For HR 4487, we used the homogeneous 
colors of Hauck &
Mermilliod (1980) with the calibration of
Napiwotzki et al. (1993) and found
K, log g = 3.92. As we do not have spectrophotometry nor a
Balmer line profile for this star, we adopted these values for our analysis.
This method also gives 
K, log g = 3.72 for 14 Hya and
K, log g = 4.40 for 3 Cen A.
For 14 Hya, comparison of
spectrophotometry (Adelman & Pyper 1979)
with the predicted fluxes of an
ATLAS9 solar composition model with the above combination of effective
temperature and surface gravity is not optimal. Better agreement is obtained
with the predictions of a 
K, log g = 3.75
model with no
microturbulence which we adopted. For 3 Cen A, Castelli et al. (1997)
derived
, log g = 3.8 after considering the problems
with a variety
of photometric data produced by contamination by the secondary which is a B
star 1.5 magnitudes fainter, but only 
away.