The Dimensions of Crystalline Germs Comparable to those of Microbes.—We may dilute the solid salol with inert powder—lactin, for example—dilute the first mixture with a second, the second with a third, and so on; then, throwing into the solution of surfused salol a tenth of a milligram from one of these various mixtures, we find that the production of crystals will not take place if the fragment thrown in weighs less than a millionth of a milligram, or measures less than ten thousandths of a millimetre in length. It would seem, then, that these are the dimensions of the crystalline particle or crystallo-*graphic molecule of salol. In the same way Ostwald satisfied himself that the crystalline germ of hypo-*sulphite of soda weighs about a thousand-millionth of a milligram, and measures a thousandth of a millimetre ; that of chlorate of soda weighs a ten-millionth of a milligram. These dimensions are entirely comparable with those of microbes.
All these phenomena have been studied with a detail into which it is impossible to enter here, and which clearly shows more and more intimate analogies between the formation of crystals and the generation of micro-organisms.
Extension and Propagation of Crystallization. Optimum Temperature of Incubation.—Crystallization which has commenced around a germ is propagated more or less rapidly, and ends by invading the whole of the liquor.
The rapidity of this movement of extension depends upon the conditions of the medium, especially upon its temperature. This is shown very well by Tammann's experiments with betol. This body, the salicylic ester of naphthol, fuses at 96° C. If it
