, following the
method of Boness et al. [1986]. For the outer core, Stacey
[1995] corrected for reduced mean atomic mass from impurities (he assumed
), 
or 778 Jkg
K
. The plot
of isobaric 
versus 
for liquid iron shock experiments given
by W.W. Anderson and Ahrens [1994] is between 
and 
at ICB conditions, in agreement with Stacey. Thus we
find 
by both theory and experiment.
Stacey showed that 
of the outer core is independent of depth because the temperature
gradient is adiabatic. To make the correction to
, we use the standard formula 
, where 
will be written in
terms of 
. Using the relationship 
, where 
is the bulk sound velocity,
Braginsky and Roberts [1995] expressed this formula
in the form 
where 
is the
bulk sound velocity. 
is calculated
for pure iron 
, where, according to
Stacey [1995], 
. By
substituting the PREM values for 
for the core at the CMB
and at the IMB, taking 
for the diluted iron core, and using 
and 
, Braginsky and Roberts [1995]
found 
and 
.
Stacey [1994] found 
(midcore)
to be 
for pure iron.
Jeanloz and Wenk [1988] proposed
for the inner core.