Why the high-lying glueball does not mix with the neighbouring f 0

Chiral symmetry restoration in high-lying hadron spectra implies that hadrons which belong to different irreducible representations of the parity-chiral group cannot mix. This explains why the $f_0(2102 \pm 13)$, which was suggested to be a glueball, and hence must belong to the scalar (0,0) representation of the chiral group, cannot mix with the neighbouring $f_0(2040 \pm 38)$, which was interpreted as a $ n\bar n$ state, and that belongs to the $(1/2,1/2)$ representation of the chiral group. If confirmed, then we have an access to a "true" glueball of QCD.
4 pages, LaTeX, final version, Eur. Phys. J. A 19 (2004) 153