Let's review the history of measurement units and what they were defined from.
Time: Second
Original: Mean solar day
1956: Year (a specific one)
1967: Cs-133 ground-state hyperfine transition
Lest someone think that this makes the second a "physics-based" unit, I'd put it like this:
Time: Second
Original: a particular fraction of a Mean solar day
1956: a particular fraction of a Year (a specific one) (arbitrary fraction chosen to be close to original definition)
1967: a particular number of Cs-133 ground-state hyperfine transitions (arbitrary number chosen for consistency with 1956 definition)
Length: Meter
1795: Earth size
1799: Platinum bar, later platinum-iridium bar (arbitrary length chosen to be close to previous definition)
1960: a particular number of Kr-86 electronic transitions: 2p10 - 5d5 (arbitary number chosen to be close to previous definition)
1983: Second with c fixed
This makes it clear that the units remain arbitrary, despite the use of non-arbitrary physics to define/measure them, and it explains how multiple unit systems can coexist and differ. For example, the meter is the distance light travels in vacuum in 1⁄299,792,458 of a second. The inch is the distance light travels in vacuum in 1/11,802,852,677 second. You can't derive either of those numbers from the laws of physics; they're both arbitrary human choices.
The SI second is the duration of 9,192,631,770 cycles of the 133Cs hyperfine transition. The French revolutionary decimal second was 7,942,433,849 cycles of the 133Cs hyperfine transition. You can't derive either of those numbers from the laws of physics, they're both arbitrary human choices.
Unless you're Andrew Worsley, in which case 1⁄299,792,458 and 9,192,631,770 are important inputs to how Nature determined the mass of the electron.
Last edited: