There is a statement in the above that the roundness was checked with a vernier,this usually does not work on ground steel bar as it was and I assume still produced by using a centreless grinder,it very basically works in a similar way to the travlling steady on a lathe,there is the grinding wheel plus two two support wheels of a reasonable diameter ,one support wheel has an axis parallel to to the grinding spindle,the other wheel has the axis set at an angle so this wheel supports the work plus plus provides endways feed to the bar. If the machine is not set correctly the result will be a bar which is lobed,the lobes will give a constant diameter,but will not fit a circular ring gauge.A quick test is to set two vee blocks on a surface plate and and place a dial indicator on the bar where it rests on one of the blocks any lobing wil show up, When say fitting a lobed bar into its mating reamed hole and it is found to be tight but can still be rotated, the lobing will show as high spots on the bar,sometimes there can be 3 or 5 high spots,may be more and always an odd number. I came across this effect when making travelling microscopes a long time ago,the fine adjustment rods were made from centreless ground stainless steel,to make a batch of rods about 100 ft of steel in 6ft lengths was used and occasionally one bar woud be found with some lobing,our cure was to lap out the reamed holes in the microscope base as the lobing did not affect function on this particular instrument.Silver steel was also produced by centreless grinding and it was very rare to find roundness errors and that produced by Stubbs was very good regarding roundness and dimensionally,