WHEN the early craftsmen first observed that the metals they worked in were made harder by hammering, and that the original softness could be again restored to the hardened metal by heat, it probably did not to them that any explanation of these useful properties was called for. At a later period, when an interest in the reasons for things became more general, it is probable that hardening was attributed to the compacting of the substance by the driving of its particles closer together so that the mass as a whole became less open or porous. In the same way heat annealing was probably assumed to act by permanently expanding the metal and opening up its texture. So many analogies to these operations were ready to hand from the most common and everyday experiences that it is not surprising that even on closer inquiry this explanation should continue for a time to be accepted as sufficient, the more so as it was obviously true that in some cases unworked metal had an openness or porosity which could be removed by hammering or working. While the researches of chemists on the density of the metals showed plainly that increase of density does not always result from compacting by pressure, these researches were probably too far removed from the ken of those who were most intimately concerned with the working of metals to arouse them to the insufficiency of the existing explanation of hardening.