I have recently been been looking at these
Celtic rings/wheels/bells etc.
I got a lot of data from online sources (mainly Artimede sales of recent years, if I can mention them) and put it into spreadsheet.
The result was surprising - the
weights of these things seem (and I emphasise seem) to be quantised, i.e, they seem to group about multiples of a basic unit
weight of c. 3.7 gm. So far I have (tentatively) identified multiples of 1/2, 1 thru 6, 8, 12, 16, 24 and 40 units. The ring at the start of this
thread fits nicely into the 8 unit group.
It is noticeable that this
weight unit is statistically the same as the
average weight of the
Celtic arrowhead money types of 3.8 gm.
Whether this idea stands up over time remains to be seen, since,
as is well known, humans have an inbuilt tendency to see order when there really isn't any. (There's a word for this but I have forgotten it for the moment*).
My immediate problem is that I haven't been able to devise a measure of the statistical reliability of the apparent groupings, through some sort of Chi-squared test for example, but I am looking into the problem.
Ross G.
* Update - the word is apophenia - see here:
https://en.wikipedia.org/wiki/Apophenia