The structure of osmotic growths is no less varied than their form. Their stems are formed of cells or vesicles juxtaposed, showing cavities separated by osmotic walls. Sometimes the component vesicles have kept their original form, so that the stem has the appearance of a row of beads. Or the cells may be more or less flattened, the divisions being widely separated. Or again, by the absorption of the divisions, a tube may be formed, a veritable vessel or canal in which liquids can circulate. {143} [Illustration: FIG. 56.--Microphotograph showing the structure of osmotic stems. (Magnified 40 diameters.)] [Illustration: FIG. 57.--Photograph of an osmotic leaf showing the veins.] The foliaceous expansions, or osmotic leaves, also present great varieties both of appearance and of structure. The veins may be longitudinal, fan-shaped, or penniform. We have occasionally met with leaves having a lined or ruled surface, giving most beautiful diffraction colours. The usual structure, however, is vesicular or cellular, as in Fig. 58. In photographs we often get the appearance of lacunae, but all these lacunae are closed cavities, the appearance being due to the transparency of the cell walls. [Illustration: FIG. 58.--Photomicrograph of an osmotic leaf showing the cellular structure.] In conclusion we may say that osmotic growths are formed of an ensemble of closed cavities of various forms, containing liquids and separated by osmotic membranes, constituting veritable tissues. This structure offers the closest {144} resemblance to that of living organisms. Is it possible to doubt that the simple conditions which produce an osmotic growth have frequently been realized during the past ages of the earth? What part has osmotic growth played in the evolution of living forms, and what traces of its action may we hope to find to-day? Osmotic growth gives us fibrous silicates, phosphatic nodules, corals, and madrepores; it also gives us formations which remind one of the "atolls," calcareous growths rising like a crown out of the water. The geologist may well consider what role osmotic growth may have played in the formation of the various rocks, siliceous, calcareous, barytic, magnesian, the fibrous and nodular rocks and atolls. The palaeontologist relies on the different forms found in his rocks to classify his specimens; from the existence of a shell, he concludes the presence of life. Since, however, forms which are apparently organic may be merely the product of osmotic growth, it is evident that he must reconsider his conclusions. The same may be said of the various forms of coral or of fungoid growths. In the {146} presence of a calcified or silicated fungus we can no longer argue with certainty as to the existence of life, without taking into consideration the possibility that the specimen in question may be an osmotic production. [Illustration: FIG. 59.--Osmotic growth with nucleated terminal organs. (One-third of the natural size.)]