Unit based on a given volume of water
All
the schemes related to the new system planned the connection between the
mass unit with the volume unit, that is to say the unit of length.
Lavoisier
and Haüy had determined the "grave" in 1793, the mass of a cubic decimetre
of water at ice melting point, for which the value was alledged to be 18,841
grains of the average marc of the Pile de Charlemagne.
This
work was entirely resumed in the early 1799 by Lefèvre-Gineau and the
Italian Fabbroni.
The
two scientists chose water, not at ice melting point but at a 4
centigrade-degree temperature, that is to say the temperature of maximum
density for this liquid.
They
weighed successively in air and water a hollow brass cylinder whose size had
been carefully determined thanks to a comparator that Fortin purposely
created. They deducted from it the mass of a cubic decimetre of water that
was distilled at its temperature of maximum density, that is to say the
kilogramme. This mass was found to be equivalent to 18,827.15 grains of the
average marc of the Pile de Charlemagne.