Four direct methods of obtaining values of stacking-fault energy from observation of faulted defects in pure face-centered cubic metals are discussed. It is shown that there is essential agreement between the method based on the observation of threefold nodes and that based on the observation of triangular Frank dislocation loops and stacking-fault tetrahedra in deformed f.c.c. metals, in the range where both methods are applicable. On the other hand, it is shown that the third method, based on the collapse size of stacking-fault tetrahedra in quenched metals, cannot yield even an upper limit. New experimental results show that the fourth method, based on the annealing rate of faulted loops, is applicable only to metals of high stacking-fault energy and then only if jog nucleation and propagation are not rate controlling; for low stacking-fault energy metals, these factors, together with the dislocation energy, must be considered, and cannot be completely taken into account.
