10 The importance of imaginary contributions to the phase shifts by dispersion corrections has been recog-The simplest expectation for mirror /3 decays is that they should have identical/£ values. It has, however, been known for some time that the ft value for N 12 decay to the ground state of C 12 is some 10% greater than that for the mirror decay of B 12 . Careful analysis 1 suggests that the discrepancy may well be a significant one, not explicable in terms of electromagnetic, secondforbidden, isospin-mixing, and binding-energy corrections, and that its resolution may lie in the reality of second-class currents, specifically the induced tensor interaction, 2 ' 3 i.e., that the 0 interaction does not respect G parity. 2 Before accepting this fundamental conclusion one must be sure both that the known corrections have been properly evaluated and that there are not others of a "structural" nature that may fluctuate from case to case.It is therefore important to investigate other cases of mirror decay to see whether or not they fall systematically into line with A = 12. If the discrepancy in ft values was due solely to an induced tensor term its magnitude, for light nuclei and W Q * » 1, should be proportional to W 0 + + W 0~ and approximately state independent. 3 * 4 We have carried out measurements on the systems A = 20 and 25 that enable us to extend the mirror test to nized previously by Rawitscher (Ref. 6). Using a model of pure monopole excitation he was able to show that the imaginary contributions to the phase shifts are bigger than the corrections of the real part by about one order of magnitude. Assuming then that the real part of the correction may be neglected, our Eq. (1) may be derived directly from the Appendix of Ref. 6. n F. Wo J. Olver, Phil. Trans 0 Roy. Soc D London A247, 307 (1954), and to be published. these cases. We find that A = 25 shows a large departure from mirror symmetry in the same sense as for A = 12; the evidence on A = 20 is conflicting but suggests that the departure from mirror symmetry could have the opposite sense. These results emphasize the caution that must be used in interpreting the failure of mirror symmetry in terms of second-class currents.The P~ decay of F 20 is 99.92% to the first excited state of Ne 20 at 1.63 MeV. 5 The /3 + decay of Na 20 has been extensively investigated 6 by measurement of its j3 rays and the subsequent y decay and a decay of Ne 20 . The Na 20 half-life is reported 6 as 408 ± 6 msec and the /3 branch to the first excited state of Ne 20 as 90.0%. However, in view of technical problems in the j3-ray measurements we have preferred to make our analysis in terms of the superallowed transition of Na 20 to its analog in Ne 20 at 10.270±0.009 MeV. 7 For the Fermi part of this transition we have taken 8 ft-3060 ± 20 sec; for the Gamow-Teller part we have based ourselves on local systematics and taken the range ft = 6.3 xl0 4 -°° sec; this leads toft = 2990 ± 70 sec for the two together.The half-life of F 20 has been determined as 11.03 ±0.06 sec and J3...