Conclusions

The major background for the $\omega \rightarrow \pi^0\gamma$ signal comes from charged $\rho$ decays. It can be reduced by requiring large energy deposition in the LGD, $E_{\mbox{thresh}}>4.0$ GeV. In addition, using the CPV this background can be almost completely eliminated. In the case of $\eta \gamma$, the main background to the $\phi$ signal comes from neutral $b_1$ decays. These cannot be eliminated by a large energy threshold in the LGD, nor by a charged-particle veto. It has been shown that a large fraction of these events that are reconstructed as $3\gamma$ have no photons in the BGV. Therefore the first line of attack on $b_1$ background has to be to optimize the clusterization in the LGD. Once most of the $b_1$ events with $N_{\mbox{LAP}} > 3$ have been eliminated from the $3\gamma$ sample then the BGV can be used to veto a large fraction of what remains. We should lower the energy threshold for clusters as much as possible. When the LGD cluster threshold is reduced then we can expect to start to reconstruct many more reactions containing charged particles. To control this background, we will need to combine improved clusterization with the use of the CPV. Concerning BGV, cutting on energy in the barrel might also hurt the $\phi$ signal. In this study, half of the $\phi$ events contain a $\pi^0$ that comes from photo-production where a $\Delta^+$ is produced. According to Ref.[2] this cross section is comparable to diffraction, although the experimental errors are large. From what we know about diffraction, photo-production of a $\phi$ via pion exchange should be highly suppressed. Therefore we hold this conclusion in doubt. In the case of $5\gamma$ ( $\pi ^0\pi ^0\gamma$) events we see a large contribution from the $a_1$, $a_2$, and $b_1$ in the region of the $\phi$ mass. It cannot be reduced by raising the LGD energy threshold. However, the CPV does suppress the $a_1$ and $a_2$ contribution. This leaves a large $b_1$ signal whose tail extends under the $\phi$. We will need to eliminate the $\omega\pi^0$ region of phase space if we are to find the $\phi$ in $5\gamma$.