This has some interesting implications, not generally recognized. The thermodynamic
entropy S(X) is, by definition and construction, a property not of any one microstate, but
of a certain reference class C(X) of microstates; it is a measure of the size of that reference class.

Then if two different microstates are in C, we would ascribe the same entropy to systems in those
microstates. But it is also possible that two experimenters assign different entropies S; S0 to what is in fact the same microstate (in the sense of the same position and velocity of every atom) without either being in error. That is, they are contemplating a different set of possible macroscopic observations on the same system, embedding that microstate in two different reference classes C; C0.

‏¹ http://en.wikipedia.org/wiki/Krypton
‏² http://en.wikipedia.org/wiki/Kryptonite

Oh come with me
to watch the first RADON
When the stars ARGON
As the day KRYPTON
And if the morn be cloudy
You won’t XENON.