Sunday, 5 November 2017

Iron Gall Ink: Recipe using Knopper galls

Iron gall ink is literally the stuff of history; it was used to write many medieval documents, including the Magna Carter. It is made from oak galls (deformities of the leaf, acorn or other part of the tree, caused by the larvae of a parasitic wasp). If you search the Internet, you will find many recipes for this ink, including several impressively venerable ones. If you dig a little deeper, you will find that the chemistry involved is only partially understood - but I found enough information from modern science that I was able to adapt the recipes to make my own version, based, it has to be said, on a series of assumptions and guesses.

Each type of gall contains different amounts of tannins. My recipe is intended specifically for use with knopper galls.

The recipe


Quantities depend on the clean weight of the galls and are indicated in the method.
  • Knopper galls
    These fall with the acorns in the autumn and can be collected from the ground beneath an affected oak tree. Some of mine had been kicked around a lot (the tree is on the edge of our village school's playground) and were rather muddy.
  • Water
    I used tap water. It is probably better to use rain water or filtered water.
  • Ferrous sulphate
    Many recipes suggest the use elemental iron (either as a cooking vessel or in the form of rusty nails). As I understood the process, ferrous sulphate (iron (II) sulphate) is a safer and surer reagent. It is inexpensive and can be bought, in crystalline form, from garden centres as a moss killer. 
  • Gum arabic
    This is a natural binding and thickening agent. It is sap from the acacia tree, and is sold in the form of dried lumps of sap or as a solution. I used lumps.
  • Cloves
    The spice is a natural presevative.


  1. Clean the galls with water and separate them from the acorns as necessary.
  2. "Break" or "crush" the galls.
    I took this to mean that the galls did not need to be ground into tiny pieces. As knopper galls are very hard, I folded them into a cotton bag and hit them with a mallet. A hammer would also work.
  3. Sort through the broken up galls and remove any larvae or foreign objects. Weigh the galls and put them in a jar.
  4. For every 10g of galls, measure out 100ml of water. Add the water to the jar. It should cover the broken galls.
  5. Put the jar in a warm place and leave it there for 10 to 14 days.

    I used the airing cupboard, where the mixture slowly and gently fermented as the tannins in the galls leached out and reacted with the water to make gallic acid (also a tannin). There was a little effervesence (bubbles) and the liquid turned dark brown. It did not smell or erupt.

  6. Strain and filter the liquid into a clean jar. Discard the solids.
  7. For every 10g of the original weight of galls, add about 5g of ferrous sulphate.

    Gallic acid and ferrous sulphate react to form ferric pyrogallate (iron (III) pyrogallate), an insoluble black solid. This is the ink pigment. Athough it is insoluble in water, it will form a suspension in water, and this suspension will be enhanced by the binding agent, gum arabic.

  8. Add the same amount of gum arabic to the mixture

    Gum arabic dissolves very slowly in water. Heating and whisking may help it dissolve, but the most effective method seems to be waiting. My ink was ready to use the following day.
  9. Decant, using a funnel, into your storage bottle(s).
  10. Add a few cloves to help preserve the ink.

The rationale

What follows is my attempts at trying to make sense of a complicated and much debated situation. Chemists, conservationists and assorted others (like artists!) have all contributed. The starting parameters are uncertain (how much of what is in this gall?), the reactions are complex, intertwined and poorly understood. Time, measured in centuries, plays a key part, and that makes judgement of results difficult. Definitive numbers are elusive.


Warning: science bit.

Iron gall ink can sometimes cause corrosion. There appears to be some confusion about how and why it does this, but from my reading, it seems likely that the corrosive nature has very little to do with the tannic acids (which are mild acids, no more dangerous than vinegar).

Edit: This article suggests that the acids do affect the ink's stability.

The corrosion causes old manuscripts to disintegrate when they are handled and seems to be caused by the oxidisation of a soluble ferrous salt that is present in the ink (it oxidises to ferric oxide, commonly known as rust). This salt is the product of an incomplete reaction between gallic acid and ferrous sulphate, which I interpreted to be the result of an excess of iron during the reaction. (Iron is dual valence; it can form compounds at +2 or +3 valencies, but the +2 compounds tend to be less stable than the +3 compounds.)

So, to make ink that does not cause the corrosion, care must be taken not to add too much iron.

I tried to calculate the ideal amount of ferrous sulphate. I found figures that suggested that British knopper galls typically contain around 40% tannins (a high value; marble galls - the other type of oak gall that I had found locally - have around 16% and the bark of the tree has around 9%). The molecular weight of gallic acid is 170.12g/mol, and that of ferrous sulphate heptahydrate is 278.01g/mol. I found the abstract of a scientific paper about the reaction, from which I inferred two things: firstly, that it was a complicated reaction, and secondly, that the reaction ratio might be 1:1.

I was using the kitchen scales to weigh everything (they are not very accurate), so it seemed reasonable to approximate the ratio by weight of gallic acid to ferrous sulphate crystals to 1:1.5 or 2:3.

Remember the 40% tannin content? If I assumed that all of the tannins were converted to gallic acid by the fermentation process, then the relative weights of galls to ferrous sulphate crystals should be 5:3. Given that I wanted to err on the side of caution regarding the amount of iron involved, I adjusted this to 2:1. This made it easier on the kitchen scales, too.


Note: According to the mostly authorative Iron Gall Ink Website,
"Recent research indicates that a 3:1 ratio of gallotannic acid to iron sulfate produces the most stable inks."
I discovered this nugget after I had made my ink. Was I wrong? (It's entirely possible; there were an awful lot of guesses and approximations.) However, they are talking about the ratio of gallotannic acid to "iron sulfate" (which is the same thing as ferrous sulphate), not gallic acid. But is this a molar ratio or does it refer to the weights used? It's impossible to tell.

Gallotannic acid is the main tannin found in oak galls. It has a molar mass of 1701.19g/mol, ten times that of gallic acid, to which it hydrolyses during fermentation. It is the gallic acid that actually reacts with the ferrous sulphate (iron sulfate).

Before I got too lost in my calculations, I discovered a more recent scholarly article ("A few thoughts on iron gall ink reconstruction) which stated that,

" a 3:1 molar ratio of FeSO4·7H2O to gallic acid [...] has been known to have the most chemical stability"

I used a molar ratio of 1:1, with the result that I had only a third of the iron that I should have had.

I stopped trying to make the numbers work then. The ink is nice and black and it works now.

Call me in 300 years if it oxidises and makes the picture fall off the paper. In the meantime, I shall treat it as "non-archival" and I shall enjoy drawing oak trees using it.

Saturday, 7 October 2017

Porth Moina

Porth Moina, oil on canvas, 100 x 73 cm
Porth Moina is in Penwith, Cornwall, on the north coast (and on the southwest coast path) . Porth means "cove" and Moina probably means "mines" in this context.

Saturday, 27 May 2017

Purbeck Eddies

Purbeck Eddies, oil on canvas, 40 x 50 cm
I started this in the last few days of Open Studios, working from a 2014 photograph taken off the beach at Winspit in the Isle of Purbeck, Dorset. I finished it today, having taken a few days off.

Monday, 22 May 2017

Recorders and Lute: Palisander and Toby Carr (triptych)

Recorders and Lute (triptych), before framing

These three pieces were painted using inkjet printer inks (without the inkjet printer).

It's not over until...

Open Studios West Berkshire and Nort Hampshire has technically finished, but there are a few satellite exhibitions lingering here and there and, of course, there are several studios that are full time affairs and which will often open by appointment.

So yes, you can still visit me. Just get in touch...
[Contact page]   [Open Studios Directory entry, includes full contact details]

The outstanding exhibitions that I know about (because I'm involved in them) are:

Art in Music at the Corn Exchange in Newbury - on until 24 May, during performances only;

Cover to Cover (exchange sketchbooks) at the West Berkshire Museum in Newbury - on until 30 May;

North Hampshire Art at the Willis Museum in Basingstoke - on until 3 June. 

Sunday, 21 May 2017

The road to Kimmeridge: some in-progress shots

Sometimes I think I might forget how I did something (and the best somethings are often the result of a happy accident) or, more often, I worry that the next bold stroke of the knife will wipe out the best bit of a painting.  So I've got into the habit of taking quick photographs of the current work-in-progress in the studio when my hands are clean enough to operate a camera (in the field is different; there's more time pressure and, usually, fewer distractions).

This post contains a sequence of such photographs, with notes, for the painting of Kimmeridge, which was painted on a clear-gessoed linen canvas with the six tube version of my standard palette: burnt umber, ultramarine, phthalo blue, lemon yellow, rose madder quiacridone. Most of the knife work was done with a Winsor and Newton no.27, with a smaller knife (Langnickel LP-1) used later on. I also used a silicone colour shaper for a few details and my signature.

I'm working from a photograph that I took in 2014. I hadn't seen the possibilities of the shot until now (except maybe in the instant that I took it). I remember the day. It was hot, very hot for England, the heat of the sun bouncing off the rocks. My husband and children were in and out of the sea but I felt unable, uncertain of my physicality, uncomfortable in the heat and unsure of my balance - and besides, I wanted to sketch. I made a very bad, overworked and tiny watercolour of the distant Clavell Tower. And I took lots of photographs.

This was one of them.

You don't get to see my source photographs very often.

Nearly three years later, I picked it out for the studio treatment. There was a lot of fiddly detail that would challenge the knife; of course, it wouldn't all go in, but if there was any chance of making the Clavell Tower recognisable in any way, I had to use a fairly large canvas. And I was trying not to think about the seaweed, but its colours anchored the image and justified the long format, so I would need to make them work... somehow.

Starting with the darkest areas, pure "dark": burnt umber and ultramaine - marking out the composition. This isn't all of the dark bits, but the position of the horizon and of the central band of rocks seemed, to me, to be important. I placed the top of the rocks quite deliberately at that "magic" 1/3 point that all the theory says is important (or is that "all the theory that I can remember"?) This image didn't need much sky. It was about the sea. And the rocks. Especially the rocks (and the way the sea shaped them).

Well, that's all of my initial darks added, with some colour on top (mainly blue). The lower right is where I imagine I'll be signing it and I try to remember to put a nice contrasty dark there in most paintings. But this isn't most paintings and I completely overlooked the fact that the foreground is all complex seaweed and shadow - too detailed for a swirly AJB right there. But it is quite shadowy in that corner, so that's fine. Hang on a minute - I also forgot to put the sky in!

Oh, thank goodness for that. There's the sky. And some pink bits.

Pink is a surprisingly important colour, Of course, it needs to be a good, strong pink, a blue-red colour - that can be used to tone down the most excessive of your greens, to make lively greys (see the sky), to create purples and oranges and reds. By this stage, I've employed the full paletted and I'm moving around the picture according to where the next bit of the colour on my knife should be. That colour changes as it picks up paint already on the canvas and I'm constantly reassessing, occasionally wiping the blade or, less often, scraping "wrong" paint off the canvas. The rocks are mostly blocked in, aong with the bluest bits of the sea (it's only now that I realise quite how much I have heightened the colour) and I've had a go at some of the seaweed.

That's most of the canvas covered. I'm leaving bits of the natural linen on purpose. What's the point of a clear gesso if you don't? The rocks are starting to look solid. That red stuff on the closest rock shelf is glorious. I exagerrate it. The seaweed is still a bit ... vague.

I'm ignoring the seaweed. I'm having fun with the rocks. Edge of the knife, incising, finding the hidden dark, or, if that is lost, adding more dark, freshly made on the palette, as it slices through the softly yielding paint already on the canvas. I lose myself in the abstracted detail of the rocks.
The rocks are working; not quite complete, but I know what's happening there. I've softened the dark shadows on the edges as they curve away from the shadows I've established the main colours, the majority of the cracks and lines. Let's have a go at that seaweed... More shadow. Greenish bits. It's blotchy and crude and it looks alien to the rest of the image.

I distract myself with the fiddly details of the tower. It's a distinctive shape, and the appropriate proprtions elude me for a while. The back down to the seaweed... Suddenly, I realise that it's too green. It's a bit green, but there is far more yellow and red - even some white to lighten beyond the powers of the lemon yellow. And the shadows are less blue than I had them originally. The foreground paint layers up, gets scraped back here and there... How do I make it look wet and under-watery? Reflections, ripples, white, nearly white, a few dark, incised with the silicon colour shaper, with the knife (a finer, subtler line, harder to curve). And finally, carefully, dragging a near-white mix gently down over the surface, barely shifting it, distorting a little, smoothing its paleness into the colours already there... Done. Signed. Complete.