredients as fine as we could; while our dragon’s blood resin crumbled quickly and easily and did not need to be sieved to appear very fine, (Fig. 6 – grinding dragon’s blood resin) grinding our amber pieces was very hard work and never quite reached a truly fine state despite being sieved several times; there were always larger, coarser pieces visible in the final powdered form (Fig. 7 – grinding amber; and Fig. 8 – powdered and sieved amber). Our turmeric powder was a commercial foodstuff. We then heated a hot plate under a fume hood (to help deal with the strong smell of our oils) to what we thought would be a boiling temperature and started our first experiment with turpentine oil. We combined 3 ounces of turpentine oil with 1.5 ounces of ground amber and stirred it continuously with a handmade tool of iron wire, with the hot plate set on a setting of 4 out of 5. Almost immediately, the turpentine oil – which has a boiling point of ~307°F14 – began to smoke and boil. The amber also immediately started to solidify in a dark brown sludge in the turpentine, turning the whole a very unpleasant color. Within moments, the clearly too-high heat had boiled away the turpentine and left us with a crusty mess of foul-smelling amber dust. An acknowledged failure! (Fig. 9 – failed experiment: congealed oil-amber mixture)
For our following attempts, we decided to keep our hot plate on its ‘minimum’ setting and to monitor the temperature of our liquids with an infrared thermometer. Our second attempt with our first mixture rose rapidly to a temperature of about 125°F, at which point we poured in our 1.5 ounces of amber and began to stir it; its temperature eventually stabilized at between 140 and 150°F. Though we stirred the mixture for several minutes, the amber did not seem to be incorporating in any way into the oil, and did not change its sludgy brown-black color (Fig. 10 – second attempt at turpentine oil and amber mixture). We therefore decided to strain the mixture before calling it completed. After more than five minutes of no change, we took the varnish off of the heat and started to stir in our turmeric, stopping when we felt we had a healthy yellow color. We then strained the varnish through cheesecloth to remove most of the amber pieces, and stored it in a glass jar (Fig. 11 – strained ground amber).
The remaining three experiments proceeded almost exactly as above, with the only difference being the variations in color produced by similar amounts of coloring agent (our two red varnishes were stored in tin cans rather than in glass jars). Each was heated to approximately 150°F and stirred for several minutes; colored; and strained. Our results were as follows:
- Turpentine oil and turmeric (5/8ths teaspoon): muddy, rich brown-yellow.(Fig. 12 – turpentine oil and turmeric varnish.)
- Lavender spike oil and turmeric (4/8ths teaspoon): bright, nearly neon yellow.(Fig. 13 – lavender spike oil and turmeric varnish.)
- Turpentine oil and dragon’s blood (6/8ths teaspoon): muddy, dark red.(Fig. 14 – turpentine oil and dragon’s blood varnish.)
- Lavender spike oil and dragon’s blood (6/8ths teaspoon): brighter red; this was also the most liquid of all four, which was heated for the shortest amount of time and only barely broke 140°F – perhaps low heat, rather than boiling, was always the key?(Fig. 15 – lavender spike oil and dragon’s blood varnish.)
On the advice of luthier Gabriela Guadalajara, we looked through the manuscript for clues as to what might be used as a sealant for the instrument’s wood before the varnish was applied to it.15 Ms Fr 640 mentions ‘glue’ dozens of times, again frequently in reference to painting, but has very few actual recipes for it beyond that of fish glue on fol. 159r. We therefore looked for its synonyms and focused briefly on mastic, which is a resin obtained from the mastic tree (Pistacia lentiscus) commonly known as ‘arabic gum’ – not to be confused with ‘gum arabic’ or acacia gum (Sengalia senegal or Vachellia seyal).
Due to time and price concerns, however, we decided to proceed with sealing our sample piece of spruce wood with liquid gum arabic instead, which was already present in our lab and has application in the thickening of inks and paints.16 We also found in fol. 3v that the manuscript’s author-practitioner recommended that although with “Lavender varnish…it is necessary to lay one coat of the said glue…and allow to dry, then varnish,” in fact “Turpentine varnish does not need any glue since it is fat and viscous and does not penetrate the wood as the spike lavender and sandarac varnishes would”. We decided to proceed with testing all four varnishes on one piece of sealed spruce, followed by testing all four on a piece of unsealed spruce. (This would not be modern practice, as Guadalajara told us one would never now proceed with varnishing without sealing beforehand.)
Each piece of spruce was sanded briefly to take off a slightly rough top layer before one was painted with one layer of gum arabic and left to dry for half an hour (Fig. 16 – prepared spruce patching woods). After labeling, we then painted each of our four colors in straight lines on both spruce pieces. To our disappointment, it became clear immediately that although our ‘red’ varnishes looked quite red to our eyes where they lay in our jars, they were barely brown when used to paint, and indeed were practically invisible on the piece of sealed spruce. The yellow varnishes showed color much more obviously, but were still rather weak. By far the strongest color was that of the lavender spike oil & turmeric mix, which, as we had seen in its mixing, painted as near-neon on the wood. All four varnishes spread quickly and widely across both pieces of wood, despite only being painted in narrow strips. In fact, the varnishes spread even further on the sealed piece than the non-sealed piece; we suspect this may have been due to the fact that gum arabic is water-soluble, and the sealant may not have been completely dry (Fig. 17 – spruce woods after one coat of varnish).
Three coats of each varnish were painted, one on top of the other, to observe how the colors might change or accumulate with repeated strengthening coats (each given approximately 30 minutes to dry).17 Only the neon lavender & turmeric varnish showed strongly on the sealed spruce. On the unsealed spruce, the neon mixture was practically garish, but the turpentine & turmeric, turpentine & dragon’s blood, & lavender & dragon’s blood varnishes began to display more pleasing traits. Though not red, the two dragon’s blood varnishes eventually approximated a typical brown one might see on many stringed instruments today. Likewise, the turpentine & turmeric mixture created a far more palatable golden-yellow shade which would not look too out of place on a lute or oud. All four varnishes continued to spread rapidly across the wood despite being painted using very little liquid, but the author’s point about turpentine oil taking better to unsealed wood seemed well-taken, as it both displayed better color and spread less than the lavender spike oil varnishes on that untreated piece of spruce (Fig. 18 – unsealed spruce with three coats of varnishes; and Fig. 19 – sealed spruce with three coats of varnishes).
Questions and Suggestions:
There are a host of questions which this annotation could attempt to answer, but in the end it seems our work has touched upon them only in fits and starts. Our first major question was about the workshop context of building, decorating, and perhaps maintaining early modern stringed instruments. The overlap of materials between the multiple varnish recipes in the manuscript suggest that a painter, used to preparing and finishing canvases, would be equal to the task of varnishing an instrument with the same resources at hand, and vice versa for a luthier – at least when it came to our liquid materials. Ground amber and dragon’s blood resin may have proved a greater challenge to obtain, at greater expense, and indicate perhaps that a lute varnish was a substance to be prized and praised for its color alone, rather than for its protective capabilities or its potential to modify a lute’s sound. The procedure for making the varnishes, assuming we designed our protocols correctly, was also simple enough that, provided the materials were available, it could be made quickly and easily, and not require overly specialized knowledge that would be out of reach of a trained luthier or even a travelling musician.
When it comes to our second major inquiry about color, we have tried to answer the question of what steps must be taken to produce colors suited to the decoration of an instrument. Unfortunately, from there, all breaks down in subjectivity: what, after all, must we discern about a particular color to determine if it is aesthetically pleasing on the body of an instrument? If the experience of color is also a subjective experience – both physically and emotionally – how are we to decide what shades are appropriate for any given situation, or indeed whether we have even approximated the shade intended to be made as described by the author-practitioner? In the end, we definitely discovered that our liquid varnishes, though they may appear to be strongly-colored, immediately weakened in brilliance upon application. This may shed light on the processes involved in instrument-making and the distillation of color, but whether we can define anything we did as a success or a failure is completely open to debate.
- Continue to look into the possibilities for sealants and consider using hide glue, though it is not mentioned in conjunction with wood/instruments in the manuscript (or indeed at all, though a more careful search should be made; mentions of ‘colle forte’ in the manuscript in particular may refer not to a glue’s strength, but in fact to a particular type of glue).
- Consider including less ground amber to improve color, or experiment with making sure it is much more finely powdered and see whether that improves its incorporation into the heated liquid of choice.
- Use much more dragon’s blood to try to strengthen the ‘red’ varnishes!
- Continue investigating nature of ‘terra merita.’
1 This statement is partly based on the author’s own experience as a string player, and on a sampling of early modern instruments housed by the Metropolitan Museum of Art (see pg. 4), which mostly range in color from brightly golden-yellow to shades of rich, dark brown. NB that these observations are necessarily subjective (see final section of this annotation for more on color and subjectivity).
2 Sourène Arakélian, The Violin: Perceptions and Observations of a Luthier – My Varnish, Based on Myrrh trans. Marie Arakélian and Peter Armitage (Frankfurt am Main: Das Musikinstrument, 1981), 41.
3 Arakélian, The Violin, 42.
4 See Gary R. Boye, “A History of the Lute from Antiquity to the Renaissance by Douglas Alton Smith, Review,” Notes 59.4 (2003): 905-907.
5 See note 11 for collection sources.
6 Interestingly, Cennini, in his brief section on varnishes for paintings, writes that one should hold off on varnishing for as long as possible before applying it as it will ruin the color and “freshness” of a work of art if applied right away once made/the painting is ready. Clearly, however, this does not apply to the treatment of instruments, though the question of time and whether the wood would need time to adjust to its environment is a good one. Early instruments, and indeed modern instruments made in old styles, are frequently more apt to warp, crack, or otherwise react badly to their environments even when fully varnished and properly maintained. Cennino d’Andrea Cennini, The Craftsman’s Handbook: The Italian “Il Libro Dell’Arte” trans. Daniel V. Thompson Jr. (New York: Dover Publications, 1933), 98-99.
7 See, for example, Gerald R. Hayes’ Musical Instruments and their Music, 1500-1750 (Oxford: Oxford University Press, 1930), which, though highly prescient in its recognition that it was necessary to re-acquire and re-cover early modern playing techniques, restricted his exploration of the development of these instruments to their repertoire and playing technique rather than their form.
8 Peter Trevelyan, A quartet of string instruments by William Baker of Oxford c. 1645-1685 (Oxford: Galpin Society, 1996).
9 Andrew Atkinson, “Building a Renaissance Lute using original methods,” The Lute Society, http://www.lutesociety.org/pages/building-lute-original-methods; accessed May 6th, 2015.
10 Frank Della Torre, “The Rediscovery of a Lost Art and a Few Notes on the Theory of the Violin,” Science 27.693 (1908): 592-593.
11 See as just one example E. Rene de la Rie’s “The Influence of Varnishes on the Appearance of Paintings,” Studies in Conservation 31: 1 (1987): 1-13.
12 Metropolitan Museum of Art, “Online Collections,” http://www.metmuseum.org/collection/the-collection-online/search?&where=Europe&ft=lute&what=Lutes&when=A.D.+1600-1800&pg=1 A search of the V&A’s collections produces similar results when one searches for ‘lute’ and then narrows the objects to ‘musical instruments.’ http://collections.vam.ac.uk/
13 This brought up an intriguing question – one far too large to try to answer in this annotation – about how, exactly, the author-practitioner could possibly have described an exact shade of color as he experienced it to his reader or apprentice in such a way that a follower of his work could then exactly reproduce it. Since so much of the experiencing of color is subjective, can words alone ever do a particular shade justice?
14 MSDS for turpentine oil, Hazardous Material Information System, IntraWEB LLC.
15 Guadalajara is a New York-based luthier who builds her own instruments and maintains both modern and early stringed instruments. She studied violin-making in her native Mexico at the studio of Luthfi Becker as well as with the Tri-State’s premier luthier for gambas, William Monical, before his retirement. Very interestingly for us, she told us that the composition or creation of varnishes had never been a part of her own lengthy training, and in her experience the application of it was likewise something learned instinctively rather than having any hard/fast or teachable rules. http://www.gabrielasbaroque.com/
16 Unfortunately we only heard from Guadalajara after the experiment had been completed that water-soluble sealants are not recommended for instrumental work; she recommended ‘hide’ glue, which is made from animal bones and tissues. NB: mastic is also water-soluble, so even if we had indeed made the effort to procure and use it our results may have remained the same.
17 A recipe for red vermilion varnish on fol. 74v describes applying 2-3 coats, and another recipe on fol. 77v for ‘Other varnish’ recommends 3.