On the basis of the usual assumption of a secondary minimum in a one-dimensional potential barrier, we derive a relationship between the half-life of a spontaneously fissioning isomer, its excitation energy, the fission-barrier height and the half-life for ground-state spontaneous fission. For 241Am, there is no inconsistency between the experimental values for these four quantities, but the computed half-life for the ground-state spontaneous fission of 242Am is considerably larger than the measured value. On the basis of this discrepancy, we discuss the possibility that the potential-energy surface may be more complicated than previously supposed and contain both a high thin barrier through which ground-state spontaneous fission proceeds and a thicker two-peaked barrier in which the isomer is located and through which induced fission proceeds. Such a two-dimensional potential-energy surface is consistent with both the isomer data and the recently observed intermediate structure in fission cross sections. To help decide whether such a potential-energy surface is indeed responsible for the observed discrepancy for 242Am, we suggest that for large deformations near tge symmetrical saddle-point shapes the nuclear potential energy be calculated as a function of reflection-asymmetrical P3-like" deformations. © 1969."