The S·spectrum of RaE is reinvestigated taking into account the finite de Broglie wave length effect pointed out by Rose et al. 'and the correction to' the nuclear matrix elements pointed out by the author, and it is shown that the conclusion of Petschek and Marshak that only the assumption of tensor + pseudoscalar, spin chan/Si 0-0, parity change yes is able to explain the experimental ,results is incorrect. Taking into account the above two effects, many other assumptions can explain the experiments. The results are as follows: TPo-o (tensor+pseudoscdar and spin change o-o), STI-O and VAt-O fit the experiments only with the finite de Broglie wave length effect, Ao-o (including APO-O) and VTI-O requires both elects, but VT1-'0 seems too artificial. Other cases can not fit the experiments_ § 1. Introduction and summary It has been well known long since that the p-spectrum of RaE deviates from the allowed shape, but there had been no definite theory to explain it. By the Mayer's nuclear shell model it is made clear that this transition is parity change yes, and last year Petschek and Marshak!) concluded that only the mixture of tensor and pseudoscalar interaction and spin change 0-0 (of course parity change yes) is able to explain the experiments. Some time after, however, Ahrens, Feenberg and Primakoff2) insisted using their method:!) to evaluate the ratios between nuclear matrix elements that the nuclear matrix element \l1r;; in pseudoscalar type is very small and the large ratio of the coupling constantslGI'/GII ~133 is required to fit the Petschek's and Marshak's interpretation.However, this large ratio causes many awkward affairs, * and recently Ruderman 6 ) has shown that if the nuclear force includes a large pro type one the relation of Ahrens, Feenberg and PrimakofPl becomes invalid and the magnitude of \J9rs is of the order of other momentum type nuclear matrix elements (\ a, I pa and frs). According to his theory it is sufficient to take the coupling constant Gp to the same order as G 7 , to fit the result of Petschek and Marshak!)On the other hand, if the 'nuclear force is entirely due to the pseudovector coupling of 1r-mesons, the evaluation of Ahrens, Feenberg and Primakoff2) is valid, and a large pseudoscalar coupling constant is required. Therefore, if the foundation of the article of Petschek and Marshak were reliable and if it becomes clear that a * First, as pointed out by Ahrens, Feenberg and Primako£f2) a new nuclear matrix element should be added to the old one, ISr". The effect of this new nuclear matrix element to the S,speetrum of RaE will be examined by Takebe. 4 ) Moreover, this large coupling constant destroys the allowed shape of high energy 6-spectra.5) Also Cf. t'eferenee 6.