242
DOC.
331
DECEMBER
1911
frequencies.
This would be
incomprehensible,
because
only one
relative
oscillation of
the
Na+ and Cl- ions
is
conceivable.
But
the
dispersion theory
yields
the
following
in
the
infrared
in
the
case
of
one
proper
frequency:
Xo
n
v
=
0,68A0
A0
the
wavelength
of
the
proper
frequency. n2
the
square
of
the
refraction
index.
For
negative n2
(region
between
0.68
A0
and
X0
for rock
salt)
the
substance
is
completely
reflective.
Reflectivity
approximately
so:
0,68p0
A0
Absorption
almost
without
influence
on
the
reflectivity.
Residual
rays
correspond to
wide
region.
But
sharply
delimited
wide
region
yields
"beats"
similar
to
those found
by
Rubens.
Thus,
in
reality
only one
proper
frequency,
its
wavelength
being longer
than both of Rubens's.
Nernst based
his
grand double-quantum theory
on
Rubens's
values.[6]
The
thing
is
very
strange.
Xo
true
curve
H.
Rubens
Habicht
demonstrated
his Maschinchen to the Berlin
Phys.
Soc. and
had
enormous
success.[7]
The
Maschinchen
is
head
and
shoulders
above
the filament
electrometer,
and
its
future
is
now
secured.
This
makes
me
tremendously
happy.
Habicht
has
already
received
quite
a
few orders.-
There
are
lots
and
lots
of
interesting
things
to
talk
shop
about,
but
paper
is too
cumbersome for that. The
most important
result that
the
physicists
have
now
firmly
established
seems
to
me
to
be
the
proof
that
spectral
lines
can
be
emitted
and
absorbed
by
neutral metal
vapors, e.g.,
sodium
lines. You must
come again
soon;
what
a pity
that
we are
not together,
there
are so
many
interesting
things
we
could
do.-