American Scientist presents an interesting proposal in its Macroscope column this month. The authors want to explicitly redefine Avogadro's number (~6.022 × 10²³, as you may remember from physics) to be a specific integer.

I remember learning that this number represented a mole, the number of molecules of water in a gram. It turns out, of course, that this depends on the definition of a gram. At the moment, the gram is defined by a specific chunk of metal held in a vault somewhere in France. This chunk of metal (Le Gran K, according to the article) is changing over time (due to handling, cleaning, etc), so the actual definition of a gram is changing, as is (consequently) Avogadro's number. Redefining Avogadro's number to be a constant would remove the Système International's dependence on this physical artifact.

I'm fascinated by our obsession with clean definitions (i have the same obsession myself), and that there are whole bureaus of people out there who just compulsively clean things, measure things, calibrate measuring equipment, and write up reports about it. The idea that we could make an explicit decision to redefine our measurements qualitatively (i.e. making the number the reference, and the physical thing the approximation instead of the other way around) is pretty cool, too, even if it could put some of these people out of work.