Primary tabs



Asia-Tropical: Australasia: E. Africa: present Mascarene Islands: present Pacific: present Tahiti: present tropical mainland Asia: present
Throughout Malesia (152 spp.), tropical mainland Asia (12 spp.), Mascarene Islands and E. Africa (4 spp.), Australia and the Pacific (17 spp., only 1 reaching Tahiti).


John Smith's name Sphaerostephanos refers to the yellow glands which fringe the indusium of the type species. Copeland and Christensen treated this name as feminine, but as the Greek stephanos is masculine, I regard the Compound name also as masculine. Smith's original specimen had broken indusia (carefully illustrated by Francis Bauer) from which he gained an erroneous idea of their structure. Soon afterwards Smith reeeived from Robert Brown a better specimen, and his comments on it were published in 1840 under Brown's name Mesochlaena which Smith regarded as having priority; he then realized that it differed from Nephrodium (as interpreted by Schott) only in having an elongate sorus. Brown's name Mesochlaena had been published in 1838 with a very slight description and without any indication of its status, so that it cannot be regarded as a valid generic name, as pointed out by Copeland in 1929. The first good illustration of the sorus was published by Mettenius () with the name Mesochlaena javanica; this was changed to Nephrodium javanicum by Hooker who published an excellent illustration of a whole plant (), with Aspidium polycarpon Bl. as a synonym and reference to Smith's comments of 1840.

When Beddome published he also reverted to the genus Nephrodium for Aspidium polycarpon Bl., and described an additional Malayan species N. larutense which also has elongate sori. In , Christensen retained the generic name Mesochlaena for M. polycarpa but transferred Beddome's N. larutense to Dryopteris. Van Alderwerelt accepted the generic name Mesochlaena in , transferring N. larutense to it and later adding two more species. In the third Supplement to Index Filicum Christensen followed Copeland in adopting the name Sphaerostephanos and transferred to it v.A.v.R.'s names.

The species thus brought together in Sphaerostephanos are not a closely allied group, and there are others (e.g. S. norrisii) which have slightly elongate sori, so that no clear line can be drawn between species with elongate and those with round sori. For these reasons, in 1955 I transferred the Malayan species with elongate sori to Cyclosorus, a genus which I accepted in the sense of Ching (1938), though commenting on the unsatisfactory nature of generic delimitation in the family Thelypteridaceae at that time. Ching however had misinterpreted the sorus of the type species of Sphaerostephanos and had wrongly reported the spores as trilete, for which reasons he proposed for it a new family ().

When I had made a preliminary study of all known species of Thelypteridaceae in preparation for the present work, I made a survey of all existing generic names and found that Sphaerostephanos was the oldest name for a large group of mainly Malesian species most of which do not have elongate sori. For this reason in 1971 I published a revised concept of the genus, which is here adopted. Species of this genus were variously placed in Iwatsuki's Classification of 1964, which makes no reference to the majority of Malesian species; I cite above only infrageneric taxa which he typified by Malesian species.

The main characters of this genus are the presence of much-reduced basal pinnae and of sessile non-resinous spherical yellow glands of an easily recognized type; these glands are not destroyed in the process of drying specimens for the herbarium as those of. Coryphopteris and Amphineuron often are. The species can be arranged in a series beginning with those that have glands on both surfaces of pinnae and on indusia and sporangia, then those lacking glands on the upper surface (there are no species bearing glands on the upper surface and not on the lower), those with glands only on sori and finally those with no glands at all. Species without glands agree with Pneumatopteris in having reduced basal pinnae but have a much greater complement of acicular hairs on all parts (especially on sporangia); they lack also short capitate hairs and the peculiar pustules on the lower surface of dried fronds which are characteristic of Pneumatopteris.

I also found that some species had very few reduced pinnae and that there is not a sharp distinction in this character between Sphaerostephanos and Pronephrium. This matter is discussed under Pronephrium; I believe that the great majority of species belong clearly to one genus or the other.

As the sessile spherical yellow glands are the most distinctive feature of the genus, I regard the development of these on all parts of the frond as a primitive character. Assuming also that the original species had an erect caudex, S. polycarpus might be regarded as a prototype for the genus. It is one of the largest species vegetatively. But it is not a forest plant (a reason for its wide distribution) and it has deeply lobed pinnae, whereas the most widely distributed Malesian forest species with glands on both surfaces is S. penniger, which has shallowly lobed pinnae with several pairs of anastomosing veins; it is allied to the arborescent S. arbuscula (Willd.) Holttum of Southern India and the Mascarene Islands. So I suggest that the primitive Sphaerostephanos had shallowly-lobed pinnae with anastomosing veins, and that S. arbuscula is the nearest surviving species. Several New Guinea species have a slender erect caudex and shallowly-lobed pinnae (e.g. S. arfakianus) but have no glands on the upper surface. At least it can be said with some certainty that (as in Christella) the few species which have free veins are probably not primitive but represent a secondary development occurring locally in a few areas (e.g. S. inconspicuus in Borneo, S. novoguineensis in New Guinea and S. pycnosorus in Samoa); they do not form a coherent natural group.

A few species have quite gradually reduced basal pinnae (e.g. S. hastatopinnatus and S. arfakianus, both with an erect caudex) but most have an abrupt (or almost abrupt) transition from normal to reduced pinnae at the base of the frond. A few species which have many reduced pinnae are somewhat intermediate, showing first a gradual reduction downwards and then an abrupt transition to much smaller pinnae (e.g. S. hispiduliformis). I suggest that the condition of gradual reduction is the more primitive arrangement, as it alone occurs in Cyathea. Apart from Sphaerostephanos and Pneumatopteris, which are predominantly Malesian in distribution and certainly allied, most other genera have basal pinnae gradually or not at all reduced.

The species with long-creeping rhizome are S. unitus, a gland-bearing species very widely distributed with distinct eastern and western varieties, and S. invisus which lacks glands and is mainly distributed in the Pacific (its distribution is similar to that of Pneumatopteris costata). S. hirtisorus (C. Chr.) Holttum in N.E. India and W. China appears to be related to S. invisus.

The following key is based on the distribution of glands on the surfaces of pinnae and on sori. I believe that this represents an evolutionary trend, but within it there are many separate bifurcations which, in such a complex group, are not easy to discern. For example, there is a group of species which are related to S. stipellatus and have closely appressed hairs on the lower surface of costae and costules; this character may be associated with elongate aerophores and lack of glands on the lower surface of pinnae (as in S. stipellatus itself) or with varying combinations of these and other characters. Varying trends may lead on to other groups (e.g. elongate aerophores may be associated with erect hairs on the lower surface of costae) and I cannot see how to follow all the trends or delimit sub-groups. It seems to me evident that the present condition of the genus is due to rapid and complex recent evolution, especially in New Guinea, resulting in a difficulty in the delimitation of species, with some necessary inequality of treatment. The destruetion of Malesian forests may put a stop to this process or alter it in unpredictable ways.

In some species where glands are not abundant there seems to be some Variation in their occurrence; in such cases a species may appear in more than one place in the key.


Base chromosome number 36; 8 spp. examined, all diploid.


Ching - in Sunyatsenia. 1940: 240
J. Sm. - in Hook., J. Bot. 3. 1840: 18
emend. Holttum - in Blumea. 1971: 39
Holttum - in Novit. Bot. Inst. Bot. Univ. Carol. Prag. 1969: 48
sensu Holttum - in Rev. Fl. Mal. 1955: 255
C. Chr., Ind. Fil. 1905: xxii.
Copel., Gen. Fil. 1947: 144
Copel. - in Univ. Cal. Publ. Bot. 1929: 60
J. Sm. - in Allertonia. 1977: 201
J. Sm. - in Kalikasan. 1975: 47
sensu Copel., Gen. Fil. 1947: 140