178
DISSERTATION
ON MOLECULAR DIMENSIONS
matter
or
electricity-can be traced back
to
his earlier
work.[64]
Einstein's
previous
work
had
touched
upon
most
aspects
of
the
physics
of
liquids
in
which
their molecular
structure
is
assumed to
play a
role,
such
as
Laplace's
theory
of
capillarity,
Van
der Waals's
theory
of
liquids,
and
Nernst's
theory
of
diffusion and
electrolytic
conduction.[65]
Before
Einstein's
dissertation,
the
application
of
hydrodynamics
to
phenomena
involv-
ing
the atomic constitution
of
matter
or electricity was
restricted to
consideration of
the
effects
of
hydrodynamical
friction
on
the motion
of
ions.
Stokes's
law
was
employed
in
methods
for
the determination
of
the
elementary
charge[66]
and
played a
role in
studies of
electrolytic
conduction.[67]
Einstein's
interest
in
the
theory
of
electrolytic
conduction
may
have been decisive for the
development
of
some
of
the main ideas in his dissertation.[68]
This interest
may
have
suggested a study
of
molecular
aggregates
in
combination
with
water, as
well
as some
of
the
techniques
used in the dissertation.
In
1903
Einstein and Besso discussed
a theory
of
dissociation that
required
the
assump-
tion of such
aggregates,
the
"hypothesis
of
ionic
hydrates" ("Ionenhydrathypothese"),
as
Besso called
it,[69]
claiming
that this
assumption
resolves difficulties with
Ostwald's
law
of
dilution.[70] The
assumption
also
opens
the
way
to
a simple
calculation of
the sizes
of
ions in
solution,
based
on
hydrodynamical
considerations. In 1902
Sutherland
had
consid-
ered
a
calculation
of
the sizes
of
ions
on
the basis
of Stokes's
formula,
but
rejected
it
as
in
disagreement
with
experimental
data.[71]
Sutherland did
not
use
the
assumption
of
ionic
hydrates,
which
can
avoid such
disagreement by permitting
ionic sizes to
vary
with such
physical
conditions
as temperature
and concentration.[72] The idea
of
determining
the sizes
of
ions
by means
of
classical
hydrodynamics
occurred to Einstein in
1903,
when he
pro-
posed
to Besso what
appears
to be
just
the calculation that Sutherland had
rejected:
Have
you already
calculated the absolute
magnitude
of
ions
on
the
assumption
that
they are spheres
and
so large
that the
hydrodynamical
equations
for
viscous
fluids
are applicable?
With
our
knowledge
of
the absolute magnitude
of
the elec–
[64]
Einstein
presumably
had
acquired a
basic
knowledge
of
diffusion in
liquids
from his
study
of
Violle 1893
(see
Vol.
1,
"Albert Einstein-
Beitrag
für
sein Lebensbild,"
p.
lxiv).
Chapter
4 contains
an
extensive treatment
of
diffusion
and osmosis.
[65]
See Vol.
1,
the editorial
note,
"Einstein
on
Molecular Forces,"
pp.
264-266,
and,
in this
volume,
the
editorial
note,
"Einstein
on
the Na-
ture
of
Molecular Forces,"
pp.
3-8.
[66]
See Townsend
1920,
pp.
209-214, for
a
review
of
the
use
of Stokes's
law in the
interpre-
tation
of
experiments on
the determination
of
atomic
charges.
[67]
For
a contemporary
discussion
of
the
ap-
plication
of Stokes's
law to
electrolytic
conduc-
tion,
see Bousfield
1905b. For
a
later
survey, see
Herzfeld 1921,
pp.
1011-1018.
[68]
For evidence
of
this interest,
see
Einstein
1902a
(Doc. 2)
and Michele Besso
to Einstein,
7-11
February
1903.
[69]
See ibid.
[70]
For
a
discussion
of
this
law,
its failure for
strong
electrolytes,
and the resolution
of
this dif-
ficulty
by
the "hypothesis that
the
ions
of
an
electrolyte
consist
of molecular
aggregates
in
combination
with water,"
see Bousfield
1905a,
p.
563.
[71]
In
Sutherland
1902,
Sutherland wrote:
"Now
this
simple theory
must have been written
down
by many a physicist
and found to be want-
ing"
(p. 167).
[72]
This conclusion
was
drawn
by
Bousfield
(see Bousfield
1905b,
p. 264).