search  current discussion  categories  glazes - chemistry 

specific gravity and density--two errors

updated wed 3 dec 97


Stuart Altmann on tue 2 dec 97

My November 25 posting on specific gravity and density--written in haste
before I left for a Thanksgiving family reunion on the farm--contained two
errors, caught by several Clayarters. As before, let me begin with the
take-home message for potters.

(1) Specific gravity (s.g.) differs from density not just in being
dimensionless. The values also differ by a factor of 1000. To convert your
s.g. values to density values, just multiply them by 1000. They are then in
standard SI metric units of kg per cubic meter. Conversely, if you have
many records in s.g., you might, for comparability, want to divide your
density values by 1000--giving you "millidensity," I suppose.

(2) To determine the density of a glaze or other liquid, weigh a liter of
it in grams (not kilograms as previously indicated). That is its density,
since a density of a gram per liter is equivalent to a kilogram per cubic

(3) By using density measurements instead of s.g., the "problem" of dealing
with units that seem too compressed--a complaint of some about s.g.--
disappears, as does the justification for degrees Twaddell. In any case,
density, not specific gravity, is the standard SI metric unit, and so should
be used instead.

Now for the details. In my previous post, I said that a liter is also a
cubic meter. Wrong: a liter is a thousandth of a cubic meter. So the
density of water is 1000 kg/m^3^ (not 1 kg/m^3^) and the definition of
density is not equivalent to a kg per liter but to a gram per liter.

You may be concerned about the shift from measuring volume in cubic meters
to measuring in liters. Don't be. Just remember that the volume of 1000
liters is equivalent, by definition, to the volume of a cubic meter.
Previously, a liter was defined as the volume occupied by 1 kg of pure water
at its maximum density. Rounding errors in measurements resulted in a
minute discrepancy: a cubic meter equaled 999.97 liters, not a thousand.
However, in the SI system, the liter has been redefined to be a thousandth
of a cubic meter. In any case, a difference in the fifth significant figure
is beyond the scales that potters use, and we certainly don't use pure water
at 4 C!

My other error was in an example that I used, which implied that water added
to a glaze would increase its density. Wrong: adding water to a glaze
reduces its density, because the substances that we add to water to make
glazes are denser than water.

My thanks to several Clayarters for their clarifying questions and comments:
Tom Buck, Terrance Lazaroff, and Denis Whitfield, and my apologies for any
confusion my errors produced.

Stuart Altmann