From the supplied photograph the same ratio of error appears to occur in the second octave even though those distances are not the same so it is unlikely to be a mechanical error low down or the repeat would completely mis-align due to the larger measurements in the lower octave.
Accurate charts for 12TET were only published by William Braid-White in 1917, indeed the first accurate sub-divisions into 'cents' had only been done by Ellis in 1864 using tuning forks. What he found was that people claiming to be tuning in ET were not in fact doing so (1885) If the board patterns repeat the ratio of error further up but not the scale, then we are most likely dealing with a different temperament approach, possibly taken from earlier patterns & decisions, rather than some simple mechanical loosening or flexibility during the Loar period as postulated by some.
Bookmarks