DOC. 211 JULY 1910 157
211.
To
Arnold Sommerfeld
Zurich, July
1910
Dear
Colleague:
I
hesitated
so
long
to write to
you
because,
on
the
one
hand,
I would
so
much like
you
to
come
to
Zurich
during
the
summer
vacation,[1]
but,
on
the other
hand, I lack the
courage
to
urge
you
to
do
so.
For
I have not
been
able
to
come up
with
anything even
partly complete
regarding
the
constitution of radiant
energy.
Still,
there
is
one thing
I
believe I
see
clearly;
it
seems
incontrovertible that
energy
of
a
periodical
nature,
wherever
it
occurs,
always occurs
in
energy quanta
that
are
multiples
of hv. Smaller
quantities
of
energy
cannot be
isolated
in
principle,
neither
as
radiation
nor as
the
oscillation
of material
structures.
To
err
is human;
I
wouldn't
be
surprised
if
you
found
a
weak
point
in
my
reasoning.
But
this would
then
also be
the
saving
idea that leads
out
of
the dilemma.
I
have
pondered
innumerable other
possibilities,
but
always
came
back
to
the
conclusion
that
one
cannot do
without the
hypothesis
of the
finite
divisibility
of
the
energy
of
periodical
processes.
What
H. A.
Lorentz
and M.
Planck
said in the
Annalen
was
by
no means new
to
me.[2]
I
also do
not
believe
that
crudely
materialistic
conception
of the
point-structure
of
radiation,
which
gives
the
simplest
account
of
the
statistical
properties,
can
be
worked
out.
The
question
whether
Maxwell's
equations
can
be
upheld
for
vacuum seems
to
me
therefore
not
to
be fundamental,
because
these
equations
have
only
a physical
content
when taken
together
with
the
expression
for
energy
and
ponderomotive
force. Planck did
not
bring
forth
any
sound
argument
of
sorts
against
my
statistical
considerations
regarding
the
energy
distribution and the radiation
momentum
and discontinued written
discussion
about
this business
(he
did
not
reply
to
my
letter).[3]
It
now seems
pretty
certain
that,
as
regards
the
heat
content,
the molecules
of
solid
substances behave
in
essence
similarly
to
Planck's
resonators.[4]
Nernst found the
relationship
confirmed
in
the
case
of
silver and
some
other
substances,[5]
while I
recently
read that diamond
really
has
an
infrared
absorption
maximum
that
is predictable
from
its
thermal behavior
(the theory
yielded
from the
specific
heat
X
=
11ß,
and the observed
value
was X
=
12^).[6]
In
view of
the
extraordinary
difficulties involved
in
the
experiments, I
do
not
attach
great importance
to the fact
that the results
have not
yet
shown
that the
dependence
of
the maximum
photoelectric
effect
on v
is independent
of
the nature
of
the metal.
For
the
present,
the fact
that
the maximum
velocity
of the
emitted electrons
is independent
of
the
intensity
of the
exciting light
seems
to
me
more
important.[7]
The
crucial
point
in
the
whole
question
seems
to
me
to be:
"Can the
energy quanta
and
Huygens' principle
be
made
compatible
with
each other?" The
appearances
are
against
it
but,
as
it
seems,
the Lord
knew all the
same
how to
get
out
of the
tight
spot.[8]
I
am
greatly
enjoying your
new paper.[9]
How could
you
think that
I
don't
know how
to
appreciate
the
beauty
of
such
an investigation?
The consideration of the formal