We obtained total radiation widths of s-wave resonances through R-matrix analysis of 147 Sm(n, γ) cross-sections. Distributions of these widths differ markedly for resonances below and above En = 300 eV, in stark contrast to long-established theory. We show that this change, as well as a similar change in the neutron-width distribution reported previously, are reflected in abrupt increases in both the average 147 Sm(n, γ) cross section and fluctuations about the average near 300 eV. Such effects could have important consequences for applications such as nuclear astrophysics and nuclear criticality safety.PACS numbers: 24.30. Gd, 24.60.Dr, 24.60.Lz, 25.40.Lw In this letter, we show that total-radiation-widths (Γ γ ) extracted from R-matrix analysis of 147 Sm(n, γ) cross sections reveal an abrupt change in the shape and average value of the Γ γ distribution near E n = 300 eV. These observations are in stark contrast with theoretical expectations that both quantities should remain essentially constant across the resonance range and beyond.The effect reported herein occurs very near the same energy as previously reported abrupt changes in the α-particle strength-function ratio [1] for the two s-wave spin states and the shape of the reduced-neutron-width (Γ 0 n ) distribution [2]. Due to the difficulty of measuring the very small α widths, the former effect was of limited statistical significance. However, the effect in the Γ 0 n data was established at about the 99% confidence level using several different tests. These two previous effects remain unexplained.As we will show below, changes in the Γ γ distribution shape and average are established with very high confidence. That three such deviations from theoretical expectations could occur by chance in the same nuclide at the same energy must be vanishingly small. Therefore, it is virtually certain that significant departure from standard theory has been observed and that all three effects may have a common origin. Given the relative paucity of high-quality Γ γ data, similar effects may exist for other nuclides and, if so, could have far-reaching consequences for both basic and applied nuclear physics. For example, as we show below, these changes in the Γ γ and Γ 0 n distributions are reflected in abrupt increases in both the average 147 Sm(n, γ) cross section and fluctuations about the average that cannot be explained by the nuclear statistical model. As this theory is used to calculate cross sections for applications, similar differences in other nuclides could have important impacts in nuclear astrophysics and nuclear criticality safety.It is expected that Γ γ distributions for medium to heavy nuclides should be very narrow, and essentially constant across the resonance energy region. Expectation [3] that the Γ γ distribution should be very narrow arises from i) the very complex wave functions of states at high excitation characteristic of neutron thresholds and ii) the very large number of channels for γ decay. Condition i) results in partial γ-decay widths...