Monday 24 April 2023

Bigging up the prefixes

I was doing more lonesome driving than usual recently and tapped into a seam of Tim Harford's BBC More or Less episodes. One was about the naming of parts prefixes for increasingly large and small numbers. If your audience is numerate it's quickest to read and most reliable to evaluate to use suffix superscripts to indicate the size of a number. Avogadro's number is

  • a) the number of molecules in a mole [= gram-molecular-weight not Talpa europea yeh daftie]
  • b) 6.02 x 1023  if you can manage superscripts <sup>23</sup> in HTML
  • c) calculators often render this as 6.02e+23
  • d) 6.02 ^ 23 is also acceptable to sciency ppl
  • e) 602,214,080,000,000,000,000,000

But these are all hard to interpret in audio mode, so there Avogadro's number is 602 sextillion. Seems that until recently Système International SI units had no official prefixes for numbers larger than 1024 or smaller than 10-24. But stellar distances and computer storage is driving to agree terms for counts larger than this. For a while ten years ago, different teams were pushing for the adoption of Hellabyte or Brontobyte for 1027 bytes but this got side-eye from the SI Metrology committee because the one letter abbreviations H for henry the unit of electrical conductance and B for byte were already spoken for; and that would lead to needless confusion.

At a November 2022 SI meeting is Versailles [France not Vərsaylz, KY yeh daftie] prefixes for 1027 ; 10-27 ; 1030 and 10-30 were agreed. Here's the full list:

quetta  Q    10^30 ** quecto  q    10^-30
ronna   R    10^27 ** ronto   r    10^-27
yotta   Y    10^24 ** yocto   y    10^-24
zetta   Z    10^21 ** zepto   z    10^-21
exa    E    10^18 ** atto    a    10^-18
peta    P    10^15 ** femto   f    10^-15
tera    T    10^12 ** pico    p    10^-12
giga    G    10^9 ** nano    n    10^-9
mega    M    10^6 ** micro   μ    10^-6
kilo    k    1,000 ** milli   m    0.001
hecto   h     100 ** centi   c    0.01
deka    da     10 ** deci    d    0.1

As a camel is said to be a horse designed by a committee, the SI General Conference on Weights and Measures has delivered a bit of a kludge. For the higher / more recent designations there is a neat symmetry: Q R Y Z reflects q r y z.  But P peta 1015 is not the reciprocal of p pico10-12 nor is M mega 106 the reciprocal of m milli 10-03

And I can tell you that μ for micron is a typographical disaster for students and scientists being far too frequently replace by u or "mu".  But my annoyance is really focused on making two-letter da the official SI abbreviation for deka when the most recent meeting redoubled the precedent for capitals being big Q R Y Z with l.c. letters being q r y z. Historical inertia has fixed k for kilo and h for hecto rather than K and H. No room for complacency here.

Strict SI logic would suggest that 10^30 should be quecca but queca means fuck in Portuguese, so they went with quetta - a city in Pakistan - instead.

No comments:

Post a Comment