Hi and welcome David,

Just picking up on the methodology of using the Ordnance Survey as a basis for your measurements, particularly the very small scale 6 inch map (1/10560).

I'm sure you're aware of all of this, but the thickness of a line on the 6 inch map equates to about 10 feet in width, so using a map of that scale as a basis for a survey of stones (that are presumably a couple of feet wide) is problematic. An error of a millimetre in your placement on a 1/10560 map would move the stone about 20 feet (sorry for mixing metric and imperial!).

The modern OS map at 1/2500 (equivalent to the 25 inch map) has a relative accuracy of a couple of metres. Rather less for a 25 inch map surveyed using chains back in say the 1950s. A couple of metres accuracy is obviously pretty good for a map that you want to show the relative position of buildings, roads, etc, but not so good if you want accurate placement of a smallish stone in a big circle. So if you've taken measurements on the ground and plotted them onto a 25 inch (or 6 inch) map, depending on how you've done the surveying and whether the ground slopes, which of the features on the map you have chosen to "fix" your measurements against, etc, you can plot your stone confidently to about 2 metres of where it actually is on the 25 inch, but probably only to 10 metres or so on the 6 inch (I'm not sure what the tolerances and relative accuracy of a 6 inch map would be, but quite low).

That still leaves quite a degree of variation to accurately plot the circle with confidence to show a "true" circle. This gets even worse if the distance between the stones is considerable, which your measurements for the circle's diameter would indicate.