SUMMARY
Prepared the canvas using glue for sizing combined with a small amount of champagne chalk. Applied two layers of glue and chalk. Applied a final layer of size only.
Longer field notes
First preparation on my canvas:
060r_1 First whitening of the painting
[<id>p060r_1</id>
<head>Blanchiment premier du tableau</head>
<ab>On couche par deulx ou trois fois de la <m>croye</m> destrempée à <m>colle<lb/>
de destrempe</m> sur le tableau, non avecq un pinceau mays avecq une broisse,<lb/>
en telle sorte comme si tu voulois poncer, & laisse seicher. Et<lb/>
reitere jusques à deulx ou trois fois, puys avecq un cousteau<lb/>
unis bien la derniere assiete, <add>puys donne dessus une main de <m>colle</m></add>, sur laquelle tu pourras aprés<lb/>
faire ton impression puys paindre. Mays garde que<lb/>
ton premier <m>blanc</m> ne soict pas trop espés, car il s’esclatteroit<lb/>
volontiers. Les painctres flamens font faire à douzaine tels tableaux.</ab></div> ]
I am working on the assumption that ‘glue’ is one option of a medium for distemper. This would most likely have been hide glue or glue made from parchment clippings but instead I am using rabbit skin. This is still glue made from animal skin and therefore I do not think it ought to make too much difference. It is easier to acquire and is a compromise I am making in order to proceed.
‘destrempee a colle de destrempe’ – mix with distemper glue.
I am heating rabbit skin glue to 60 degrees c.
Heating water for a ban marie for mixture.
Mixing with Champagne chalk. Quantity not given – but enough to be white after 3 layers, yet not too thick.
I used approximately 1tbsp of champagne chalk with 40ml rabbit skin glue, heated to approx. 50 degrees. I mixed them together with a metal spatula. I found it quite thick and difficult to apply evenly. I became concerned about not having put down a separate size layer first, but the recipe did not mention a glue only layer so I proceeded with this. I used a large brush.
I am now leaving it to dry. It does seem to be functioning as a size layer, yet is whiteish. The next layer might want a little more chalk to glue.
For next layer, I remembered to wet the brush as the Ms suggests that wetting brushes 'couleurs a destrempe' spread more easily.
The ms. says leave it to dry between each layer.
I left it to dry.
It went from white coloured to transparent.
In order to try and achieve a white colour, as the Ms seems to indicate, I put in more chalk to the second layer. It still dried transparent.
Finally I brushed on a layer of glue only and left it to dry.
Monday 4th April
I used a print out of the face of the Madonna in Dieric Bouts' Annunciation as a 'cartoon'. I rubbed charcoal powder on the reverse and traced through with a sharp pencil as we had done in our skill building exercises. This method, of tracing with carbon through a cartoon, has a long history (...). I traced four to give myself the opportunity to try out different possible media for 'destrempe' - glue, gum, and egg white.
Wednesday 6th April
SUMMARY
4 'ombre' layers of portraits done on canvas prepared with size and chalk.
1) - Highly concentrated animal glue as medium. Used warm. Only painted where there would be shadows in the final work. Dampened back with wet sponge.
Results: gritty paint, coagulated quickly.
2) - Highly concentrated animal glue as medium. Used warm. Painted dark layer all over the face and hair. Did not dampen back with sponge.
Results: some grittiness of paint. Paint still coagulated quite quickly but I was slightly more used to it and so able to control it better. As the glue dried, the canvas buckled.
3) - Less concentrated animal glue as medium. Used warm. Only painted where there would be shadows in the final work. Dampened back with wet sponge.
Results: still some grittiness of paint. Paint coagulated and dried less quickly than in the other two examples and was therefore easier to use.
4) - Windsor and Newton preprepared gum arabic with water. The concentration was already chosen for me by the manufacturer, I added a little water while working but the medium was already very easy to use. It did not dry out, or clump and it was possible to achieve both thin washes and darker passages by manipulating the proportion of pigment to medium. I am used to using watercolour of this kind and think this is the main reason that it was more successful. That, and the pre-set concentration determined by experience paint manufacturers.
Longer Field Notes
Returning to the canvas. I had prepared it with glue and chalk, and traced the outlines of the faces using a 'cartoon' and charcoal. I had then 'sealed' them with glue or gum, so that the charcoal would not run into the painting later. Whereas when I had just finished the lines of glue were very raised in raking light, after 3 days of drying they have gone down considerably and are almost, though not quite, flat.
I wanted to first try distemper with glue. I used rabbit skin glue, 1:15.
But this has been heated and reheated so many times that I doubt it is really such a low concentration. It seems very concentrated.
First I ground the combination of pigments suggested for woman's shadows in 093v. Bistre, brown ochre (in place of 'ochre de run) and sap green. I put them together according to estimates of quantities. There was no particular detail given except that the quantity of sap green should be less than the bistre or brown ochre.
As the Ms suggests the bistre really does make a very gritty sound on the slab. Is it difficult to grind? This is from a pigment bought from Kremer that is already very finely ground, but I think it is still more gritty sounding than the other pigments. Also very dominating hue and not as thirsty for water as I expected.
Once I wet the back of the canvas with a sponge, as the Ms directs, I realised that it does not work to dampen the front of the canvas. At all. The moisture does not penetrate the canvas. This is because I sized it. I had feared this would happen, but still thought to size the canvas because the thought of working on an unprepared canvas seemed so strange to me.
I had hoped that dampening the back would affect the way that the glue flowed as I painted it on the front, but since it did not get through to the front of the canvas it did not. I did try to use the glue and pigment without warming it, as the Ms did not suggest heating, but it was far too glutinous and could not work as a paint. The flow of the distemper was really impossible. So then I started to use it all warm, making sure to put my glue and pigment mixture into a bain marie of very hot water. I also found myself putting the paint onto the palm of my hand to mix it further before applying to the canvas. Interestingly this is the same instruction given by Jean le Begue for using glue based paint - to use the warmth of your hand. The paint did flow onto the sized canvas when it was warm, but not easily, and I still found it often clumped and the affect on the canvas was not smooth and paint like. I was aiming to achieve a thin wash of shadow on the first face, something like the undermodelling visible in the unfinished Tuchlein attributed to Jorg Breu the Younger, but a little darker. But in fact the glue became thick and dried quickly on the canvas leading to clumps, and despite grinding the pigment it still looked gritty on the canvas.
So, the next one I tried to paint glue on the front, warm, but without wetting the back. To see whether the wetting of the canvas made any difference. It did not seem to make any difference to the flow of my paint. However, as the two dried, the one where I did not wet the back of the canvas buckled – because the drying glue pulled the canvas. Whereas on the quarter where I did wet the back, it seems to be taking longer to dry and not to be bucking.
In this second case I also painted over the entire face with the 'shading' pigment, more like a Verdaccio layer perhaps, to see how this would work in the layering system suggested by the author-practitioner.
I became convinced that at least one of the problems was the concentration of the glue I was using. So I took some glue that said it was 10:1 rather than 15:1, that is, more concentrated, but I could tell to look at it that it was less, because it was paler in colour. I do not know precisely what concentration of glue to water this mixture was, but when I heated it it flowed better and seemed more dilute. So I broke my protocol, in which I had planned to do just two different 'glue' faces, and instead did a third. I used the less concentrated glue and kept all of it warm while painting in the bain marie system. I put down the dark paint only where there would eventually be some shadowing on the face, much like an undermodelling layer. This was the most successful of the three tried so far, although I still did not have brilliant control of my medium.
At this point, I paused, and considered that I should never have prepared the canvas at all. I should have followed 093v as given, and then perhaps the water from the sponge on the back of the canvas would have worked to dampen the front. So I took a small piece of unprepared linen and tried painting onto it with warmed glue and pigment and cold glue and pigment. I found that both were more successful than on the sized prepared canvas. Additionally, the bistre really did 'stain' the canvas like the Ms said it would when I used an unprepared one.
So, I want to try following the instructions on 093v instructions without preparing the canvas with any size layer at all and see whether it works.
Additional thoughts during the process - trying to use a totally new paint medium following the instructions in the text feels very strange. How commonly did people in the past actually use painting manuals to learn the basics of a craft? I felt as though I would be able to learn much faster by following the example of someone who knew what they were doing with distemper paint. When you watch someone work, you see them make a lot of tiny 'micro-decisions' and that information is conveyed to you very quickly. I also felt very frustrated with my own limitations in handling the medium, particularly in relation to its drying time.
Annotation field notes 11th April 2016-04-10 Time 10 am – 2pm.
Summary - Painted with various media options on unprepared canvas. This was most successful with animal skin glue. This was more successful than it had been painting in glue on the sized canvas.
Longer field notes
Tried using glue and gum Arabic medium on canvas that has not been prepared in any way.
I stretched the linen canvas over a stretcher, making it as tight as possible. It is not dyed linen nor sized at all. I first tried using glue. I chose to use the glue at a lower concentration (said 1:10, see previous) as that had been easier to handle last time. I kept it warm in a bain marie. I soaked the back of the canvas in luke-warm water just before I started to paint. This time the water did penetrate to the front. I felt as though this did have a beneficial effect on my ability to paint with the glue and pigment mixture, that it helped with the flow because the glue did not seem to dry out as quickly.
The effect was that the bistre –sap green- ochre de rux combination did stain the canvas. Much more as the manuscript described. I think it will work better for painting on with ‘carnation’ layers.
Still though, my lack of experience means that the medium still clumped at times. I also tried doing the same process but with gum Arabic. It flowed on easily and was quick to use. But, interestingly, as it dried the gum Arabic spread out within the water and made a vague, very soft effect. The gum Arabic medium clearly ran into the water and extended, leaving water marks on the canvas.
Also, regarding the darkness of the colour – both the Ms and De Mayerne make a point about wetting the back with the sponge and both note that with distemper the colour changes between when it is wet and when it is dry. They have the idea what working with the colours damp prevents you from being ‘deceived’. But for me, it is difficult to judge how dark the final results will be because I apply them wet. So what do they mean by avoiding deception? Why is the true colour not the light one? I tried leaving a small section of the gum Arabic one without wetting the back to see how the result was different. I obviously made it darker from the beginning, because when they all dried that section was darker.
Painting on the 'Carnation' layer
I began to apply the layer above the shadow base layer.
According to the instructions at 093v:
'Apres on pose la carnation qui nempesche point que les<lb/> ombres desja faicts ne parisien'. This has been translated as:
'Then one lays the flesh tones, which does not prevent the already painted shading/shadowing/shadows from showing [through].'
At 093v there are no specific instructions for what combination of pigments ought to be used for this.
I looked elsewhere in the manuscript for places where the author-practitioner discusses colours for skin.
At 058r he suggests using Florence lake for a 'beautiful flesh colour':
<ab><m><pl>Florence</pl> lake</m> is better than <sup><m>lake</m></sup> from <pl>Flanders</pl> because in <pl>Florence</pl> they make the best <m>dyes</m>. To make a beautiful flesh color, the reddest and brightest <m>lake</m> is the best, because those that are purple and violet as a result of the addition of too much <m>alum</m> look like the flesh of someone that is very cold. That is why women who want to color their cheeks crush <m><pl>Florence</pl> lake</m> very finely, then fill a little <m>cotton</m> with it, which they then wrap in a little fine <m>cambric</m>. And thus they rub the lake on their cheeks and then, with another clean <m>cotton</m>, they soften it.</ab>
Again at 061r_1 the author-practitioner suggests Florence lake for skin colours
<id>p061r_1</id> <head>Skin colors</head> <ab>It is necessary that you make it of two kinds, one more red to make the main layer, the other more pale for the highlights, as around the eyes. And then on this last skin color, you will touch lightly the main lights with a little <m>white lead</m>. But avoid applying too much of it because this will look like a face of death. The beautiful <m><pl>Florence</pl> lake</m> makes a beautiful vivid skin color that resembles the complexion of <m>alexandrine pink</m> and </m>incarnadine</m>.</ab>
From these, I deduced that a bright red ('reddest and brightest') lake was one that he considered good for flesh tones. Since the author-practitioner indicated that a purple tone would make someone look cold, I considered that he preferred a warmer red tint rather than a slightly cooler, or more purple, pink colour. The madder lake produced in the laboratory (ask for further details) is a red, warm colour, and I selected this for use in these flesh tones. I also deduced that some lead white would be used later for the highlights.
At 062r_2 the author practitioner wrote:
<id>p062r_2</id> <head><m>Ocher</m></head> <ab>This is used for faces, hair, skulls and rocks.</ab>
Just above an instruction:
<id>p062r_3</id> <head>Distant people and animals</head> <ab>First one roughs them out in gray or in purple, which is made of <m>azure</m> ash and <m>lake</m>. Once dry, one highlights and finishes with flesh tones and other colors and white. And it will look better and is quicker to do than in black and white. Armies are painted the same way.</ab>
Again at 064r_3 the author-practitioner reinforces the need for ochre in flesh tones:
<id>p064r_6</id> <margin>left-top</margin> <note> <head><m>Yellow ochre</m></head> You need a bit in every flesh tone.</note>
From these I decided that I would need some ochre, and could use yellow ochre specifically, and madder lake.
I had to make a decision whether to use lead white in this carnation layer, or whether to use chalk for the white colour.
Although chalk becomes too transparent to use in oil, it can work as a relatively opaque layer in animal skin glue, and it was used as a white in tuchlein paintings.
I therefore decided to use chalk for this layer and lead white only in the highlights.
I ground champagne chalk, madder lake, and yellow ochre together. I decided on the proportions of pigments to use by eye, bearing in mind the fact that I knew there would be a colour change as the distemper dried. The colours look darker when they are applied and then dry lighter. (This colour change is discussed in De Mayerne manuscript - translation: ).
I combined the pigments with the warm rabbit skin glue in 1:10 approximately, with extra water to dilute it as required into a usable paint. I added the pigments to the glue sufficiently to create a workable paint, this was not done by proportion but feel and eye. This seemed appropriate as this would have been the method most likely used by artisans, and because no proportions or even indications of the ratio of pigment to binding medium are given in the manuscript. But an alternative approach would have been to use the ratios suggested in conservation literature. (details).
I wet the back of the canvas with a sponge as before and painted on the layer. When it was wet, the layer as though it would cover the shadows too opaquely, preventing them from showing through. But as it dried, the shadows beneath showed through beautifully where they had been painted on. It seems the system is working.
Wednesday 27th April
Summary: Painting the 'vive' touches on the face. Carefully following the instructions, the painting is 'coming alive'. The process increasingly seems to match secondary source accounts of distemper painting in the 14th and 15th centuries. I feel more and more certain that animal glue is the 'right' binder for 093v, or at least one that works well in the way the author-practitioner describes. I am also gaining more mastery over the glue through experience of handling it, and it no longer seems such a counter-intuitive medium to work with.
Longer field notes:
Added lead white, madder lake, and yellow ochre in combination with animal skin glue for the highlights on the face. I found myself putting quite a lot of this down. Did I risk putting too much white and making a a face ‘of death’ as author practitioner suggests at 061?
‘And one touches again the more prominent areas with brighter flesh tones’ is the approximate translation.
However looking back at the French, the manuscript actually says: ‘Et encores dune carnation<lb/>
plus vive on touche les endroits plus releves’
I would like to ask for further translation help because I think it is more interesting to consider the ‘plus vive’ as more ‘alive’.
As I am becoming more accustomed to using rabbit skin glue as a medium, I find it easier to manipulate, end even to ‘blend’ or ‘soften’ the edges between the flesh tones and shadows to some extent. I wonder whether the author-practitioner would want me to do this? Since in the final shadow layer he suggests hatching for shadows. Elsewhere in the manuscript he refers to hatching as the Italian method. Is hatching an alternative system to ‘softening’, with the former being used in tempera/ distemper and the latter only with oil? This is a question to investigate in early treatises.
The above is answered by a description of glue painting by Dunkerton in Giotto to Durer.
The description of glue painting given almost seems to describe 093v exactly: 'Glue size painting
The most straightforward of the easel-painting techniques of the fourteenth and fifteenth centuries was using a medium of glue size or distemper on canvas. The pigments (previously ground in water) were simply combined with the same animal skin glue used to size canvases or panels and to bind gesso or chalk grounds. Because the technique did not involve the use of a ground the paint seems to have partially soaked into the canvas, the pigment particles becoming to some extent incorporated into the fibres of the fabric. This can lead to the appearance of an image on the reverse of the canvas, sometimes visible even when the painting has been lined. The retention of moisture by the canvas allowed a certain amount of blending of the brushwork, enabling the painter to achieve smooth transitions in the modelling. The placing of wool cloths beneath the canvas to absorb and retain moisture during the painting would have furthered this purpose, as well as preventing the bleeding of the colours into one another. This practice is described in the collection of manuscripts on painting techniques compiled and revised by Jehan Alcherius from 1398 to 1411, and apparently taken from cloth painters in London. A legal decision of 1458 resolving a dispute between the panel painters and the cloth painters of Bruges… actually stipulated that the latter should only work on the linen ‘while wet in the threads’. However, any subsequent application of paint, including details such as facial features and strands of hair or superimposed patterns on textiles and marbled stone, would have had to be applied with a lighter touch so as not to redissolve and disturb the first, by then dry, paint layer. This would account for the hatched shading, almost like drawing, to be seen, for example, on the faces in Bouts’ Entombment.’ P188
‘Because of the optical properties of the size medium, all pigments appear more or less opaque, event those pigments transparent in oil like the lakes or indeed the chalk used as a white pigment by Bouts. For similar optical reasons the expensive mineral pigments, and in particular azurite and ultramarine, appear at their most brilliant in a glue medium, and there is evidence that glue size may occasionally have been used for these pigments on works which are otherwise in oil or egg tempera.
The relative opacity of pigments in a glue medium results in good coverage with a reasonably thin layer of paint, an advantage if a canvas is to be rolled for transport. But as soon as a varnish is applied it is absorbed by the inherently lean and dry paint, and many of the colours will become dark, saturated and unintentionally translucent.’ They should not have ever been varnished, but without such a protective coat they are very likely to be damaged by water. ‘Therefore the surviving examples, with a few notable exceptions, represent little more than the shadowy remains of the original paintings.’ P188
I did not want to project these descriptions of 14th and 15th century techniques onto my reading of the Ms 093v recipe. But the longer I have been trying to interpret the recipe and use the methods he suggests, the more it seems to correspond to other descriptions in secondary sources of relatively standard processes of painting with glue. The combination of the possibility of softening and blending paint passages with hatching in the upper layers of shadow seems to match 093v precisely. It seems that he is describing a traditional method, and is evidence that such a process may still have been current in late 16th century France.
Monday 2nd May
Summary
Final hatching shadow layer was applied to the face in distemper.
Different handling quality of using the glue paint on the dry canvas, as opposed to previous method where the canvas was kept wet.
The hatching was somewhat difficult to achieve but once I had become used to it, worked well as a system for reinforcing and emphasising shadows.
Also tested out the pigment combination for women's shadows suggested at 066r to evaluate how different a hue resulted from this combination. Also to compare oil and glue handling.
Longer field notes
To put in the final layer of paint on the face of Mary.
This is the final layer of shading, done by hatching.
I tried to do this layer without wetting the canvas, thinking about what Dunkerton said about the top layers being applied with a lighter touch, through hatching, dry, so as not to disturb the already dry glue paint layers below. Quote:
Dunker ton: 'However, any subsequent application of paint, including details such as facial features and strands of hair or superimposed patterns on textiles and marbled stone, would have had to be applied with a lighter touch so as not to redissolve and disturb the first, by then dry, paint layer. This would account for the hatched shading, almost like drawing, to be seen, for example, on the faces in Bouts//'.
The author practitioner specifies hatching at this final stage. I made the paint as before, 1:10 rabbit skin glue combined with bistre, sap green, and ochre de rux. I applied this in a light hatching motion over areas of the face where I wanted to reinforce the shadows. This worked well on the face. In my opinion, the areas where the layers have been built up to 3 layers (base shadow layer, carnation layer, then final hatching layer) look relatively subtle. Although, it would have been better if I had used an even smaller brush, or worked on a larger scale, or perhaps had just developed greater manual facility with hatching in this medium.
The manual experience of hatching on a dry canvas was different, felt different, from applying the medium to a canvas dampened from behind.
There was more 'drag' in the medium, it felt as though it skimmed across the tops of the canvas threads rather than sinking into the threads as much.
I think that applying a final layer of darks gave a greater sense of modelled form to the face.
In the hair, I just hatched over the top of the previously applied base shadow. I did this area first to 'try out' the hatching and my initial difficulties in keeping a sufficiently fine tip to my brush are evident.
I chose to leave one side of the hair with just the base shadow layer, because it shows the kind of initial shadow layer that the rest of the layers go on top of. Also, you can see a spot where the glue clumped and came off. It is interesting to see how difficult I found the medium at the beginning of this process.
I also tried out 066r_4
<id>p066r_2</id>
<head>Umbres</head>
<ab>Pour femme <m>Aspalte</m> <m>terre dumbre</m> & un peu de <m>laque</m></ab>
</div>
Shadows
<ab>For woman: <m>asphaltum</m>, <m>umber</m> and a bit of <m>red lake</m>.</ab>
I combined ashphaltum pigment, umber pigment and a madder red lake.
I used approximately equal parts asphaltum and umber, and half as much red lake. (See notes on pigments used for shadows).
I ground these together in linseed stand oil for half an hour. I repeatedly tested the paint as I was grinding by painting it out and seeing whether granular pigment particles were still visible. After half an hour this had improved, but it would still have been further improved if I had ground the paint for longer.
The resulting paint was glossy and quite transparent. This combination of pigments in linseed oil does not create an opaque film but rather a translucent 'glaze' like one.
I had thought, on reading the recipe, that the colour produced would be somewhat red, or notably warm.
However, the colour produced was brown, with if anything a cool rather than a warm tendency.
It was not very different from the kind of colour produced by the mixture of pigments suggested at 093v for faces in distemper.
This could hint at some consistency in the author-practitioner's understanding of what hue women's shadows should be, that the 'greenish' quality that he suggests they have runs through his conception on female shadows.
But if this is the case, why add the red lake? What does that do to this mixture? Is there any symbolic meaning?
Handling properties notes, personal reflections: using animal glue paint and linseed oil paint during the same morning allowed me to reflect on the experience of using the different media. As I used them, both were quite translucent. The fact that oil does not dry allowed me to make micro decisions as I went along. With the glue, even as I have developed some relative facility with it, I have to make a decision first and then carry it out at reasonable speed, because even with a lot of water and keeping the glue warm it still dries much more quickly than oil paint. Though the possibilities of layering afforded by both media, optimising their relative translucency, are slightly similar, overall the handling properties are very different. I am intrigued by the fact that using animal glue I was able to both blend, and hatch, depending on whether I wet the canvas before or not. If I wet the canvas, blending and softening between the layers could work to some extent. When I did not wet the canvas, the hatching system worked. This seems to be a somewhat significant flexibility. Softening works very well in oil paint, it is easier to soften with oils as they allow wet-in-wet blending due to their slow drying nature. But can one use a hatching system in oil? Not at all easily.
These thoughts on handling properties are not necessarily 'historical' because while I have attempted to reconstruct the directions in the manuscript, there are so very many variables that must have altered in my re-creation of this, and also I am not trained as early modern practitioners would have been. It may be that different concentrations of glue would result in greater or less great similarities to oil.
Monday 9th
Summary
It was interesting that the author-practitioner suggests 'jet' as a black pigment.
Ground pigment from a piece of jet jewellery. It was possible to grind it to a fine powder and use in oil, but more time grinding would be needed to really use it as a pigment. Also, the colour was surprising.
Followed workflow as has been edited, see workflow for step by step process.
This was the piece of jet jewelry that we began with.
We then sawed it into pieces. 
These seemed surprisingly blue, I had expected the inside of the piece of jet to be black.The blue colour could be to do with desiccation. I then used a hammer to break up these pieces into smaller fragments which looked like shards of blue glass.
I then took these smaller fragments and pounded them in a mortar and pestle to break them down further. 
As I continued grinding the jet, the color gradually changed, and the yellowish stain around the inside of the bowl intensified. It seemed as though the jet was producing this yellowish substance and becoming more green as I ground it. Then I added linseed oil, and continued grinding it in the pestle and mortar.
After grinding with linseed oil, this blue-grey substance was formed.I then used the muller to grind it further with the linseed oil on the 'porphyry' slab. Gradually the colour seemed to darken, although the final color arrived at was only ever a dark teal, it never came to appear like it would work well as a black paint.
In total, around two hours were spent hammering, grinding and otherwise breaking down the jet. Had I continued for longer it would most likely have become more pigment-like.
I experimented combining this jet paint with lead white oil paint and yellow ochre as the author-practitioner suggested for men's shadows, and tried painting them out on pre-prepared boards. The colour arrived at was greyish, but not the dark yellowish black that I had anticipated from the manuscript. This seems like a subject that would merit further time and research.
