An interesting project came to the Books department last week: Brittany made this beautiful silver case for a tiny automaton, and the client asked for it to be covered in shagreen. We use a lot of different skins on a regular basis, but this one is unusual:

Shagreen is a bit of a loose term for a range of rough-textured skins. The 17th century Persians made it by pressing pebbles or seeds into horse, camel, or donkey skin to give them a bumpy grain; since then it has been used for a variety of coarse-grained skins. Now it most frequently refers to the skins of rays as well as (less frequently) sharks, dogfish, and other cartilaginous fish. These are covered in protrusions called placoid scales; they're very similar to teeth, with an inner core of pulp surrounded by dentine-like material and a thinner layer of enamel-like material. In Japan shagreen was used (undyed) for sword handles and armor, and in China for bows, in both instances its texture offered the user a better grip. In Europe for a long time it was simply an inexpensive byproduct of the fishing industry, used for abrasive: check out the description of making this modern shagreen sanding block.

Though it had been imported to Europe for about a century already, in mid-18th century Paris Jacques Galuchat popularized green vegetable-dyed shagreen when he covered hundreds of small objects for Louis XV and Madame de Pompadour. The material experienced a renewal in the 1920s on Art Deco furniture and small objects, often either green or undyed. The skins are small to begin with, and the useable parts even smaller, so it tends to be used to cover only small items, or in multiple panels if on furniture: there are some nice photos here.

Shagreen is prepared by scraping, stretching, and drying the skin rather than tanning or tawing it. When used for decoration rather than grip, the dermal denticles are usually sanded down flat, as in the skin above. This one is an adult ray from Ed Tanner in London. You can see that the (vegetable) dyes are largely picked up by the skin rather than the scales, which are larger on older animals. Rays also feature this knot of bigger, thicker scales along the spine. The bigger the scales, the more difficult to cut the skin in a straight line: they're hard and brittle, and put up a fight when subjected to my breakaway knife, whose blade had to be changed constantly. Conservation and identification was covered quite effectively by Margot Brunn here on the Conservation DistList. Since Brittany's case was new, we didn't have to consider adhesives and ethics as carefully as if it was original to the automaton.

On to the covering:

My strategy was to wrap one piece each around the sides of the lid and bottom, cut away corners as it wrapped up around the rounded edges, then fill the center with another piece. One of the challenges of shagreen is that the scales are bigger towards the center, becoming smaller gradually but consistently towards the edge; virtually nowhere could I cut a piece that didn't include some kind of size gradation, which would be painfully visible at the seams. And, like I said before, the larger scales were much harder to cut; the knife tended to sheer away, creating a protrusion from the cutting line, or the brittle scales would chip or even ping off of the skin, making dips in the line. It was near impossible to pare, because there was actually very little skin---most of the thickness is from the scales. The best I could do was to cut at an angle to begin with, where necessary. Swearing was involved.
I positioned the pieces so that the scales would get smaller towards the rounded ends of the case, where making cuts for the shape would be more difficult, and made the cuts through the bigger scales in several passes, hoping to score enough on a preliminary pass that subsequent ones could evenly split the scales, which was successful enough after a few attempts. The pieces of the sides of the box came from the sides of the skin; you can see the corresponding losses in the image below. Even though it behaves nothing like a nice bit of goat (doesn't stretch or compress nearly as much, obviously) there is a tiny bit of play in the dimensions and I could mold it to fit once the cuts were in the right place.

I used Paraloid B72, an ethyl methacrylate co-polymer adhesive, to glue the skin to the case; it's removable with acetone if it gets on the face of the skin, and sticks to the metal but not so much that I couldn't peel it off if I needed to re-position. It holds the skin to itself at the cut edges tightly, and when everything is finished we'll put the outer edges down with epoxy, which will bond permanently to the metal. I held the rectangular strip ends in a circle with masking tape while I was working, so I could mark the corners to be cut while on the box, then slip it off to do the cutting, slip back on to check, etc.
The winding holes in the base added an extra challenge. Once the strip was the right size, I glued it in place at one end, wrapped it around, and marked the holes with a permanent marker from inside the box. I used permanent marker generally for marking, because I could get it off again with the same acetone I used to make up the B72---as long as the skin didn't get saturated, in which case (oops) it took the pigment with it and made a nice black spot on the inside. Had to re-do that piece...

Brittany of the mysterious clock tools found some punches the sizes of the holes, which worked really well on the small one and well enough on the large one. The small one came out in a nice clean circle (the edges of which will get covered with a bezel), but the large one (also through larger scales) only cracked at the edges without separating, so I had to cut it with a knife. That was still easier than trying to cut a circle from the start by hand!

Here you can see how the corners came together after the sides were glued down; the cuts themselves were slightly curved, because taking out a pure triangle would leave a little pucker at the point when the sides were pushed together. The edges had to be squished together quite firmly, and I found that an excess of B72 here would fill any tiny gaps left behind. Larger gaps, or holes inherent in the skin (breathing holes of course would not be covered in scales) could actually be filled by cutting a single scale out of a scrap, and plugging it in the loss.

As to the center, it required a complicated shape even though those edges really were quite straight to begin with: the stretching and compressing to make the corners meet resulted in a uneven edge. I used a piece of Melinex, taped in place to the metal so that it wouldn't wiggle around. The same tape sufficed to hold it against the skin while I was cutting; I took the piece out of an edge near the eyes that had smaller bumps overall, to match the small bumps of that edge already on the box. By this time I was swearing less and feeling pretty pleased with myself.

I went over the seams with a rotary drill with various sanding/buffing bits to try to smooth out the scales that stuck out as they rounded the corners. When I finished, there was still a bit left to do: because the box wasn't designed for covering, there was no lip on the bottom against which to butt the skin, leaving this raw edge. I deposited it with Brittany for metals expertise---stay posted!