We show that the density of states for localized 4f electrons coupled to a
conduction band calculated in the framework of Bethe ansatz solution for th
e degenerate Anderson model qualitatively disagree with the well-known resu
lts obtained for the same model but using a variational approach. The scale
s of parameters used in our Bethe ansatz calculations to fit the experiment
s disagree with the commonly accepted values from other studies. This impli
es narrower conduction bands hybridized with 4f orbitals or questions the a
pplicability of the Bethe ansatz for a degenerate Anderson model for the hi
gh-energy characteristics of some rare-earth systems.