DOC.
6
FEBRUARY
1903 9
LiCl
P
i1
i2
i3 i4
1.17 1.95
-
- 1.75
2.-
2.-
1.85
1.72
1.70
4.38
-
1.93
1.78
1.61
NaI
P
i2 i4i3
6.38
1.76 1.86
1.74
16.18 1.99
2.01 1.70
KI
p
i2 i3 i4
7.85
1.85
1.89 1.79
15.86 1.90 1.87 1.78
This
was
the
basis
of
my
statement last
summer
that
the
experimental
confirmation
of
the dissociation
hypothesis
leaves much
to
be
desired. But
we
are
dealing
here
with
what
are
already
old observations
and
probably
also with
printing
errors
(transposition
of lines in the
i1
i2
i3
columns).
Could
you verify
that
in
Bern?
In
Roloff's short but
elegant
article
on
electrolytic
dissociation,[13]
there
are
tables of results that
are
at
least
somewhat
in
agreement
with
the
theory-I
have not
yet
seen a
good
critique
of the
experimental
results.
2.
Your
experimental
criterion
for
Sutherland's
hypothesis[14]
had
not
occurred
to
me.
It
might
be difficult
to
apply
it
because
of
the
gaps
in
our
knowledge
about
semipermeable
membranes[15]-still,
according to
your
criterion,
it suffices
for the
rejection
of the
hypothesis
if
one can
find
two
membranes
one
of
which
is
more
"perfect"
for the
dissolved substance
a
than
for
b,
while the
opposite
is true
for the
other
one-which
does
not
seems
to
me
implausible,
whereby
the
hypothesis
would
collapse.
I
am
enclosing
two
slips
of
paper,
from which
you
can see
how I
figured
out
two
possibilities
for
pressure
reduction
in
cases
of
incompl. semipermeability.[16]
Please
return the two
slips
to
me, preferably together
with
the
sketch I
sent
you
earlier
(if you
can
still find
it),
because
I
did
not
make
any copies
as
I
was
in
hurry.
It remains
to
be
investigated,
and this
seems
to
be
a simple thing,
how
one
should
conceive
the
change
in
the free
energy
in the
course
of solution and dilution
according
to
my
"coarse-grained"
model.