APPROXIMATIONS
APPROXIMATIONS
In principle, even if one uses the apparatus as described,
there is a disturbance due to gravity.
More precisely: the diffusion process is slackened
by gravity.
Haskell [3] measured the systematic error caused by this effect and
demonstrated that it was negligible.
What does it means: "an error is negligible?"
A systematic error is negligible when it is much smaller than the error due to measurements executed in that particular experiment.
A priori, the use of any instruments inevitably involves the introduction of systematic errors.
It is possible to estimate the total systematic error which will affect
the measurement (diffusion coefficient, in this case) considering the
instruments used in the experiment and all possible effects due to
physics.
Any factor which induces a systematic error lower than those due to
the instrumentation is negligible.
If one wants to obtain more precise measurement in the same experiment
one must use a more sensitive instruments
The resulting total sistematic error
will be lower and some error that was neglected before,
could then become relevant.
The method described for the calculation of
the diffusion coefficient is not rigorous
because we consider constant concentrations in the two reservoirs
but on the same time we need a variation (even if little) to measure the
flux.
However the results obtained can be used in many cases.
Back to the measurement of the diffusion coefficient with Clack's
apparatus.