The reaction π − p→ π + π − n has been measured in a high-statistics experiment on a transversely polarized proton target at 17.2 GeV, and unexpectedly large nucleon polarization effects have been observed. Combining the results of this experiment with a measurement on a hydrogen target allows a model-independent partial-wave analysis in terms of the “nucleon transversity” amplitudes. Unique or at most twofold ambiguous solutions are obtained. In particular we find a high lower limit ( ⪆30% ) of the spin non-flip unnatural exchange amplitudes at low | t |. These amplitudes, interpreted as being due to the exchange of an object with the quantum numbers of the A 1 , have been assumed to be absent in previous analyses. In checking the consequences of this finding on the old results, we test the validity of the rank-two assumotions for the density matrix. We find a small but significant deviation, which shows the need for a new phase-shift analysis including the A 1 exchange contribution.
MASS DEPENDENCE OF NORMALIZED T-CHANNEL MOMENTS SCALED TO 100 PCT POLARIZED PROTONS.
T DEPENDENCE OF NORMALIZED T-CHANNEL MOMENTS IN THE RHO REGION SCALED TO 100 PCT POLARIZED PROTONS.