In Fact 7.9 and the reference to Kaplanski-Levitzky in its proof, the matrix 
ring should be over a field, not an arbitrary division ring. The fact is 
applied only to the field of complex numbers. (Considering M_1(\Delta) shows 
that the statement of Fact 7.9 is obviously too strong.)

Thanks to Geva Yashfe for pointing this out.