The E is not a glueball - but flavor symmetry shows how to find them

Glueballs are flavor singlets and can be identified by flavor symmetry predictions for selection rules and branching ratios. Flavor SU(3) predictions for glueball decays are expected to hold since SU(3) predictions for similar glue to light quark transitions in charmonium decays agree with experiment. The absence of a strong ηππ signal relative to KoverlineKπ in E decay rules out the δπ dominance of KoverlineK and the flavor singlet classification for E. Enhancement of a low-mass is shown to follow from G parity without the presence of the δ, while the known branching ratios of δ decay require a large ηπ signal. Data in the E region are consistent with a normal 1 ⁺ quarkonium nonet and additional broken nonets of quarkonium pseudoscalars with nearest-neighbor radial mixing.