plants; and although there are still indications of vast gaps in our knowledge, due, no doubt, to the very exceptional conditions required for the preservation of plant remains, we now possess evidence of a more continuous development of the various types of vegetation. According to Mr. Lester F. Ward, between 8000 and 9000 species of fossil plants have been described or indicated; and, owing to the careful study of the nervation of leaves, a large number of these are referable to their proper orders or genera, and therefore give us some notion—which, though very imperfect, is probably accurate in its main outlines—of the progressive development of vegetation on the earth. The following is a summary of the facts as given by Mr. Ward:—
The lowest forms of vegetable life—the cellular plants—have been found in Lower Silurian deposits in the form of three species of marine algae; and in the whole Silurian formation fifty species have been recognised. We cannot for a moment suppose, however, that this indicates the first appearance of vegetable life upon the earth, for in these same Lower Silurian beds the more highly organised vascular cryptogams appear in the form of rhizocarps—plants allied to Marsilea and Azolla,—and a very little higher, ferns, lycopods, and even conifers appear. We have indications, however, of a still more ancient vegetation, in the carbonaceous shales and thick beds of graphite far down in the Middle Laurentian, since there is no other known agency than the vegetable cell by means of which carbon can be extracted from the atmo-
- Sketch of Palaeobotany in Fifth Annual Report of U.S. Geological Survey, 1883-84, pp. 363-452, with diagrams. Sir J. William Dawson, speaking of the value of leaves for the determination of fossil plants, says: "In my own experience I have often found determinations of the leaves of trees confirmed by the discovery of their fruits or of the structure of their stems. Thus, in the rich cretaceous plant-beds of the Dunvegan series, we have beech-nuts associated in the same bed with leaves referred to Fagus. In the Laramie beds I determined many years ago nuts of the Trapa or water-chestnut, and subsequently Lesquereux found in beds in the United States leaves which he referred to the same genus. Later, I found in collections made on the Red Deer River of Canada my fruits and Lesquereux's leaves on the same slab. The presence of trees of the genera Carya and Juglans in the same formation was inferred from their leaves, and specimens have since been obtained of silicified wood with the microscopic structure of the modern butternut. Still we are willing to admit that determinations from leaves alone are liable to doubt."—The Geological History of Plants, p. 196.