I started this post 2 years ago, but clearly it is time to do an update.
Since then I’ve not managed to perfect new whorl moulds, but have done more research into them, which has thrown up a few things.
The critical issue I have had is in making holes for wooden spindles that are vertical and at right angles to the faces of the mould stone. This really needs a column drill or lots of skill and practise, so in my big mould of new whorls, things aren’t quite right and this has slowed my production of new designs of them.
The other issue is running out of good stone to carve them from. Tracking some down is difficult, and always has been. I need fine grained stone which is moderately soft, so that I can carve it without too much trouble. I don’t want to pay lots of money for carefully cut soapstone online.
This isn’t usually too much of a problem, but some skill is needed because I have made the moulds out of irregularly shaped stone. Sometimes I have to tip the mould at an angle such that it fills from the bottom up and the air is pushed towards the to, which is really important when doing the large 14th century one.
The biggest thing I have found is the 2016 paper by Eleanor Standley (Spinning Yarns:The Archaeological Evidence for Hand Spinning and its Social Implications, cAD 1200-‐1500), which is basically a literature review of archaeological information about spindle whorls in the medieval period.
It is interesting, informative and wide ranging. She has done a good deep dive into the literature and available sources, covering the materials used, when they were used and puts forwards some ideas which are worth thinking about. If you are interested in spindle whorls you should go and read it.
However there is one major howler that I have noted within it.
She refers to half a mould found in Dunkeld,
and suggests that the lead was poured into it and wiped off the top, then this was repeated and the two halves soldered together. This is a strange thing to suggest, because of the many ramifications of it and just how hard it would be to do. Firstly you have to get both sides of the whorl to a nearly liquid temperature. Which would be very fiddly to do. Then you have to stick them together without any obvious overlap or joining ridge, which would be almost impossible. If we had found a lot of them that were slightly mismatched in terms of one side or the other being offset, that would be one thing. Or else, if they were slid down a wooden stick so that they matched together perfectly that might solve that problem but still leaves the issue of the perfect join that they exhibit.
This method of course would also work only for biconical ones, not the asymmetric ones which are seen more in the high medieval period.
There is a general lack of lines marking joins in moulds in lead spindle whorls, you can see this even in ones I’ve made myself. I think this is partly to do with how viscous liquid lead can be, and also how it shrinks when it cools.
Another important thing to note with manufacture is the case of whorls which have two different patterns on them, as if two different halves which fitted together were used. It does rather suggest to me that the manufacture was carried on by a small number of professionals whose moulds were well made and interchangeable in some ways. To me this is much more likely than two different halves made in different moulds and soldered together. These photos are taken from the Portable Antiquities scheme Finds database, finds.org.uk and show either side of the same spindle whorl, as tweeted by Graham Rawson last year or so.
Anyway, back to Standley.
The condition of the Dunkeld mould, as just one part of a probably set of moulds, is entirely normal for stone moulds for metal casting. In the UK we don’t have many stone moulds; one of the biggest collections is at the Herbert Museum in Coventry. I have examined their moulds, and they are mostly 1 or sometimes 2 parts of what were clearly 2 or 3 part moulds. This is the case almost everywhere there are stone moulds, only 1 out or 2 or 3 parts has been recovered. It seems a causal factor for their disposal in a context we eventually excavate, that most of the mould is broken or lost. So in this case I see no reason to confuse matters by thinking that whorls can be made using just one side of a mould.
She tries to bolster her case by referring to whorls found that are clearly incomplete or damaged, and registered on finds.org.
I have examined all the relevant ones that I can see on there, and they are more like mis-cast ones that I have made using a 2 part mould. The strange shapes are what you get when the mould is opened when the lead is still liquid, or not held together properly or not enough metal is poured into it. A key point to note though is that I have been informed by my expert spinners that it doesn’t matter if the whorl is not balanced, it will keep turning nonetheless. Plus they have seen many partly hollow metal detector finds themselves, which were clearly poor castings yet were still sold on and used.
Standley does rather leap ahead of me in her paper with reference to post-medieval use of spindle whorls. This I had heard about from Beth the weaver and after some thought reckoned it was very likely, and Standley’s linkage of it to lead whorls found more in certain parts of the UK is definitely worth pursuing further. She also points out that there are dateable imported ceramic whorls that clearly continued in use for a long time after the medieval period, making it likely that the lead ones continued in use too.
So my next step, when we are allowed to meet other people, is to use a friend’s column drill to make perfectly vertical holes in soapstone in order to make 2 or 3 different whorls of greater perfection than before.
I note too that there is someone selling recreated ones for £12 or more on ebay, clearly I am in the wrong job because I only charge 2 or 3£ to people I know.
I suspect also that bronze moulds might be used, especially for more early modern ones, enabling thousands to be made without any trouble. Ultimately it would be nice to compare all the whorls on the finds database, for different sizes and patterns and shaped. As of writing this, April 2020, there are 6,602, of which 3,970 are lead. Not all will be medieval, but I think most will be that or post-medieval. I believe further research on the regional distributions of patterns is underway or being considered. In the meantime I shall continue to research and make my own ones.