D O C U M E N T 3 8 M A Y 1 9 2 0 1 7 7
the system is already completely determined by those attributes, but perhaps Miss
Edith ought to work this out a bit more. For as it stands, although one can always
arrive at a constancy of by suitable choice of the density distribution, it is not
convincing to everyone that this alternative is the right one. The complete theory
would have to be able to indicate from a constant the full distribution of tem-
peratures, densities, etc. I fear, though, that Miss Edith will not be equal to this task,
and think that an explanation that only one more constant was involved here can
make the matter plausible.
How would it be, though, if one regarded instead of a single layer the whole
breadth between the boundaries? The entropies of the various layers cannot be
summated, of course, as no equilibrium, in the normal sense, subsists. But one can
certainly sum together the heat amounts that are introduced into the various layers
with an infinitesimal rise in all the temperatures and obtain
It thus seems to me that a correct minimum principle is obtained when one multi-
plies Boltzmann’s H for each layer with the relevant temperature and integrates
over all the layers[6]
The advantage would be that the density distribution would be obtained directly.
Yet, even if you should express approval of this consideration, I would not recom-
mend it for Miss Edith, because it would take forever and she must finally finish
up.[7]
This semester I am lecturing on “boundary value problems,” or, in other words,
partial differential equations, and am very pleased with my class because they are
working with real diligence and interest.[8] I am sorry that I shall probably not reap
the fruits of labors of this course, for if all indications do not deceive me, this will
be my last semester in Zurich.[9]
Meyer and Weyl were in Germany and returned the day before yesterday. You
probably spoke with Weyl in Berlin;[10] Meyer was in Tübingen. He was able to re-
lax well but is appalled by the reactionary spirit at the university.[11]
Lately I delivered some popular lectures on the theory of relativity to psychiatric
circles. It is strange why precisely these people are suddenly so extraordinarily rel-
ativistically concerned. Most interesting is Bleuler,[12] who is really trying to get to
the bottom of the issues.
With respectful greetings, I remain, yours very truly,
Paul Epstein.
xx
x
dQ dQn ; TdS TndSn = =
T f logf f – d d d dx minimum=