similar aptitude for repairing their mutilations. Pasteur, in an early work, discussed these curious facts. Other experimenters, Gernez a little later and Rauber more recently, took up the same subject, but could do no more than extend and confirm his observations. Crystals are formed from a primitive nucleus, as the animal is formed from an egg; their integral particles are disposed according to efficient geometrical laws, so as to produce the typical form by a constructive process that may be compared to the embryogenic process which builds up the body of an animal. Now this operation may be disturbed by accidents in the surrounding medium or by the deliberate intervention of the experimenter. The crystal is then mutilated. Pasteur saw that these mutilations repaired themselves. "When," said he, "a crystal from which a piece has been broken off is replaced in the mother liquor, we see that while it increases in every direction by a deposit of crystalline particles, activity occurs at the place where it was broken off or deformed; and in a few hours this suffices not only to build up the regular amount required for the increase of all parts of the crystal, but to re-establish regularity of form in the mutilated part." In other words, the work of formation of the crystal is carried on much more actively at the point of lesion than it would have been had there been no lesion. The same thing would have occurred with a living being.
Mechanism of Reparation.—Gernez some years later made known the mechanism of this reparation, or, at least, its immediate cause. He showed that on the injured surface the crystal becomes less soluble than on the other facets. This is not, however, an ex-
