Now I can begin. We safely moved the entire hoard (weighing over a ton) into the lab and I can now work on it in a controlled environment. We deliberately left about 5cm of earth all over the hoard to protect its surface during excavation and transit so the first job is to remove this soil. It's actually a very fine grained clay and I take it off with small hand tools and a mist sprayer to keep the coins damp. After an hour or so we get our first look at a cleaned part of the hoard's surface. The coins are a beautiful deep green and seem in fair condition. Coins can already be seen that show the classic "hoarse and head" decoration of Armorican coinage.
Wow, an exciting find we didn't expect! Carrying along the cleaning of the hoard's surface today I saw something that obviously wasn't a coin like the rest. The coins are made of a mix of silver and large amounts of copper which is why they are coated with green corrosion. The object just revealed has a shiny silver surface. When washed back a little further it was revealed to be a piece of silver jewellery, probably a brooch or cloak pin of some type. It must be almost pure silver as it was uncorroded and needed little more than a wash to reveal its surface. In previous hoards found in Jersey there was a small amount of similar jewellery, often twisted or broken. We won't be able to take the new piece out until we begin to systematically take the hoard apart, probably early next year.
It keeps getting better. I've finished cleaning the area of the top surface I'd chosen and have now starting cleaning the vertical side face of the hoard. As expected it's composed of tight packed coins, obviously still in the same position they were buried, flat against the side of the pit. About half way done the side of the hoard I could see something else, a larger object, flat and twisted. A light wash and clean with a soft brush and suddenly there was gold, gleaming among the dark green coins. In twenty six years as a conservator I've never seen anything like this. You know intellectually that all information we can gather from the hoard is important and that such research is what drives the project but I wasn't prepared for how viscerally exciting finding gold is. It's not its financial worth, it's just spellbinding to see. The object is made of a sheet of thin gold, now apparently twisted and flattened. it looks like this is deliberate but it might have simply been crushed by the weight of coins on top of it. It's uncanny how gold doesn't corrode. It's been there two thousand years and all most of it needs is a wash. In some parts it was coated with a brittle green corrosion layer but this was simply corrosion from the surrounding coins and slipped off the gold surface when touched, leaving it looking as new as something in a shop window.
I've carried on removing the earth from the area to the left of the exposed gold. The crushed gold object is longer than we expected at about 20cm. We've shown pictures of it to various people and it seems it may be a torc (a large necklace). It is made of sheet gold and would originally have been formed over an inner core of wood so that it looked like a large solid gold object. We can see now that it has fine decoration on its ends and was probably deliberately crushed and twisted before burial.
Another gold piece! To the left of the first torc we have found another one. This one is solid gold and what is exciting is that the curve on what we can see of it (only five or six cm) is exactly right for its original necklace size. If therefore it hasn't been twisted like the other torc it might just be complete and unbroken.
I've continued to remove earth from around the hoard. I've now exposed about half the coin surface. No more gold in over a week. I never thought I'd get so spoilt that a week "only" exposing iron age silver coins could feel like a let down.
An interesting find. On the opposite side of the mass to the gold pieces the coins on the surface are very different. Over most of the exposed surface the coins are cemented together by the hard green copper corrosion that has built up on their surfaces. This makes them one hard mass of corroded metal. In this new area however the coins are still corroded but are much less tightly bonded together. There is a fine organic material between them. It's black and fibrous, like peat. What's interesting about this material is that it's only between the coins and not in any of the surrounding earth. This means that the material is the remains of something buried at the same time a the coins themselves. We don't know what it is yet but I'd guess it's the remains of bags or baskets that the coins were carried in. We'll do micoscopic analysis later and hope to get a better idea.
During the excavation of the hoard and in the removal of the earth here in the lab we have collected a large number of loose coins from the hoard. having these loose ones allows me to do something I've wanted to try since getting the hoard here. That is working out just how many coins we think we have in it. There are several ways of working this out but the way I'm doing it is to see how many coins fit into a certain volume and then working out the volume of the hoard itself. I've made a container thats inner walls are 10 x 10 x 10cm. It has a volume of 1000cm3 therefore. I poured in the 364 loose coins, making sure they fitted as tightly as possible and found we had enough to fill 700cm3. This works out as 0.52 coins per cm3. I then worked out the hoard volume by printing a good photo of it's top surface and dividing this into a series of rectangles and squares of known size. The only guess work is the depth of the hoard. In the lab it still sits of several cm of earth and we've never seen its underside. All of the exposed edges show it's about 15cm deep however so assuming it's got a flat bottom that gives up a volume of 127,500cm3. This works out at 66,000 coins. However I'm adding 5% as I know the coins will pack tighter in the hoard than in my container. This gives us just under 70,000 coins which if it's correct makes it Britain's biggest coin hoard and the world's biggest celtic one by a factor of about five or six. Interstingly when I announed this figure to my colleagues I found out one had already done an estimate by weight and one by coin thickness. We had all got the same figure to within 5%!
With all these loose coins on my hands it seemed time to try cleaning some of them. As soon as we knew we were onto a hoard I contacted Julia Tubman of the British Museum Conservation Department. She and her colleagues are working on the Bath roman coin hoard and she very kindly had given me a lot of information about how they treated their coins. As soon as the hoard was up I went over to see them and it was great to see the techniques being used and to get loads of tips which I otherwise would have had to work out myself. Our hoard coins are made from a mixture of about 40% silver and 60% copper. They would have originally looked shiny and indistinguishable from pure silver but two thousand years buried in the soil has made the copper corrode, giving them a detail-obscuring lumpy green coating. The British Museum treatment, whch I adopted, uses a dilute formic acid bath to strip away the green copper corrosion and reveal the original silver surface. I tried this and it worked very well, allowing us to see all the original surface detail that had been hidden. As the BM had intimated however it has to be a very "hands on" process, as coins react at different rates and if the reaction goes too far copper can come out the acid solution and start to plate the coins. Therefore I found I could only do a small number at a time. After stripping, the coins are thoroughly washed again and again to remove any traces of the acid which might make them corrode even after they had been dried.
This is the one point in this blog where I will do the usual museum warning.
PLEASE DO NOT TRY THIS YOURSELVES. I KNOW FROM TRYING IT HOW EASY IT IS TO GO TOO FAR AND DAMAGE COINS. IF YOU HAVE ANY HISTORIC METALWORK TAKE IT TO A MUSEUM AND ASK THEM. DO NOT TRY DIY CONSERVATION, YOU'LL RUIN IT.
I showed images of the cleaned coins to our coin expert colleague Dr Philip De Jersey (he must get some stick living in Guernsey with that name!) and he said that they were all Coriosolite coins. This was the tribe who controlled the area of the french coast closest to Jersey at the time of the roman invasion.
I had a first microscope look at the organic material between the coins today. This is the blacky peaty stuff that may be the remains of the bags or baskets the coins were moved in. Using our Olympus compound microscope at magnifications of 100X and 400X I could see quite a lot. This is not my area of expertise but some of the material was clearly plant remains, with the cell structure well preserved. Other parts of it were animal hairs. The most interesting find on this first look was the leg of an arthropod. This must have been extremely small, like a mite, as the whole leg was visible at 100X. Preservation appears to be very good so hopefully we will find a lot more inside the hoard. I'm sending the initial images I've taken to the British Museum in the hope that they will be able to make more of them than I can.
This image resulted in the first truly dreadful hoard joke I've heard when I was asked if this proved we'd found the "Mite of Rome"
We've just had an absolutely mad weekend showing the hoard to the public. We thought the people of Jersey deserved a look at it while it was still a single block so we arranged a three day show at Jersey Museum. We displayed the hoard out in the open in our museum concourse and always had a member of staff next to it at all times, frequently spraying it with water to keep it damp. For security reasons we didn't advertise the event until a couple of days before but the response was amazing. I've never seen a queue outside the building waiting for opening time before and from then till closing time it stayed solid. We gave a talk with video and still images every couple of hours and the museum theatre was packed for all of them. I did this with Olga Finch, our curator of archaeology and Reg Meade and Richard Miles, the hoard's finders, and in the end we had over three thousand people come to see it. We heard about so many people who missed it that we may repeat the show again before Christmas.
Today we have had a demonstration of a Faro laser scanning/metrology arm. As its name suggests this is a device which can both laser scan the surface of the hoard to build up a 3D virtual model but can also measure the position of any point in the hoard to within fractions of a millimeter. We hope to purchase a unit like this to use as we deconstruct the hoard, logging the original position of every object that comes out of it. As far as we've been able to find out, no hoard of anything like this size has ever been fully recorded as it's taken apart coin by coin. If we do this we will be able to know exactly where every coin or piece of jewellery fits in relation to all the others, even once they are all in seperate bags or boxes. This will allow us to investigate the hoard after its dissasembly so that we can look up positions or groupings of coins by type, weight, condition etc. This way we should be able to tell how the hoard was deposited. Whether for example it came in twenty bags or fifty. Whether they all came from the same tribe or several and if so, which tribes left the jewellery etc. The Faro unit is a fantastic tool and will allow this research to be done in a way that would have taken far too long by any other technique. As a bit of a gadget fan I must admit it was also a great toy and I kept finding myself looking around the lab to find things to scan.
It's hard to know just how long the project will take but we are extrapolating out from the British Museum's experience with their roman hoard and think there might be six to eight years or work for one person. We will try to get the work done by a team therefore in order to complete the hoard in a more manageable time. If all goes well we will employ two new staff members next year who will work with me on the hoard's disassembly for perhaps three years. The work will be split between the chemical/physical work of cleaning the coins and the recording work of scanning, position logging, photography etc
Today we had a visit from Doctors Karl Harrison and Andrew Shortland from Cranfield University. Karl has a background in forensic archaeology and is very interested in the hoard in the context of its landscape and was keen to see the burial site. His colleague Andrew was interested in the material science side of things and rather splendidly brought over a handheld X Ray Fluorescence device. This fantastic machine is carefully pointed at the surface of any desired metal/ceramic etc. and will give you a readout of what percentages of elements it is made from. In an hour's work Andrew was able to tell us that the bulk of the coins were about 40% silver/60 % copper, that the two large gold objects were indeed 75 and 80% gold and that the bowl of corroded metal recently found at the hoard's lower edge was bronze! I want one! Sadly the machine is beyond our price range but are talking with Cranfield about one of their students or staff coming over periodically to do a 5% XRF survey of all the coins as they are cleaned. Hopefully Cranfield can also work with us to do a geophysical survay of the area where the hoard was found so that we can get our first idea of how the site relates to nearby buildings, roads etc.
There was one of those interesting moments when you get a glimpse of someone else's expertise when I was talking with Karl Harrison. There is a feature on the hoard's surface that is hard to see in photos but looks very much like a footprint when seen close up. I'd rather hesitantly put this idea to Karl, not wanting to look like a burk if I was wrong but he immediately said yes, he thought it was a footprint. Then in a matter of seconds he said it was a left foot, that the owner's other foot was outside the hoard pit, the man was bent down and that he'd been smoothing about 2-3cm of earth over the top of the hoard with his hand at the time. Very impressive, very Sherlock!
The XRF machine.......I want one
We've just had another weekend where we put the hoard on show for the public again. Once again very successful with good numbers of locals. This time in addition to Reg and Richard, the finders, we also had another local metal detector, Ken Rive, with his new hoard. They're just oozing out of the ground this year in Jersey. Ken's is a pot full of 25 late bronze age axe heads in beautiful condition. We excavated it with him, bringing up the pot, still full of axes, by first wrapping it in bandages as it was exposed and then, after protecting it with clingfilm, encasing it in a block of polyeurethane foam. After X raying it to check its contents I then took the whole thing apart, cleaned all the pot pieces and reconstructed the pot. The axes were washed and then treated with a solution of benzotriazole in alcohol to stop further copper corrosion. Don't try this at home, benzotriazole is very toxic and must be handled and disposed of correctly. The axes were then given a coat of protective lacquer, matted down so as not to affect their appearance.
Still a bit of a quiet time with the coin hoard. We are all waiting on developments over the next month or so that should allow us to start the conservation work in earnest. We did have a slight wobble with the coin numbers last week however. For some time I'd been thinking that my initial estimate of their numbers (70,000) was begining to look a bit high. This is because it was based on the hoard being an average of 15cm thick, top to bottom. The more revealing of its base I did the more it became apparent that it probably averaged closer to 13cm. This shouldn't matter of course as only the accurate figure is important but it was with a slightly heavy heart that I re-did the maths and came up with a new, lower figure of 61,000 coins. Still Britain's biggest hoard but coming perilously close to the Frome hoard of 52,000. Competitive, me? My worries were put to an end a couple of days later however when I was thinking about the hoard's base. As it still sits on a layer of earth, we had just worked on the assumption that it has a flat bottom. I realized all I had to do was dig a small trench under a part of the hoard to check this. I duly did so and found that the belly of the hoard is actually boat shaped, going down another 6cm from the bottom of the hoard's walls. Some rough maths on this depth adds slightly more coins than I took off a few days before. Back to a conservative estimate of 70,000 then, hurray!
The hole under the hoard revealing its new base level
I've just finished an experiment making a replica of a part of the hoard's surface. We want to have a replica of the hoard "block" as excavated because the conservation treatment will of course turn this into 70,000 seperate coins, probably spending most of their lives in tupperware boxes in a store. To have a permanent record of it in one piece we decided therefore to laser scan the entire surface so a to have a virtal 3D replica but also to make a more old fashioned physical cast. In order to do this I painted a liquid silicone rubber onto a part of the surface and let this harden after a couple of hours. I then applied a second coat of rubber, this time thicked to a consistency like Icing. Once this too had gone hard after twenty four hours I was able to peel this mould off the surface of the hoard. This was the important bit as it was a test to see if the rubber would damage the surface and pull off coins. In the end only one coin came loose and it may actually have been loose before. The coin was carefully laid back in place for later recording.
I now had silicone rubber mould of about 15 x 20cm with which to make the copy. I did this by painting a thin epoxy resin onto the mould's surface. Once this had set I applied a second layer again thickened almost to a paste. Once this set I peeled off the rubber to reveal the new resin copy. The detail was very good but the resin was of course uncoloured. I therefore painted the surface with a variety of enamel paints to copy the various browns and greens. The whole excercise worked well and means that quite soon I'll go ahead and do the whole hoard. It's so big that I will have to cast it in sections and join them later but this shouldn't be a problem. What is a problem however is the very fragile gold jewellery. The thin crumpled gold sheet would be dragged right out from the hoard with the rubber if I tried to mould it so I shall have to leave these areas alone. Plan B is to have these sections laser scanned and resin 3D prints made from the results. I will then make epoxy casts of these areas from their prints and incorporate them into the whole replica structure. I will probably gold leaf the jewllery in the replica to get a good finish.
The resin copy sits here on top of the real hoard. Can you see the join?
Last week I went ahead and did the moulding of the hoard's surface in order to produce a replica. I did it in the same way as the test piece described above, the only difference being that I inserted a flexible plastic foam into any crevice that would have "trapped" the rubber too much and caused damage on pulling it away. For instance I filled the inside of the gold "bracelet" loop and behind the gold torc. When the replica is cast I will grind away the replicated foam areas to match the original surface.
The protective foam applied under fragile areas
In the end the best way to do the whole surface proved to be in three sections. I did the entire top surface of the hoard in one go, making sure that I overlapped the side wall edges slightly. Before it set I inserted a long strip of 150mm high clear plastic sheet into the rubber all around the hoard's top edge. This acted as a wall all around the top of the hoard. This was needed because I had decided to use polyeuethane foam as a backing for the rubber. A backing is needed because although the rubber preserves every detail on the hoard's surface, it is very flexible and floppy. By itself therefore it wouldn't preserve the hoard's shape well. When it sits on a hard foam backing that is moulded in place however it is rigid and accurate. The foam is made by mixing two resins and pouring them onto the rubber, allowing the mix to then start bubbling, expanding and then hardening. Doing this in a series of small stages I built up a 100mm thick rigid foam backing. Once this was hard I then pulled it off the rubber and then carefully peeled the silicone rubber sheet off the hoard top. Pulling it off in this way, however carefully, did dislodge more coins than I was happy about but each was placed back on the surface where it came from. Although the process has changed the surface of the hoard slightly the replica produced from it will preserve the original surface perfectly for all time.
The rubber mould of the hoard's top on its foam support
Having done the top surface I then went onto the sides. I did about 80% of the edge in one long rubber strip, leaving only the area with the exposed gold torcs. I inserted a protective polyester wall again to protect the top surface coins and I then backed the side in foam in the same way as the top. I later cut the hardened foam off in four sections but left the rubber in one piece which I peeled off as before. When I come to do the casting I intend to make the edge strip in one go but will do the work in four sections, each bit suppported by the appropriate foam backing. The removal of the rubber did little damage except for one area to the left of the gold pieces where a section of the surface about 2cm deep and 5 by 10cm wide did fall away. The only remaining area left to do was the one with the gold torcs. I had initially thought that the rubber removal would be too damaging to allow its use here but I worked out a safe way to do it. All the areas so far had been given a deeply penetrating thin liquid coat of silicone rubber and it was this penetration that caused some damage on removal. For the gold area I did not use the thin liquid rubber but went straight onto the thickened paste-like mix I had previously used for backing the first layer. This paste was applied with gloved fingers and a small trowel and was not pushed deeply into the surface. This led to some compromise of deep seated detail but allowed easy removal with no damage to the jewellery.
Here is the rubber strip and foam backing still on the hoard. The gap at the top is the area with the gold torcs
Pulling all the rubber strips and sheets off was the scariest thing I've done since we lifted the hoard out of it's original hole in the field. I'm pleased with the results however and will start producing the epoxy cast after easter.
Hello, sorry it's been so long since my last blog. This has been a combination of illness and the fact that the making of the replica has been the only real progress we've made on the hoard in the recent months. As to its actual conservation, we are still waiting on funding/permission issues to be resolved. I have however finished the replica which turned out, as ever, to be a bigger job than anticipated. The making of the moulds themselves went OK and they didn't need too much cleaning up before they could be used for the casting. Once again I began with the hoard top as this is a nice flat area. I mixed up a pourable epoxy resin in 200ml batches and painted it onto the silicone rubber mould with a 2" brush. I let a first thin coat harden before I applied a second. This time I stippled in strips of glass fibre to strengthen the cast. I repeated this process till there were generally about three layers of glass and a total epoxy thickness of between about 7-10mm. I did the sides in the same way but in the end I had to do it it more sections than I had originally planned. The problem I hadn't considered was gravity. Painting the liquid epoxy onto the silicone rubber curved mould meant it kept pooling at the bottom of the area I was doing. In the end I produced lengths of about one meter.
Three replica sections
A close up of the gold torque area
The epoxy cured quite slowly, strong the next morning but still flexible, only going totally rigid after a week or so. This proved useful as it meant I was able to cast each of the pieces and begin to join them up while they were still somewhat flexible.
The joining was trickier than I had anticipated. Once again I had not foreseen something, specifically that the removal of the top surface rubber mould dislodged some of the edge coins. This meant that when I moulded the abutting side sections those edge coins were no longer there so the two overlapping areas of the final replica sections weren't identical. In the end I cut the sections with a Dremel tool to get the best possible fit, then glued and filled the joins with the same epoxy resin. Finally I reinstated "missing" coins from the edges with epoxy resin copies I'd cast from elsewhere on the hoard, using high resolution images I previously taken to place them correctly. Written up in one paragraph this all sounds very controlled and well thought out but it was a tricky process and at one stage I was using every cramp, G clamp and length of rope I could get my hands on to hold the whole thing together while it set.
More clamps, now!
Once it was all together I went round the bottom edge, again with the Dremel to give it a flat bottom so that it would stand on a table etc. I was worried (unnecessarily as it turned out) that the whole thing would flex too much so I then filled its interior with expanded polyeurethane foam. Then came the fun bit, painting it. As with my test piece earlier I used Humbrol enamel colours. I airbrushed a brown base coat all over to replicate the soil between the coins. I then painted the green coins by hand. I referred to images of the hoard and replicated the correct appearance of the individual coins to a reasonable degree. I applied several different shades of green to replicate the varying corrosion types. I tried gold paint for the gold pieces but it looked flat and unrealistic so in the end I gilded those areas with 24Ct gold leaf which, as you might expect, looked much better. As a final touch I painted a very watery mix of mud from the original hoard over the whole cast and then wiped it off with a microfibre cloth. This give a very effective finish so I then sealed the whole surface with a satin finish lacquer which gave it a good "damp" finish.
Replica surface showing gilded and painted finishes
Me as Obelix, doing something I couldn't do with the original