Links With the Past in the Plant World

by A. C. Seward

An enquiry into the geographical distribution of living Ferns reveals facts of special interest in connexion with the relative antiquity of different genera and families. The wide distribution of the Bracken fern has already been referred to: it is abundant in Tasmania; its vigour in the island is well illustrated by Mr Geoffrey Smith's statement that constant attention is necessary to keep it from invading newly opened country([45]). On Mount Ophir in the Malay Peninsula the cosmopolitan bracken occurs in association with the two genera Matonia and Dipteris, ferns which are among the most striking examples of links with a remote past and have a restricted geographical range. With Osmunda regalis, the Royal Fern, the Bracken is conspicuous in the marsh vegetation of the Bermudas; it flourishes on the Atlas Mountains, in the Canary Islands, in Abyssinia, on Mt Kenia, in British East Africa, in the Himalayas, and is in fact generally distributed in the tropics in both the north and south temperate zones.

The Royal Fern ([Fig. 9]) is another British species with a wide distribution; it occurs in Northern Asia and in North America; it is common in the Siberian forests and lives in several tropical countries, extending to Southern India and Cape Colony, and in South America it is represented by a closely allied species. Though at the present day Osmunda regalis is one of the rare English Ferns, its occurrence in the submerged forest-beds round our coasts and in pre-Glacial beds points to its former abundance in the British area generally. The Royal Fern is a member of a family now represented by two genera, Osmunda and Todea.

With the exception of Todea barbara, with its large spreading fronds and a short root-covered stem, which occurs in Australia and Cape Colony, all the species of this genus are filmy ferns with semitransparent fronds adapted to a moisture-laden atmosphere. The maximum development of the genus is in New Zealand.

Todea barbara affords an instance of discontinuous distribution; it was no doubt once widely spread in circumpolar regions and now survives only in South Africa and in Australia.

There are satisfactory reasons for regarding the Bracken Fern, with its world-wide range in present-day floras, as a comparatively modern species now in full vigour. Its anatomical and other features are consistent with the view that it is a late product of evolution, and as yet no indication has been given by the records of the rocks of an ancient lineage. The Osmunda family, on the other hand, is undoubtedly an extremely old branch of the fern group. A comparison of the Royal Fern with the Bracken shows that their stems are constructed on very different plans, and we have good reasons for speaking of the structural peculiarities of the former as those of a more primitive type. Moreover, the discontinuous geographical range of some members of the Osmunda family is in itself an indication of antiquity. There is another point which may have a bearing in this question of antiquity, namely the fact that the spores of Osmunda are green and do not possess the powers of indurance inherent in the spores of the majority of ferns which are not green. It has recently been contended by Professor Campbell of Stanford University that the delicate green spores of the Liverworts, plants closely allied to the Mosses, constitute an argument in favour of the antiquity of these plants([46]). Certain Liverworts are cosmopolitan in their range, e.g. the genera Riccia and Marchantia.

If certain genera are widely distributed, notwithstanding the fact that their reproductive cells, by which dispersal is effected, are ill-adapted to withstand unfavourable conditions or to endure prolonged desiccation, it would seem reasonable to conclude that their emigration has been accomplished slowly and with difficulty. Ferns such as Osmunda, with green and short-lived spores, would thus be handicapped in competition with other genera provided with more efficient means of dispersal and better equipped for the vicissitudes of travel.

The inferences as to antiquity deduced from a study of the existing species of Osmunda and Todea receive striking confirmation from the testimony of fossils. Some of the oldest known Palaeozoic ferns, though differing too widely from the existing Osmundas and Todeas to be included in the same family, afford distinct glimmerings of Osmundaceous characters, which at a later period became individualised in the direct ancestors of the modern forms. Our knowledge of the past history of the Osmunda family has recently been considerably extended and placed on a firmer basis by the researches of Dr Kidston and Professor Gwynne-Vaughan. These authors have recognised in some exceptionally well-preserved fern-stems from Permian rocks in Russia, anatomical features which point unmistakably to close relationship with the recent members of the family([47])([48]).

Passing higher up the geological series, fertile fern fronds with spore-capsules and spores practically identical with those of Osmunda have been found in the Jurassic plant-beds of Yorkshire and in rocks of approximately the same age in many parts of the world. From Jurassic strata in New Zealand a petrified fern-stem has been described (Osmundites Dunlopi), almost identical in structure with the surviving species. Cretaceous and Tertiary examples of similar ferns might be quoted; but enough has been said to establish the claim of the Royal Fern and other members of the Osmunda-family to an ancestry which possibly extends even farther back than that of any other existing family of Ferns.