In the course of the experiments with the two-arm DISTO [1] spectrometer at the SATURNE accelerator in Saclay the exclusive production of w mesons was measured at excess energies Q = 89, 207, and 320 MeV (Q = Ös-Ö[(s₀)]). These beam energies are well suitable to close the considerable gap between the near-threshold data obtained in SPEC III experiments [2] at excess energies Q  <  30 MeV and data at larger Q  >  400 MeV (see e.g. [3]).
Data at the highest beam energy in DISTO experiments were analyzed and published so far [4-7]. The investigation of all data is still in progress. In this contribution first preliminary results on the analysis of the data at the lowest excitation energy for the w meson are presented.
The exclusive reactions pp®pph and pp®ppw were identified via the three pion decay (i.e. h, w® p⁺ p^- p⁰), where the braching ratios are taken from [8]. Only events with one negative and three positive charged particles were used in the data processing of h and w mesons. A good separation was achieved for p⁺ and p, also the p^- was clearly observed as shown in [4-6] using the light signal from the water filled Cerenkov detectors together with the particle momentum. Fig. 1 shows the (M^ppp+_inv)² - (M^p-_miss)² distribution, whereas the symbol (M^x_inv/miss) denotes the invariant or the missing mass of the particle or particle combination x. In Monte Carlo simulations only the region above 0.4 (GeV/c²)² was populated with h and w mesons. Therefore, this behavior was used to suppress pure 4 body events (ppp⁺ p^-) in the data analysis, which are peaked at (M^ppp+_inv)² - (M^p-_miss)² =  0 (GeV/c²)².
To improve the resolution of the (M^ppp+p-_miss)² spectrum a kinematical refit assuming a missing p⁰ meson was applied in a second fit procedure. The spectrum after this refit is shown in Fig. 2. The arrows depict the expected positions for the mesons taken from [8]. The main source of the remaining background is the non-resonant p⁺ p^- p⁰ production. Unfortunately, the exact form of the background is unknown. Therefore, the spectrum was fit with Gauss peaks for the h and w signals and the background was parametrized with a polynomial of third order and a wide Gaussian curve. In Fig. 2 are also depicted the fit result (full), the h and w peak (dashed), and the background (dotted), respectively.

Fig. 1 Distribution of (Mppp+inv)2 - (Mp-miss)2.	Fig. 2 (Mppmiss)2 spectrum after kinematic refit.

¹II. Physikalisches Institut, Justus-Liebig-Universität Gießen

References

[1] F. Balestra et al., Nucl. Instr. Meth. A 426, 385 (1999)

[2] F. Hibou et al., Phys. Rev. Lett. 83, 492 (1999)

[3] R. Baldi et al., Phys. Lett. B 68 (1977) 381

[4] F. Balestra et al., Phys. Rev. Lett. 81, 4572 (1998)

[5] F. Balestra et al., Phys. Lett. B 468, 7 (1999)

[6] F. Balestra et al., Phys. Lett. B 491, 29-35 (2000)

[7] F. Balestra et al., Phys. Rev. C 63, 024004 (2001)

[8] D.E. Groom et al., Eur. Phys. J. C 15, 1 (2000)  IKH 05/31/01 © R. Dressler