Re: RADSAFE digest 1231

[Dick King <king@reasoning.com>]

> 
> Date: Thu, 16 Jan 1997 12:26:32 -0700
> From: James.Langsted@rfets.gov (James Langsted)
> Subject: SI Units Stumper
> 
> RADSAFE'rs:
> 
> The following question was printed in the "Wacky Questions" column of Denver's 
> Rocky Mountain News, in a year-end column subtitled "Queries that stumped the 
> staff."  In the column, Rebecca Jones (Miss-Becky@aol.com) and her staff 
> research and answer all sorts of questions.
> 
> She's stumped on this one, and I am curious as well.  From previous postings, I 
> know there's a group of units experts reading this.  Post your thoughts on the 
> list and I'll forward them on to the columnist (with credit to you) or e-mail 
> her directly.
> 
> Thanks
> Jim Langsted, CHP
> MH Chew & Associates
> Golden, Colorado
> 73517.351@compuserve.com
> 
> "In the International System of Units, almost everything is defined in terms of 
> something universally observable, like specific atomic wavelengths or 
> frequencies.  There are two exceptions.  The kilogram is defined in terms of an 
> artifact (which is, I believe, a platinum/iridium cylinder that resides in 
> Paris).  The other is the mole, which is defined as the number of atoms in 12 
> grams of carbon-12, and is thus derived from the kilogram.  It seems to me it 
> would make more sense to just specify Avogadro's number once and for all, use it 
> as the definition of the mole, and define the kilogram in terms of the mass of a 
> mole of carbon-12.  Why doesn't it work this way? -- J-J Cote, Boulder"
> 
> 
> Date: Thu, 16 Jan 97 15:19:56 -0500
> From: Lester.Slaback@nist.gov
> Subject: Re: SI Units Stumper
> 
> As opposed to time or length which are people invented devices the number
> of atoms as specified by Avagadro's number is set by nature.
> 
> Doing that would be someting like defining Pi to be 3 without
> appropriately changing or number base.

No.

Time is a unit created by nature; a day is a particular length and there are
86400 seconds in a day.  If you define the second by defining some spectral
line to be so many hertz you must make that measurement extremely accurately
lest the sun refuse to reach zenith at noon two days in a row: as it is we add
a leap second to the beginning of the year every now and again because either
the measurement was imperfect or the Earth slowed down.

Length units are, indeed, created by man.  If you defined a meter to be the
length of the UN chief's foot then we would have to recalibrate all of our
instruments and rewrite all of our textbooks when the UN changed
administrations but nothing more ;-).  

By defining Avogadro's number by fiat you would actually be specifying the
gram; if i said that Avogadro's number was 1 then a gram would be the mass of a
Hydrogen atom [approximately].

Now i will answer the original question:

The problem with defining Avogadro's number by fiat is not that we would be
attempting to tell an atom what it weighs.  We would be telling an atom what
number _we_ will assign to its weight; that is our privilege.  

We always define units in terms of the most reproducible measurements, whatever
that may be at a particular technological level.  The length standard was
changed from a metal rod to a spectral line when the precision of
interferometry began to exceed the precision of measuring metal rods [and the
consistency with which the rod maintained its length, which is not perfect].
The time standard was changed from Earth's rotation to a spectral line when we
came to be able to measure time by a spectral line well enough to show that the
Earth was a mediocre clock.  However, we can measure the mass of a particular
standard with more precision than we can measure the mass of a Carbon atom; if
we can measure the mass of a kilogram to, say, eight significant figures but
the mass of a Hydrogen atom to only six significant figures then two
first-generation mass measuring devices will agree with each other to 20PPB if
we use a metal standard, but only to 2000PPB if we use an Avogadro's number
standard.  The need for second- and third-generation standards if we use a
single standard mass only costs us another 20PPB or so.

When we come to be able to measure the mass of an atom with about a third of
the precision we can measure the mass of a metal standard, we will switch over.
Indeed using a physical standard has the advantage that there need be no
second- and third-generation standard objects, but that is an advantage that
can be outweighed if the physical standard constitutes a difficult measurement.

-dk