The Ferns as a whole represent a section of the vegetable kingdom which traces its ancestry as far into the past as any group of plants. Impressions of leaves on the shales of the Coal-measures and on rocks of the earlier Devonian period are hardly distinguishable in form and in the venation and shape of the leaflets from the finely divided fronds of modern ferns. Until a few years ago these Palaeozoic fossils were generally regarded as true ferns, and it was believed that ferns played a conspicuous part in the vegetation of the earliest periods, of which we have any botanical knowledge. Conclusions based on external form must frequently be revised in the light of more trustworthy evidence. It was shown in the later part of the nineteenth century by the late Professor Williamson of Manchester, whose researches into the plants of the Coal age shed a flood of light on the ancestry and inter-relationship of many existing plants, that some of the fern-like leaves which have long been familiar to those who search among the shales of the refuse heaps of collieries, were borne on stems differing in anatomical features from those of any known fern. The investigation of the structure of the leaves and their supporting stems led to the recognition of certain extinct genera of Palaeozoic plants of exceptional interest, to which the term generalised type is aptly applied. Associated with anatomical and other characters such as we now regard as the attributes of ferns, these plants exhibit other features not met with in modern ferns but characteristic of a group of seed-bearing plants known as the Cycads. Recent research has revealed the existence of several such generalised types which, by their combination of characters now met with in distinct sub-divisions of the plant-kingdom, clearly indicate the derivation of Ferns, and Cycads as we know them to-day, from a common stock. It was in the first instance by means of anatomical evidence—obtained by the microscopical examination of sections of petrified fragments of stems and leaves—that the generalised nature of these Palaeozoic plants was recognised. Nothing was known as to the reproductive organs. Ferns as now represented in the floras of the world are essentially seed-less plants. As the author of Hudibras wrote:
'Who would believe what strange bugbears
Mankind creates itself, of fears?
That spring like fern, that insect weed,
Equivocally, without seed.'
The reproductive organs or spores borne on the fronds of a fern produce, on germination, a thin green structure, known as the prothallus, less than an inch in length: this bears the sexual organs, and as the result of the union of the male and female cells, the embryo fern-plant begins its existence as a parasite on the inconspicuous prothallus, until after unfolding its first green leaf and thrusting a slender root into the ground, it starts its career as an independent organism[2]. In this life-cycle the seed plays no part.
[2] The life-history of a Fern is clearly described by Prof. Bower in a recent volume in this series.
It is noteworthy that the absence of any indication of spore-capsules and spores, in the case of some of the supposed fern leaves from the Coal-measures, caused some suspicion in the mind of an Austrian Palaeobotanist as to the right of such specimens to be classed among the ferns. This opinion, based in the first place on negative evidence and but little regarded by other authors, has in recent years been proved correct. In 1904 a paper was read before the Royal Society by Professor Oliver and Dr Scott([43]) in which evidence was brought forward pointing to the conclusion that one of these generalised plants bore true seeds. Subsequently Dr Kidston published an account of some specimens of another of these Palaeozoic plants in which was actually shown an organic connexion between undoubted seeds and pieces of a fern-like frond([44]). Without entering into further details, these and similar discoveries may be summarised as follows:—Many of the supposed Fern-fronds of Palaeozoic age, particularly those characteristic of the Coal-measures, are the leaves of plants which in their anatomical characters combined features now shared by true Ferns and by the Cycads. The reproductive organs of these Palaeozoic genera differed widely from those of existing ferns; the male organs, while not unlike the spore-capsules and spores of certain ferns, recall the male organs of living Conifers and Cycads, and the female organs were represented by seeds of a highly complex form. These seed-bearing plants have been called Pteridosperms, a name which expresses the combination of fern-like features with one of the distinguishing attributes of the higher plants, namely the possession of seeds. The ancestors of Pteridosperms are as yet unknown; it is, however, reasonable to assume that there existed in some pre-Carboniferous epoch a group of simple plants from which both Ferns and Pteridosperms were derived. In the forests of the Coal age true Ferns probably occupied a subordinate position in relation to the Pteridosperms.
The question of the relationship between different families of recent ferns and the older known fossil members of the group is beyond the scope of this book. Evidence has been discovered in recent years which warrants the statement that, although none of those Carboniferous ferns were generically identical with existing forms, they very clearly foreshadowed some of those structural features which characterise more than one family of present-day Ferns. The records of the older Mesozoic formations afford abundant evidence of the existence of certain types of Ferns showing a very close resemblance to recent species.
Fig. 9. Osmunda regalis Linn. Fertile frond, (2/5 nat. size.)