We shall be concerned here
with the nature of subrings of a ring with involution which are invariant
with respect to certain combinations of elements. To be more precise, let R
be a ring with involution ∗and suppose that A is a subring of R such that
xAx∗⊂ A for all x ∈ R. Can we say something definitive about the structure of
A? We shall see that if R is semi-prime then we do get a dichotomy of the
Brauer-Cartan-Hua type, namely, A must contain a nonzero ideal of R or A must be
central.