DOCS.
394,
395 MAY
1912
297
**
=
Đ
dŪ
=
Đ
30
=
Đ
'
-
'
dy
dt
a
=
ŪyA
+
ŪB
+
a
y i
ß
=
B
@rA
+
ŪyB
+
Đ,i
a y
From the
above
condition
I
then obtain the
following
equation
[8]
395.
To
Wilhelm
Wien
Prague,
17
May
1912
Highly
esteemed
Colleague:
Please
forgive
me.
I did
indeed
forget
to enclose
the third sheet of
my
note.[1]
After
many
fruitless
attempts,
I
too
came
to
the
conclusion
that
one
will not be able to
help
the
theory
of radiation
to its
feet
merely by
constructing
models. This
is
why
I
tried
to
arrive
at
a new
formulation of
the
question
in
a
purely thermodynamic
way,
without
making
use
of
a
model.
Planck's
new
theories contain
so
many hypotheses
that
they
seem
almost worthless
to
me;[2]
for
one
cannot assert
that the
validity
of
the
radiation
equation
can
support
these
hypotheses
to
any
appreciable degree,
and I also do not
see
how
one
could otherwise
apply
these
hypotheses
in
a
useful
way.
The
important question
seems
to
me
at
the
moment to be
whether
a
non-monochromatically
sensitive structure satisfies
the
hv
law for the
v
of
the effective
radiation
or
for its
mean
proper
frequency.[3]
An
important question
connected
with this
is
the
following.
One
can
very
well
assume
that
in
a
selective
photochemical
effect,[4]
which extends
over
a
broad
frequency range,
the
same
elementary
structures
of the metal
should be viewed
as
the carriers of the
effect
for
the entire
range.
If
I
successively
irradiate
with
the
two frequencies
v1
and
v2
of
the
range,
then the
first time
I
will
get
for the
kinetic
energy
of
the
electrons
(obtained
from
very
thin
foils)
hv1
-
p,
the second
time
hv2-
p.
If
I
irradiate
with
both
frequencies
simultaneously,
I will
get
either
(a) two
electron
energies,
namely
hv1
-
p
for
one
part
hv2-
p
for
the
remaining part
or
(b)
kinetic
energies
arise
that
lie
between these
two
values,
even if,
in
case
of
monochromatic
irradiation, the slow
cathode
rays
were
homogeneous.
It
now seems
to
me
that
in
every
accumulation
theory
both
frequencies
must
take
part
in
the
ejection
of
every
single
electron
(e.g.,
in
Sommerfeld's
theory).[5]
Wouldn't
one