cover the uncertainty in the calibration of the proton beam polarization. Their data shown for 119° (lab) were obtained by interpolation of measurements at 117.5° and 120°. In Figs. 2 and 3 the comparison is made for the same "compound-nuclear" energies by using two energy scales for E n and E p that are connected by the relation E n = E p -1.1 MeV. Clearly the measured values of A y (0) for these two charge-symmetric reactions are identical within the accuracy of the experimental values. This is the first time n + 4 He data have been compared to the data of Bacher et at, for p + 4 He and the first illustration of the equality of A y (0) for n + 4 He and p + 4 He when they are compared in this way.In summary, the reaction 3 H(d, n) 4 Ke can now be used to provide neutron beams from 20 to 30 MeV with a polarization known to about ± 2%. Such neutron beams were used to determine the 4 He(w, rc) 4 He analyzing power A y (0) at 119° (lab) which is near the back-angle maximum. The data show that the phase shifts of Hoop and Barschall predict A y ( 119°) fairly well above 22 MeV, but not in the immediate region below the 22-MeV resonance. The angular distribution data at 20.9 MeV favor the phase-shift sets of Lisowski and Walter 9 and of Stammbach and Walter 10 over earlier sets. Comparison to 4 Ue(p,p) 4 Re experimental data at the same "compound-nuclear" energies shows that the results for the two charge-sym-Experimental systems capable of detecting nuclear reaction products in resonant final states with good efficiency and energy resolution can open up a wide range of unexplored nuclear reactions. Although at present such studies are largely confined to the detection of 8 Be nuclei/* 2 Robson 3 has pointed out many other interesting resonant systems which can be detected as reaction products. Additionally, the well known finalstate interaction in the two-nucleon 1 S 0 , T = 1 system can be utilized; in particular, this interac-metric reactions are identical within the accuracy of the measurements.