1 7 6 D O C U M E N T 3 8 M A Y 1 9 2 0
38. From Paul Epstein
Zurich, 6 Physik St., 30 May 1920
Highly esteemed Professor,
Permit me to direct a few questions to you regarding the business of Miss Edith’s
dissertation.[1]
Miss Edith did not exactly expedite the matter; it is also correct that
due to Meissner’s illness she was completely prevented from working and studying
for months on end. No doubt she has a share in the success of nursing Meissner
back to
health,[2]
but obviously this cannot be combined with progress in the
project; and that is how it came about that only now did she complete the calcula-
tions that you fully described. It is not much, of course, but at least something.
With reference to some notes she made following your
instructions,[3]
I would,
nevertheless, still like to request some explanation. It involves what can be regarded
as constant in each cross section. Let us look at the integral
; the same thus gives the momentum that is carried
through the unit of area perpendicular to the x axis.[4] When we take a parallel
plane, no momentum can form between the two if the collisions satisfy the laws of
mechanics and, in the stationary state, the same momentum must be transported
through the unit area of the second plane. If I understand Miss Edith’s notes
correctly, however, she regarded a different quantity as constant, namely (if we de-
note it as ), . This, as far as I can see, would involve
that the quantity enclosed in parentheses be constant, which, however, scarcely al-
lows agreement with the condition of constancy of . In hydrodynamics and
elasticity theory, constancy of this quantity is required, but there the terms stem-
ming from the temperature gradients are simply not taken into account.
I am completely aware that my mind has been a little dull recently: it seems that
the nervous tension from which I can’t escape here in Zurich has gone to my head
and has impeded my ability to concentrate. It is therefore easily possible that I have
been writing nonsense and, in the latter case, I ask you please not to read my further
questions, since they are all based on the former train of thought. For if it is correct
that = const., then the reader could ask why we introduced just three secondary
conditions: number constant, energy = (or, to be precise, ), and ener-
gy flow = and not also the fourth = const., which would alter the form of
the function f( ) even further.[5] The reason obviously lies in that the state of
m f 2 ddd
-
+
xx xx
1
3
-- -
xx yy zz
+ + –
xx
xx
kT
2
----- -
dU
T
------- ds =
x
xx