Effects of ageing and conservation efforts on Renaissance egg tempera paintings

​Thank you very much, I find both articles very interesting. 
It seems that the question about pigments used to paint his hat is answered: contents of mercury and tin were found in a closely related painting, and it it said that lead-tin yellow type I was identified in other places, probably the hat is so orange because lead-tin yellow was added to vermillion.

However, I have new questions. It is said that flesh colors were mostly painted using green earth and lead white underpainting; yellow earth, lead white and vegetable black subsequent layers; lead white with vermillion highlights. I have several green earths: chlorite, celadonite and glauconite, all are really green – very suitable to paint a zombie. Did they used some other form of green earth of dark-yellow color without such pronounced green component? Perhaps what is now known as “Antica green earth” or “tobacco coloured chlorite”?

Also it is said that zones of monotonous color were applied in first stages of painting process. Is this technique close toByzantine icon painting sequence? 

Hello,
These are great questions, albeit difficult to answer, especially without scientific analysis or an in person examination. I’m not a conservator, but I have worked as an egg tempera painter and teacher for many years – so I’ll offer some ideas with the caveat that they are guesses.  Here goes:
1. What do I see at the edges of paint flakes divided by craquelure? I think that when the ground layer cracked, separate flakes slightly curved and their edges raised. Next, paint near edges of flakes was abraded during handling and cleaning of the painting, upper layers of paint were stripped. So at the edges near minuscule cracks I see lower paint layers. Is this correct?
When you say “ground layer”, perhaps you’re referring to the initial layer of paint; however if, by ground layer, you mean gesso, I want to clarify that a traditional gesso ground only tends to crack shortly after being made if it contains too much glue or a bad glue; or it can crack later if it’s dropped or hit very hard (then you see a localized crack that spreads out from point of impact), or the wood panel underneath moves and/or the grain pattern telegraphs through (although generally fabric was/is applied over wood to mitigate this). But a well-made gesso itself, as far as I know, doesn’t crack into craquelure as it ages.  It’s the paint, as it shrinks, that develops craquelure. Is that accurate, Brian? 
It’s hard to say what appears within the craquelure: is it underpainting, gesso or wood showing through; or dirt in the cracks accumulated over the centuries sealed by later varnishes; or a later varnish that’s darkened…? A painting of this age (500+ years) has been through a lot. I would guess a conservator can answer. What I can say is that the brown within the craquelure on the flesh would not be an underpainting of brown; tempera layers are so thin, and so many have to be accumulated to render a form, that you don’t begin with a dark layer to render a local light.  
Within the long vertical and horizontal cracks (which don’t appear to be craquelure) on the chest, I’d say the darker brown color of the cracks is, in fact, the initial shirt color. On top of an initial layer of brown, Botticelli layered more paint to model the form – it’s those upper color layers that have been abraded or delaminated in the cracks, revealing Botticelli’s initial layers of paint within the cracks.  
2. How transitions between light and dark parts of the face were applied, particularly at the nose and cheek? I see darker spots of paint without clearly defined edges. They appear to be spatially oriented as if they are parts of longer brush strokes. What causes this interrupted appearance? Are they a result of paint abrasion? Or his panel was grounded with some texture and we see an effect similar to watercolour granulation on rough paper?
Traditional gesso was sanded very smooth, so I don’t think it’s granulated texture in the ground itself you’re seeing. Some dark spots may be pinholes in the gesso.  It’s not uncommon to get pinholes when building up gesso layers (although pinholes aren’t inevitable with proper gesso technique).  Pinholes tend to suck up concentrated amounts of paint, so if Botticelli glazed umber over the flesh at any point, a pinhole would turn into a dark dot.  Attached here is close-up from a Botticelli portrait – there are pinholes all over the surface (apparently his apprentice wasn’t very careful with the gesso that day). 
Pinholes, Botticelli.jpg
Another explanation for the granulation effect is complicated, so bear with me. Tempera is built up in many, many, many layers.  With so many distinct layers and brushstrokes, a certain amount of texture is created along the way. (In particular when applying paint with a sponge, as I do. I know of no evidence to support Renaissance painters using sponges to speed up the otherwise slow development of tempera paint layers; but I do know that Annigoni, a renown early 20thc. tempera painter, ended up doing this.)  Anyhow, tempera is also a very telegraphic medium – underlying textures readily show through – so texture from accumulated paint layers starts to influence how fresh paint lies on the surface.  Finally, because tempera paint is applied so thinly, even opaque colors (such as white) tend to behave more or less semi-transparently. In other words: innumerable brushstrokes, multiplying transparent or semi-transparent layers, accumulating textures – all, to some degree, are visibly interacting with and affecting one another.  It can create a sort of consistently inconsistent, granulated look to a surface. 
3. Opening the image in GIMP and using a CIELch colour picker reveals complex variations of hues, particularly at skin areas. How such variability was achieved? 
As described above.  I’ll also attach a close up of a portrait of mine – looking closely, you’ll see a similar weave of colors within the generally local flesh color.  
K. S. Flesh Closeup.jpg
I apply anywhere from 40 to 200 distinct layers to paint flesh.  This sounds impossibly laborious, but because a thin layer of tempera dries to the touch within seconds, it’s possible to apply many, even scores, of layers within a day. Botticelli, being such a master, probably didn’t need as many layers as I do!  Still, given the thinness of tempera paint, he undoubtedly applied a lot of layers in order to achieve as developed, realized a surface as he did. 
In my understanding of egg tempera technique, a painter mixes pigments in raw state, grinds them on a stone with binder and puts resulting paints in dishes. 
I don’t mean to nitpick, but in general colors were/are ground first, then stored as powders; so when a powdered pigment is combined with binder it’s being dispersed (not ground). It’s a small yet important distinction; appreciating it helps one to better understand the paint making process.
Without mixing on a palette since tempera dries fast. I have several guesses, but what is correct? Perhaps, many pigment mixtures were composed in raw state? Or he composed just a few pigment mixtures and applied them in several very thin layers of varying density? Or he used tempera grassa which dries slower and mixed it on a palette like oil paints? Or the painting was heavily overpainted during many conservation efforts and we see a result of using slightly different pigment mixtures by restorers? This is certainly the case near some major cracks and scratches, but to what extent other areas are overpainted?
As Brian notes, the indirect (many layers of paint applied over time) method used by Renaissance tempera painters developed out of the icon tradition.  (A good, contemporary book on the Icon tradition is: Hart, Aidan.  Technique of Icon and Wall Painting, Freedom Publishing, Australia, 2011.)   There is more variability between icon traditions than is sometimes appreciated; still, there is a general working method that continues to this day, so if you read over any contemporary icon practice you’ll have a general grasp of the tradition.  It was (and still is) a very precisely ordered method; it is also a spiritual practice, so deviation is discouraged. 
Renaissance tempera painting arose out of the icon working method, but there was more freedom to elaborate on it.  Renaissance painters did work within a studio’s established practices; still, painters were individuals and able, specifically masters, to modify a method to suit his individual nature and goals. All of this is to say that I don’t think there was a single, unaltered way to work in the Renaissance; within the influence of the icon tradition and a studio’s general practice, there was some individuality and improvisation.  This is apparent when one looks close up at portraits by Ghirlandaio, Botticelli, Cosimo Tura, Fra Angelico, etc. 
It’s not possible to physically blend tempera paint on the surface of a painting; but you can intermix dry pigments (as they did to make “cinabrese”, a Caucasian flesh color made from lead white, yellow ocher, and vermillion), and/or intermix tempered colors on a palette, and/or layer them one atop the other within a painting – or do all three at various points within a painting. Renaissance artists were generally reluctant to intermix expensive, high chroma colors with a dirtying earth colors (high chroma was much harder to come by in those days); in those instances they generally layered.  Finally, Renaissance painters generally worked efficiently (Leonardo being an exception), were masterfully trained, and probably wouldn’t have endlessly noodled with layers (as I, sigh, sometimes do). These are some of the things I think about when trying to understand how they painted.   
4. Is there a darker layer of paint under light areas of the skin? Or lighter parts were painted straight on a white ground with some sketch made with dark coloured lines? I guess I see darker underlayer in a lower left corner of his neck near the fur collar.
As mentioned above, they did not underpaint a local light with dark in tempera – too much work to overcome the dark.  They generally (tho’ not always) modeled a form first in ink (watered to a range of values) to establish the value pattern and help the darks along (faster to overcome the white of gesso with initial wash of ink); atop that, they modeled the form with green (either clean or dirty), which helps create halftones.  Next they built up light values using  a warm, relatively opaque (almost nothing is fully opaque in tempera; it’s too thin a medium) paint; and deepened darks with warm, low chroma, transparent paint.  Between the light and dark – where the edges of the warm, opaque flesh paint overlap the green earth underpainting – a cool pearlesence, as described by Brian, is created. This is a scumbling effect (akin to how mist turns a distant green landscape blue) that conveys the cool halftone of flesh.  Amidst this systematic method there were lots of opportunities (given how many layers tempera necessitates) to glaze, scumble, improvise.  The variability of brushstroke and how much you do or don’t thin tempera paint also contributes to the variability among Renaissance tempera painters.   
An excellent book for understanding the related yet varied ways in which Renaissance artists began their paintings is by Billlinge, Rachael (and others), Art in the Making: Underdrawings in Renaissance Painting, National Gallery London, 2002. As I say above, within general visual principles (such as “cool halftones”) that guided traditional  imagery, and whatever studio tradition an artist worked, individual natures and goals manifested themselves; more so in the Renaissance than in prior eras because of the increasing individuality of artists that emerges at the time.   

5. What happened with some hairs of the left eyebrow (model’s right eyebrow)? I see crisp lines of dark coloured paint but a couple of lines are very dim. Were they abraded? Or they are parts of preliminary drawing covered by subsequent paint layers?
My guess is that they are thin strokes of tempera – this is what a single layer of tempera generally looks like (and why one needs to build many layers to fully render a form and achieve any sense of solidity).  Botticelli left a few thin, individual strokes both above and below (see in the shadows) the eyebrow, probably to soften the edges so the eyebrow doesn’t look penciled on. 
6. The overall skin color is very yellow, about 1/3 more yellow than a real skin color of a Caucasian man could potentially be – I am judging by CIELab colour coordinates and limits given in an article by A. Chardon «Skin colour typology and suntanning pathways». I wonder is this a result of binder or varnish yellowing? Or an artistic choice to make it look more saturated?
The majority (maybe all?) of Renaissance panel paintings were varnished at some point in their long life, and most old varnishes yellow, so it might be an old varnish (unless it was recently cleaned).  If the painting contains either oil and/or tempera grassa (Botticelli worked in both at times) oil could have yellowed the painting. However he most often worked in egg tempera, and egg yolk does not yellow with age. 
Regardless, while the painting may be a tad warmed with age (hard to say) I don’t think it appears overly so.  Painters of the past nearly always created relationships between warm and cool colors.  In a painting yellowed by age everything is overlaid with a warm film and cools are lost. In this painting the browns, reds and flesh colors are warm; but the fur trim, white shirt ruffles, highlight above the lips, whites of the eyes are cool greys – they don’t look overtly warmed. So I don’t think the color temps in the painting are necessarily too far from the original. 
It’s possible to argue the subject’s skin in fact is a warm, olive color not atypical of many Italians.  But that misses the point that Renaissance artists did not paint realistically, they painted representationally.  I very much doubt Botticelli was overly concerned with getting the real skin color of a Caucasian man.  As much as he was creating a portrait, he was creating an image, and making it visually interesting and exciting was part of the job. The warm yellow chroma of the skin is played beautifully against the achromatic black background; and cool halftones within warm flesh each accentuate the other.  So I agree with your last statement, it was an artistic choice. 
7. The color of his fancy hat is puzzling. Too orange for an ordinary cinnabar. I guess its color is an effect of particle size – somewhere I’ve heard that very finely ground cinnabar turns almost to carrot-orange. Or is it a result of fine glaze applied atop of cinnabar? Or another pigment was used too, like realgar or orpiment?
My experience is that vermillion, which is the manufactured version of cinnabar, and was, I believe, more common in the Renaissance, is an orangey red – so the hat is not far off, maybe a bit warmer, which could be accounted for with a glaze of yellow on top. 
Probalby more than you needed to know, but it’s a worthwhile challenge to try to understand Botticelli.
Koo Schadler

​Thanks a lot for your replies!