fragment can of course be made by any one who can intelligently compare a specimen with a drawing or a description ; but the interpretation of the affinities of a fossil or the restoration of a plant from the few fragments known, can be accomplished only by one who has some acquaintance with living organisms, and with the essential and non-essential characters which combine or separate them. When this knowledge is wanting, a lively imagination supplies its place to the complete satisfaction of the investigator, but to the great injury of science. From its very nature, moreover, an imaginative interpretation is more tenaciously adhered to by its author than if it were the legitimate deduction from known facts ; and it is more satisfactory, because it does not present the difficulties that are always encountered in real life. It is quite in keeping with this that a considerable contributor to the subject of fossil-botany has declared that a knowledge of recent plants is a serious hindrance to the investigation of fossil vegetables — and that another has recently expounded in a science-lecture his important determination of the affinities of Lepidodendron and Calamites to Lycopodium and Equisetum, although his descriptions make it evident that he never has examined, and probably never has seen a single specimen of either a club-moss or a mare's-tail in his life.
I propose in this paper to call attention to some of the specimens which I have either met with in collections, or have had sent to me as fossil plants, but which have no connexion with the vegetable kingdom. Instead of setting them aside as non-vegetable, I have taken advantage of the opportunities afforded by my connexion with the British Museum, and especially of the important assistance of my friend Mr. Thomas .Davies, to determine what they are.
It would be curious to trace such errors in scientific works, and show how frequently careful observers have thought that they had the most perfect foliage in dendritic crystals, and beautiful wood- structure in stalagmites. But I have encountered errors more remarkable than these, now happily exploded. Among others I may mention : — a curious form of calamite from a particular bed in the South Wales coal-field which turned out to be the fragment of the handle of a Wedgewood basalt tea-pot ; a branching part of the root of a great tree, the remainder of which was yet in situ and could be obtained, converted into metallic lead ; and a fragment of exogenous wood showing the openings of the medullary rays, which was a singularly altered piece of shale or slate from the wall of a vitrified fort.
The first specimen to which I would ask special attention is a supposed fruit figured and described by Sternberg in his ' Flora der Vorwelt,' tab. ix. fig. 2, as Carpolites umbonatus. These are round, flattened bodies, with a glazy polished surface and a central nucleus (Pl. XIX. figs. 12-17) ; they sometimes separate from the matrix enveloping them, and then appear to be fruits, with their pericarp converted into a thin shining layer of coal, like the Trigonocarpons that are found in similar beds. On careful examination, however, it is seen that the glazy surface is not produced by a foreign substance,
