Attention is called to a new phenomenon: in certain materials, specimen size (or, its energy-storage capacity) influences its brittle-ductile transition and strength. This effect is not the recognized statistical one (which only concerns the nucleation of fracture), blot derives from the strain-energy in the system and concerns the stability of sIow-growing cracks after nucleation. Recent experimental observations, in which this effect was noted, are cited and a theory proposed to account for them. This theory is based on the fact that a highly unstable equilibrium exists between the respective rates of strain-energy release and energy demand. When the system is over-stocked with strain-energy, any sudden drop in energy demand creates an excess of energy released. This then takes the form of kinetic energy capable of doing work against the remaining resistance, in turn, resulting in a lower fracture load and reduced ductility. Specimen size enters these considerations only in so far as it governs the amount of stored energy; consequently the effect is most pronounced in flexure.
Although the effect was so far observed in materials with moderate ductility, it is speculated that fine techniques might reveal its presence both "ideally" brittle and "ideally" ductile materials.