IMPORTANCE OF THEηπDECAY MODE IN THE SPLITA2DECAY

If SU(3) symmetry is assumed and singlet-octet mixing is neglected, (a) the $\eta \pi$ decay mode is forbidden for all negative $\mathrm{CP}$ states, (b) $\eta \pi$ cannot appear in decays where the $K\overline{K}$ mode is absent, (c) the $\frac{\eta \pi }{K\overline{K}}$ branching ratio is the same for all isovector octet bosons allowed to decay into $\rho \pi$. Violation of these predictions would suggest serious re-examination of other SU(3) decay predictions. Prediction (b) seems to be violated by preliminary data. If there are no violations, and $\eta \pi$ and $\rho \pi$ are both allowed, the only possible classification for either $A_2$ peak with $J<\mathrm{}4\mathrm{}$ is $2^{++}$.