its convex surface. The cells adjacent to the inner concave surface of the vascular bundle have a deposit of colouring-matter in them ; but this disappears towards the base of the petiole. I am able to confirm the correctness of Dr. Ogilvie's description of this dark band in the petioles of Osmunda regalis ; but it is probably not always present, as Prof. Church was not able to detect it in the specimens he examined (Linn. Soc. Journal, Botany, vol. vii. 1864, p. 88). Bundles of coloured cells occur, also, irregularly scattered through the parenchyma of the wings.
The axis gives off in the older portions numerous adventitious roots. These proceed, as in tree ferns, from the interpetiolar spaces, and are not developed from the base of each petiole as figured and described by Duval Jouve in the living plant (Billot's Annotations a la Flore de France, 1855, p. 51, pl. 1. fig. 5 A). In the fossil I have a preparation showing their independent origin ; and in the living Royal Fern I have dissected them out, tracing them into the vascular cylinder. They have a different structure from that of the petioles, being more or less circular in transverse section, and composed of a small central vascular bundle imbedded in a very little pale parenchyma, and both surrounded by a cylinder of very dark brown elongated cellular tissue. On issuing from the axis the root pushes through the cellular tissues of the petioles, until it is able to pass upwards between two. It then repeatedly branches ; and when it escapes beyond the bases of the petioles the branches become long and wiry, and form a dense covering around the recent stems. This has disappeared from the fossil, because of the rubbing to which it has been subjected ; but the branching roots are obvious among the petioles in the least-worn surface.
The singularly perfect preservation of the tissues of this fossil is very remarkable. Not only are all the cells and vessels intact, but even the starch-granules, which abound in the parenchyma of ferns, still fill the cells in which they were originally formed ; or rather, I should say, the silica by which they are replaced, and which assumes their form, is there. In the form of the granules, and in the method in which they are packed in the cells, the fossil agrees exactly with the recent species. Many of the cells contain the mycelium threads of a parasitic fungus, which are inarticulate, and probably belong to the genus Peronospora, one species of which is too familiar from the injury it has brought on the potato crops. The dead stem of the fossil must have been at once attacked by this parasite ; it could never have been desiccated, as the most delicate tissues are perfectly preserved. Buried in the moist clay, the silica must have speedily replaced the organic tissues before the most delicate cells, the mycelium threads, or even the starch-granules were disorganized.
The position of the fossil cannot be doubted. It certainly belongs to Osmundaceoe, and most probably to the genus Osmunda. It would be referred without hesitation to this genus by some workers in fossil botany ; but it seems to me most desirable not to refer positively a fossil to a recent genus which has been established upon characters that are not present, and consequently cannot be determined, in the
