First the Process, second the Backstory, and then we’ll talk about the Why.

1: the Process

The sea-horse you see above:

1. You apply a wash of paint, and blend. (“Paint mixed with what?” you ask: good question – we’ll get there soon.)
2. You wait several hours – note: this depends on how warm and dry your work-place is – then copy-trace i.e. put your glass on top of the design and trace the main lines.
3. The tracing finished, you can immediately strengthen your lines and embolden them.
4. You wait a few hours for this paint to dry, then cut the highlights, and soften them by hand or use a scrub.
5. For shadows, you can immediately apply a wash or indeed some wide lines of tone, and blend.
6. You wait several hours for the paint to dry, then reinstate old highlights or cut new ones.
7. Then you fire – just the once.

And yes, you’re right: there is a lot of waiting.

But it’s not a problem when you have 5 or 10 bits of glass you’re painting: by the time you’ve done the last one, you can return to the first.

The main challenge is actually the liquid you mix the paint with.

• It isn’t water and gum Arabic.
• It isn’t glycol.
• It isn’t Lavender or Oil of Tar.

When using this liquid whose name I’ll tell you in a moment, we always wear a mask.

And I want to be clear: if I talk with you like this in public, I’m assuming you’ll be sensible. It takes a lot for me or Stephen to wear masks. But, when working with this liquid, we do. I hope that’s clear.

2: the Backstory

So two months ago I spent a morning looking for two bits of glass I painted 6 years ago when a client asked us to undertake a demanding forgery of 4 pieces from a window which a vandal threw a brick at.

The bits of glass I sought were just the usual tests I always do before the actual work.

You may imagine that in our studio there are tens of boxes filled with glass-experiments of every kind:

• Different techniques.
• Different paints.
• Different media we mixed our paint with.

And the upstairs walk-in cupboards are crammed with yet more boxes.

Thus the morning’s search.

But in the end I found the two bits painted glass I wanted.

One of them you see above: the sea-horse.

It wasn’t connected with the client’s window except for the liquid which I mixed my paint with.

All that waiting time, right?

Well, this liquid was suggested to me by an entry in an old Victorian journal Stephen owns: a long-forgotten painter, Nathaniel Somers, dissolved Shellac (a kind of resin) in warm Linseed Oil, then combined it with his glass paint, and worked according to the steps I gave you earlier.

Now next door to us there works a gilder.

And when we mentioned this Victorian recipe to our next-door gilder, he suggested we experiment with the same ‘glue’ – the correct term is gold size – he uses to stick gold leaf to wood or metal:

And that’s what I did: I mixed my paint with a particular gold size which remains ‘open’ (i.e. doesn’t dry) for about 3 hours.

For the sea-horse, I then went one step further, because a gilder doesn’t fire his work. (He doesn’t need to. Gold leaf is metal: when the glue dries, the leaf bonds securely to the surface.)

But of course I fired mine.

And as you see above, the results were promising.

All those layers.

That’s why, back in 2011, we used the same technique for our client’s vandal-damaged stained-glass window.

You see, if the end-result is what we want, we’re thrilled to change the process, because otherwise it gets boring – always doing things the same way, time and time again.

Besides, you never know when a skill might actually prove useful.

Like now …

3: the Why

The Why – as in “Why did I spend a morning hunting for the glass?” – is: because there are situations when ancient glass paint deteriorates.

Yes, sometimes the fired paint just fails.

And you can’t re-paint it and fire it again, for all kinds of good reasons.

So what can you do instead?

Well, one way to restore its appearance is by painting on the back, and leaving it unfired:

1. You need a glue to mix your paint with which is very strong and yet not permanent.
2. This liquid must be capable of all kinds of lines and shadows.
3. The back must be protected in various ways or else that paint will also fail.

Enter gold size for a second time.

Because you’ll remember I looked long and hard for two bits of glass I painted those 6 years ago.

The first was the fired sea-horse we’ve now discussed – the experiment I fired.

The second was a crown which, just by way of investigating the effect and seeing just how tough the “glue” was, I’d painted on the back and left unfired – it’s a fragment from a window I rescued a long time ago.

And I agree – from behind, it all looks somewhat rough:

But behind is not where you mostly see it from.

It actually looks very good in front. Not perfect. But you can read it (which you couldn’t do before):

You can rub hard at the unfired paint behind, and yes, it’ll eventually come off, but if you protect it, it will last: gold size is considerably more robust than glass paint mixed with water and gum Arabic.

And this is a method which we’re using now for real.

As I said, it pays to learn before you must.

P.S. When you use this unfired back-painting method for restoration, there are a hundred details to consider which I won’t go into here, and of course, you can’t just “do it”. After all, the paint has faded which means it’s difficult to see the image, like this inscription here:

That’s why you usually re-draw the image onto tracing paper first – you reconstruct it from the ghost-lines which remain.

So here’s a drawing I prepared for a lion on one of the windows we’re working on right now:

He’ll be fun to paint.

But I’ll finish all the drawings – the windows are full of images like him – before we don our masks and start again to paint with gold size.