Meh. To me, it just doesn't seem like that much of a mystery. Every single man-made process involves tolerances, and when these things were manufactured (and, I suspect, today) we were unable to count the individual platinum atoms that were going into them. We can't make a perfect vacuum, and we were probably less able to make a perfect vacuum back when the prototypes were being created. I'm sure there were quite a few errand molecules floating around in the near-vacuum surrounding each kilogram, and some of them probably settled on them, nestling in interstices between platinum atoms. And remember, we are talking about (in the case of the U.S. kilogram), a difference of 19 parts in a billion
Like ko_kyi, I am far readier to believe that this discrepancy is due to the inevitable imperfection of human manufacturing, storage, and measurement processes than some fundamental flaw in scientific theory.
But I would assume from the story that the LAST time they were compared, the differences were “different”. Are we really that much better at this now than we were a couple years ago when they were last compared?
And the chart suggests that all the variations have been in the same general directions. That actually lends credence to the idea that there were more extra particles in the seals for some than others. But do they break the seals to do the mass check?
Maybe there is a mundane explanation — there should be one, because this isn’t an area where we have a lot of unknowns.
But it is certainly a curiosity.