Untitled 256 R E I N W A R D T I A [VOL. 4 Umbellularia Nutt. 195, 201, 202, 203, 216, 235; californica (Nees) Nutt. 202, 205, 235. Urbanodendron Mez 195, 200, 207, 210, 217, 238; verrucosum (Nees) Mez 238. Virola 220. Volutella Forsk. 245. Wimmeria Nees ex Meissn. 229. Yushuma Kam. 234. R E I N W A R D T I A Published by Herbarium Bogoriense, Kebun Raya Indonesia Volume 4, Part. 2, pp. 257-280 (1957) FLORAE MALESIANAE PRAECURSORES XVI ON THE TAXONOMIC SUBDIVISION OF THE GLEICHENIACEAE, WITH DESCRIPTIONS OF NEW MALAYSIAN SPECIES AND VARIETIES by R. E. HOLTTUM * SUMMARY A new subdivision is given of the fern family Gleicheniaceae. The genus Platyzoma R. Br is excluded from the family. The genus Stroma- topteris from New Caledonia is arranged in a distinct subfamily. In the remainder of the family, subfamily Gleichenioideae, two genera are recog- nized, Gleichenia (with subgenera Diplopterygium, Gleichenia, and Mertensia) and Dicranopteris (with subgenera Acropterygium and Dicra- nopteris) . The problem of subdividing the family is discussed with reference to former treatments and to new data, and a conspectus of the new system, with synonymy and key to the genera and subgenera, is given. A number of new species, new varieties, and new combinations is made both in Gleichenia and Dicranopteris. In preparing a taxonomic revision of the family Gleicheniaceae for Flora Malesiana, I have reviewed the status of the genera proposed within the family by previous authors, and as a result have been led to take a position midway between the arrangement of Christensen (Index Filicum 1905) and of Copeland (Genera Filieum 1947). The present paper gives a summary of the facts on which this decision was reached; a fuller comparative treatment of the subject, with a discussion on morphology and growth-habit in this and other families of primitive ferns, will be published elsewhere. The genus Platyzoma R. Br., which has usually be included in Gleich- eniaceae (even in the genus Gleichenia), appears to me so different that it should be excluded. A statement on this subject has been published separately (Kew Bulletin 1956; 551) ; the genus will not be further men- tioned in the present paper. The genus Stromatoptetis, confined to New Caledonia, is peculiar in various ways, but has superficial sori of sporangia which agree with 1 Formerly Professor of Botany in the University of Malaya. — 257 — 258 REINWARDTIA [VOL. 4 those of Gieichenia. In dermal appendages and growth-habit it is quite distinct from other members of the family, so that I would agree with Nakai in assigning it to a distinct subfamily. A summary of its important characters is given in the conspectus below. The remainder of the family is clearly divisible into four groups, which are all given generic status by Copeland, as Gieichenia, Hicriopteris, Sticherus, and Dicranopteris. Many previous authors, from Willdenow onwards, were impressed by the difference between the genus Gieichenia and the other three. A main distinction (repeated by Presl, Hasskarl, Diels, Underwood and Nakai) is said to be that in Gieichenia proper the sori are terminal on the veins, in the other three groups the sori are not terminal. This statement is not true. The veins are short in Gieichenia proper, but the sorus is not terminal, as can be seen in cleared specimens (examined by me in G. microphylla R. Br., G. vulcanica Bl., and G. dicarpa R. Br.). If one looks for another distinction between Gieichenia proper and Copeland's genus Hicriopteris, the only one is the very small size of the lobes of the lamina of Gieichenia, and this is hardly a generic distinction. A basic division of the family is therefore not between Gieichenia proper and the rest. As has been shown in various other groups of ferns, the nature of the dermal appendages is often significant as an indicator of relation- ships. In the present family it is usually stated that Dicranoptens (sensu Copeland) differs from the other three groups in having only hairs, the others having both hairs and scales. This is however an undue simpli- fication, because the hairs of Dicranopteris are not like the hairs of the other genera of Copeland. In Gieichenia, Hicriopteris, and Sticherus, the hairs are all stellate, with subequal unicellular rays, and there is a gradual transition from such stellate hairs to peltate scales the marginal cells of which have outgrowths like the rays of the stellate hairs. In Dicranopteris most of the hairs consist of a row of cells, with one or more branches (unicellular or not) at the base, and often also outgrowths from the other cells of the hair; nothing like this occurs in Copeland's other three genera. Dicranopteris differs from the other groups also in two other im- portant ways. The sporangia in Dicranopteris are much smaller than in the other groups, and there are more sporangia in each sorus (see Bower, The Ferns 2, ch. XXIV) ; the sori also lack paraphyses which are usual in the other groups. The veins in Dicranopteris are always forked two or three (even four) times, whereas they are simple or once forked in the other groups. 1957] R. E. HOLTTTJM: Gleicheniaceae 259 Dicranopteris is therefore sharply distinct from the other groups, and in my opinion should rank as a separate genus. The other three groups agree in scales and hairs, in sori and sporangia, and (apart from the reduced condition of Gieichenia proper) in venation. Gieichenia proper is distinct only in the very small lobes of its lamina, and in its simple veins, which are clearly a consequence of the reduction of the lamina. The only important differences between the three groups are different modifications of the same basic habit of branching. Hicriopteris (sensu Copeland) has the simplest branching-habit. It bears successive pairs of bipinnatifid pinnae; while such a pair of pinnae is developing, the apex of the main rhachis rests, resuming growth later, the whole frond being of indefinite growth in length. Gieichenia proper has at first an exactly similar method of branching, on a miniature scale, but in large fronds each primary rhachis-branch is again branched, bearing a pair of secondary branches (usually with a dormant apex be- tween them), and strong plants may have ternary or even quaternary branches; in all cases the ultimate branches are bipinnatifid as in Hicri- opteris. In Sticherus the same repeated branching occurs, with dormancy of the apex between each pair of branches, but the ultimate branches are simply pinnatifid, not bipinnatifid. I have however seen a monstrous specimen of Gieichenia {Sticherus) milnei Bak. in which the ultimate branches are bipinnatifid. This bridges the gap between Sticherus and Copeland's other two genera. It a|gsars to me therefore that the differences between Copeland's Gleich Hicriopteris, and Sticherus are relatively unimportant as compared with their common differences from Dicranopteris, and I propose to recognize two genera, Gieichenia (with subgenera Gieichenia, Diplopterygium, and Mertensia) and Dicranopteris. The name Diplopterygium was used by Diels as a sectional one, in his Gieichenia subg. Mertensia. I use it in place of Hicriopteris because the species originally called Hicriopteris by Presl (H. speciosa) was in my opinion a Dicranopteris. Presl's description is very detailed, and agrees closely with the very peculiar structure of Gieichenia opposita v.A.v.R., which is like that of no other member of the family (see my Ferns of Malaya, p. 70 and fig. 14 F). The use of Mertensia as a subgeneric name in place of Sticherus is explained in a later section of the present paper. The above discussion makes no mention of the most recent re- arrangement of Gleicheniaceae, by Nakai, who attempted to use spore- torm as a basic character for the subdivision of the family. He made 260 R E I N W A R D T I A [VOL. 4 1957] R. E. HOLTTUM: Gleicheniaceae 2:61 a mis-statement of fact in this respect; a rectification of this error com- pletely upsets his scheme. Having separated Stromatopteiis, Nakai divided the rest of the family into two subfamilies, Sticherioideae with bilateral spores and Gleichenioideae with tetrahedral spores. In Sticherioideae he placed Acro- pterygium {Dicranopteris subgenus Acropterygium of the present scheme) and Sticherus; in Gleichenioideae he placed Gleichenia proper (subdivided into three genera), Dicranopteris (including only D. linearis and its near allies) and Diplopterygiwn. But within the group of D. linearis and its near allies, all supposed by Nakai to have tetrahedral spores, some have in fact bilateral spores. To remove these into the other subfamily would clearly be very unnatural. The scheme thus breaks down; and in fact all other evidence, which I have set forth above and summarize in the conspectus below, is against it, and in my view it is a completely unnatural arrangement. CONSPECTUS OF THE FAMILY GLEICHENIACEAE 1. Fronds unbranched (apart from occasional true dichotomy), borne on erect branched stems arising from a creeping rhizome; stellate or branched hairs lacking on vege- tative parts of plant; paraphyses in the form of small irregular scales with projecting marginal cells; frond-bearing stems partly covered with non-peltate scales and bearing long simple hairs. Subfamily Stromatopteridoideae: only genus Stromatopteris 1. Fronds of fully developed plants branched pseudo-dichotomously (forked), often many times, with dormancy (periodic or permanent) of the apices within the forks; fronds borne directly on a creeping rhizome; stellate or branched hairs, or fringed scales, or both, always present on vegetative part of plant. Subfamily Gleichenioideae 2. Hairs stellate with rays consisting of single cells, or simple and very short; scales peltate, with fringe of outgrowths from marginal cells (in G. laevissima margins entire); sori of 2—4 large sporangia, with paraphyses; lateral veins in a lobe of the lamina simple or once forked Gleichenia 3. Ultimate branches pinnate with deeply lobed leaflets 4. Dormancy confined to a periodic condition of the branch which will con- tinue the main rachis of the frond; lobes of lamina each with costule and once-forked lateral veins; sori several on each lobe. Subgen. 2. Diplopterygium 4. Dormancy often occurring, and persistent, on other branches of the frond; tinue the main rachis of the frond; lobes of lamina each with costule and veinlets all simple; one sorus on each lobe. . . . Subgen. 1. Gleichenia, 3. Ultimate branches bearing a deeply pinnatifid lamina, its lobes entire or slightly toothed Subgen. 3. Mertensia 2. Hairs consisting of a row of cells, with one or more branches at the base and often outgrowths also from other cells; no scales; sori of 8—15 or more sporangia, lacking paraphyses;' lateral veins in a lobe of the lamina forked 2 4 times Dicranopteris 5. An accessory branch, of same form as the ultimate branches, often present on the basiscopic side at the base of each branch from a fork; vascular system of rhizome a protostele. . . Subgen. 1. Dicranopteris 5. Accessory branches lacking; vascular system of rhizome a solenostele. Subgen. 2. Acropterygium STATUS OF NAMES, AND IMPORTANT SYNONYMS (for references to literature prior to 1933, see Christensen, Index Filicum) Subfamily Stromatopteridoideae Nakai, Bull. Nat. Sc. Mus. Tokyo 29: 32. 1950. Genus Stromatopteris Mettenius 1861 Only species: S. moniliformis Mett. Subfamily Gleiehenioideae Nakai, I.e. (non sensu Nakai) Genus Gleichenia Smith 1793 (conserved name) TYPE SPECIES.—G. polypodioides (L.) Sm. (Onoclea Linn. 1771) Subgenus 1. Gleichenia Calymella Presl 1836, Ching 1940 (Sunyatsenia 5: 285), Nakai I.e. 40 Gleicheniastrum Presl 1848, Nakai I.e. 42 Gleichenia subgenus Eugleichenia Diels 1900, Bower 1926 Subgenus 2. Diplopterygium (Diels) Holtt., stat. nov. Gleichenia subgenus Mertensia sect. Diplopterygium Diels 1900 Dicranopteris sensu Underwood 1907, p.p. Gleichenia subgenus Dicranopteris, p.p., Bower 1926 Hicriopteris sensu Ching I.e. 277, Copeland (Gen. Fil. 1947; p. 28), non Presl Diplopterygium Nakai I.e. 47 TYPE SPECIES.—G. glauca (Thunb.) Hook. 1844 (Polypodium Thunb. 1784) Subgenus 3. Mertensia (Willd.) Diels, p.p. Mertensia Willd. 1804, p.p. Sticherus Presl 1836, Ching I.e. 281, Copel. I.e. 27, Nakai I.e. 7 Gleichenia subgenus Mertensia sect. Holopterygium Diels 1900 Dicranopteris sensu Underwood 1907, p.p. Gleichenia subgenus Dicranopteris, p.p. Bower 1926 TYPE SPECIES.—G. truncata (Willd.) Spr. 1827 {Mertens'.a Willd. 1804) Genus Dicranopteris Bernhardi 1806, Copeland I.e. 28, Underwood 1907 (p.p.) Gleichenia subgenus Eudicranopteris Bower 1926 TYPE SPECIES.—D. dichotoma (Thunb.) Bernh. {Polypodium Thunb. 1784) (Polypodium lineare Burm. 1768 is an earlier name) Subgenus 1. Dicranopteris Gleichenia subgenus Mertensia sect. Heteropterygium Diels 1900 Hicriopteris Presl 1851 (non Ching nee Copel.) Dicranopteris sensu Nakai I.e. 56, sensu Ching I.e. 272 Subgenus 2. Acropterygium (Diels) Holtt., stat. nov. Gleichenia subgenus Mertensia sect. Acropterygium Diels 1900 Acropterygium Nakai, I.e. 5 Gleichenella Ching, I.e. 276 TYPE SPECIES.—Mertensia pectinata Willd. 1810. 262 R E I N W A R D T I A [VOL. 4 G L E I C H E N I A Subgenus 1. GLEICHENIA In his revision of 1940 (Sunyatsenia 5: 269-288), Ching used the name Calymella Presl for this group of species, as the name Gleichenia is a later homonym (conserved in 1954). Ching subdivided his genus into two sections, Eu-Calymella and Gleicheniastrum (which latter Presl had regarded as a separate genus). The basis of the subdivision was that in Eu-Calymella, the sori are immersed, in Gleicheniastrum they are super- ficial. The species in Ching's Eu-Calymella have the margins of the leaflet-lobes so much reflexed that these lobes form pouches with only small apertures on the lower surface; the sporangia are superficial on the inner surface of the pouch. The lobes of the lamina of Ching's sub- genus Gleicheniastrum are almost flat; but some of them have the sorus sunk in a distinct depression in the substance of the lamina; these sori are thus immersed in a different sense, not indicated by Ching. Nakai in 1950 (I.e.) revived both Presl's genera Calymella and Gleicheniastrum as distinct from Gleichenia proper. The basic of the distinction was that in Gleichenia proper the sori are immersed in cavities in the substance of the lamina, in Calymella they are in pouches as above described, and in Gleicheniastrum they are on the surface of flat leaflet- lobes. But in one species of Nakai's Gleicheniastrum {Gleichenia micro- phylla R. Br.), each sorus is in a distinct cavity, though the edges of this are not raised as they are in the species included by Nakai in Gleichenia. The distinction between Gleichenia and Gleicheniastrum is thus not a sharp one; and indeed all these distinctions are relatively trivial, not of generic significance. Among Malaysian representatives of Gleichenia subgen. Gleichenia, I have found one new variety which needs to be described. GLEICHENIA PELTOPHORA Copel. var. schizolepis C. Chr. ex Holtt., var. nov. Paleae omnes ciliatae; segmenta laminae infra pilis stelliformibus, non paleis orbicularibus, instructa. BORNEO. S a r a w a k , Mt Murud, 5000—6000 ft, Mjoberg 9 (P, type). . In the typical form of this species (both in New Guinea and in Borneo) there are almost entire circular peltate scales on the rhizome, and also smaller ones on the lower surface of the segments of the lamina; in this variety the rhizome-scales are fringed, the smaller scales of the 195.7] R. E. HOLTTUM: Gleicheniaceae 263 lamina are replaced by stellate hairs. The epithet "schizolepis" was given in sched. by Christensen. Subgenus 2. DIPLOPTERYGIUM The species of Malaysia fall very distinctly into two groups, which are distinguished by the nature of the scales on the dormant apex of the main rhachis. In the group of G. longissima Bl. the scales are broad, with long flexuous pale marginal hairs. This group includes few species, and G. longissima itself is distributed throughout almost the whole of Malaysia, and beyond Malaysia both northwards and into the Pacific. In the other group of species, of which G. norrisii Mett. was first de- scribed, the scales are narrower, with short, rigid, rather oblique marginal setae which are very dark, or concolorous with the rest of the scale. This group is polymorphic, with a number of local species and no species which is widely distributed. Five new species of this group are described below, and a species originally placed in the genus Hicriopteris is transferred to Gleichenia. Gleichenia angustiloba Holtt., sp. nov. Rhizoma 7 mm diametiente, primo paleis nitidis atrobrunneis 5—7 mm longis 1.5 mm latis vestitum, denique basibus palearum verrucosum; stipes 6—7 mm diam., basi primo modo rhizomatis paleaceus, denique verrucosus, parte superiore modo rhacheos paleis parvis yestitus, denique asperulus; rami rhacheos (pinnae) 100 cm vel ultra longi; foliola stipuli- formia late deltoidea, ad 3 cm longa, pinnatifida, lobi infimi profunde anguste-lobati; paleae rhacheos atrobrunneae, nitidae, 3—4 mm longae, angustae, margine setis brevibus rigidis concoloribus obliquis instructae; rami costaeque utrinque (et costulae infra) paleis minoribus multiformibus et pilis stelliformibus rufo-brunneis plus minusve persistentibus vestiti; pinnulae ad 15 cm longae et 3 cm latae, inter se 2.0—2.5 cm distantes (i.e., pinnulae contiguae vel imbricatae); segmenta omnia laminae c. 2.5 mm lata, sinubus latis separata; segmenta plurima basi constricta, ala laminae 0.2 mm lata costae utrinque commissa; segmenta infima foliola disjuncta constituta; costulae 4 mm inter se distantes; venae utrinque prominentes; sori sporangiis 4 constituti, paraphysibus crispis longis rufo-brunneis instructi; venae frondis adultae praeter paraphyses gla- brae. EAST NEW GUINEA. Mt Tafa, 2400 m, Brass 4960 (BM, type; K). This species is nearly allied to G. novoguineensis Brause (with which Hicriopteris astrotricha Copel. is identical), differing in the broader pinnules of which all the segments are narrow, separated by rather wide sinuses, several pairs of segments being constricted at the base and con- nected only by a very narrow wing of the lamina. 264 RE I N WAR D T I A [ V O L . 4 Gleichenia brevipinnula Holtt., sp. nov. Rhachis primaria c. 5 mm diametiente, apex dormitans paleis 5 mm longis, haud 0.5 mm latis, brunneis, margine nigra setiferis (setae breves, nigrae, obliquae) vestitus; rami c. 70 cm longi; foliola stipuliformia 3.5 cm longa, late deltoidea, profunde lobata, lobi omnes anguste caudati, lobi infimi profunde et oblique lobulati; segmenta pinnularum inferiorum similiter anguste caudata; pinnulae c. 2.0 cm inter se distantes, ad 10 cm longae et 2.0 cm latae, superiores non deflexae; costulae 3—3.5 mm inter se distantes, leviter obliquae; lamina usque 1 mm costae incisa: segmenta laminae supra basin haud angustata, sinubus angustissimis separata, rigida, apice late rotundata, apicibus venularum prominentibus leviter dentata, marginibus non reflexis, subtus pallide glauca; venae utroque prominentes; rami rhacheos utrinque pilis stelliformibus rigidis atro- brunneis plus minusve persistentibus, paleis angustis paucis intermixtis, vestiti; costae et costulae infra pilis similibus praeditae; venae subtus pilos stelliformes ferentes; sori 3 vel 4 sporangiis constituti, paraphyses nullae. BORNEO. S a r a w a k , Pueh (Poi) Range, summit ridge, 4400 ft, P.R. Bell 2042 (BM, type); Tinjar, Baram River, 3000—4000 ft, Mjoberg 8 (P); Bau, Anderson SJf (SING.). N o r t h B o r n e o , Kinabalu, Marei Parei Spur, Topping- 1872 (US); Kinabalu 8000 ft, Clemens 31994 (K); Kinabalu 8000 ft, L.S. Gibbs 4.256 (BM). Gleichenia sumatrana Holtt., sp. nov. Rhachis primaria c. 8 mm diametiente, paleis parvis coarctis et paleis majoribus adspersis persistentibus vestita; rami rhacheos infra similiter vestiti; paleae maximae 10 mm longae, basi 0.5 mm latae, apice filiformes, margine atrobrunneo setis rigidis atrobrunneis instructae, medio pallidae; paleae minimae et pili stelliformes omnino pallide ferrugineae; rami rhacheos ad 150 cm longi; pinnulae maximae c. 15 cm longae, 2.5 cm latae, 2.5—3.0cm inter se distantes; pinnulae superiores parvae, proximiores, rectangulariter patentes; costulae 4—4.5 mm inter se distantes; segmenta laminae c. 3.5 mm lata, inferiora 7—8-juga basi constricta ala laminae angustissima conjuncta; segmenta media sinubus haud l m m latis sepa- rata ; lamina tenuis, infra in vivo non glauca, venulae utrinque prominen- tes ; foliola stipuliformia 2.5 cm longa, 1.5 cm lata, bipinnatifida, lobi angustissimi; costae pinnularum infra modo ramorum rhacheos paleaceae (paleae minores), apicem versus glabrescentes, pallidae; costulae paleis parvis fimbriatis (f imbriae atrobrunneae) persistentibus vestitae; venulae infimae pilos stelliformes ferentes, venulae superiores plerumque glabrae; costae supra leviter canaliculatae, pallidae, paleis parvulis atrobrunneis fimbriatis sat dense vestitae. N. SUMATRA. Near Parbuluan, south of Sidikalang (west of Toba Lake), 1500 m, Alston 14.84 (BM, type); G. Talamau, 1900 m, BUnnemeyer 884 (BO). Karo Plateau, Bartlett 8543 (US). 1957] R. E. HOLTTUM: Gleieheniaoeae 265 This species resembles most nearly G. sordida Copel. from Mindanao; the latter is larger, with fewer constricted basal lamina-segments, and has more small pale rusty stellate hairs or small scales on the rhachis and veins. Gleichenia matthewii Holtt., sp. nov. Apex dormitans rhacheos paleis haud 0.5 mm latis, margine atro- brunneis, nitidis, setiferis, medio pallidioribus, vestitus; rami rhacheos 70 cm longi, praeter basis valde flexuosi; pinnulae omnes deflexae, super- iores angulo 45° deflexae, supremae in lobos rectangulariter patentes laminae terminalis transgredientes; pinnulae inferiores inter se 4 cm distantes, c. 12 cm longae et 2.5 cm latae; costulae 4.5 mm inter se distantes; lamina usque 1—1.5 mm costae incisa; segmenta laminae ob- longa, contigua, apice fere truncata (vulgo leviter retusa); foliola stipuli- formia deltoidea, c. 2 cm longa, anguste lobata; lamina firma, infra glauca; venulae utrinque prominentes; costae basin versus paleas parvas paucas ferentes, folium cetera omnino glabra; sori non visi. C. SUMATRA, G. Singgalang, Matthew s.n., 31 Jan. 1912 (K, type); same locality at 1800 m. alt., Schiffner P. H (L). This species agrees with G. norrisii in its almost glabrous lamina and in the deflexed pinnules, the latter character (and the strongly zig-zag rhachis) more pronounced than in G. norrisii. It differs from G. norrisii in its dark-edged scales, in the narrowly-lobed stipular leaflets, and in the almost oblong segments of the lamina which are almost or quite conti- guous. The name matthewii commemorates Fleet-Surgeon C. G. Matthew, who collected ferns with discrimination during his travels in SE. Asia while on naval service and subsequently. GLEICHENIA VOLUBILIS Jungh. var. peninsularis Holtt., var. nov. Paleae ramorum rhacheos et costarum ferrugineae, ciliis concoloribus praeditae. MALAY PENINSULA. K e d a h - P e r a k boundary, G. Bintang, F.M.S. Museum s.n., June 1917 (K, type). In the typical form of the species (in Java and Sumatra) the scales are very dark, with pale fringes, and contrast with the dense felt of rusty stellate hairs with which they are mixed. In this variety, the scales and hairs are of about the same colour. Gleichenia paleacea (Copel.) Holtt., comb. nov. Hicriopteris paleacea Copeland in Philip. J. Sci. 81: 3. 1952. 266 REIN WARDTI A [VOL. 4 1957] R. E. HOLTTUM: Gleicheniaceae 267 Gleichenia papuana Holtt., sp. nov. Var. PAPUANA Apex dormitans rhacheos primarii paleis atrobrunneis nitidis 4 mm longis, 1 mm latis, margine pilis pallidis patentibus ad 0.5 mm longis ciliatis vestitus; rami rhacheos costaeque infra paleis numerosis similibus minoribus diu persistentibus praediti, tendem glabri et verruculosi (paleae costarum c. 1.5 mm longae, 0.5 mm latae); costulae infra paleis minutis fimbriatis multis vestitae; rami rhacheos supra paleis angustissimis pilisque ferrugineis sat dense vestiti; rami rhacheos 120 cm longi; pin- nulae rectangulariter patentes, 2.5 cm inter se distantes, maximae 12 cm longae, 2 cm latae; costulae 2.5—3 mm inter se distantes; lamina tenuis; venulae infra leviter prominentes, supra planae; sori vulgo sporangiis 3, paraphysibus crispatis brevibus intermixtis, constati. NEW GUINEA. P a p u a . Milne Bay Distr., Mt Dayman, 2250 m, Brass 22597 (BM, type); scrambling 2—3 m; leaves grey below; forming dense tangles, edge of mossy forest. GLEICHENIA PAPUANA Holtt. var. membranacea Holtt., var. nov. Paleae apicis rhacheos margine pallidae tenuissimae, ciliis brevissimis praeditae; paleae ramorum rhacheos costarumque numerosiores, basin versus tantum atrobrunneae, cetera pallidae. NEW GUINEA. P a p u a . Milne Bay Distr., Goodenough I., 1600 m. Brass 24768 (BM, type); scrambling to 2—3 m, plentiful in openings in forest This species is allied to the very widespread G. longissima BL, but is smaller and persistently scaly. The dark scales are quite different from those of G. paleacea (Copel.) Holtt. The variety membranacea differs only in the very thin translucent margins of the scales; these margins take the place of the long fringe of pale hairs in the typical form of the species. Subgenus 3. MERTENSIA The generic name Mertensia was proposed by Willdenow for all members of the family Gleicheniaceae other than Gleichenia subgenus Gleichenia of the present treatment. As argued above, it appears to me that this is an unnatural division of the family, the true division being between Dicranopteris and the rest. Diels used Mertensia as a subgenus in a sense exactly equivalent to its use by Willdenow as a genus. Here it is used in a restricted sense. As a generic name, Mertensia Willd. is antedated by Mertensia Roth (Boraginaceae) and is therefore illegitimate. If Mertensia is used as a subgeneric name in the restricted sense now proposed, it takes the place of Sticherus Presl of Copeland's treatment. The disappearance of Sticherus is not to be regretted, as Presl's descrip- tion, based on published descriptions of specimens he had not seen, is partly erroneous and altogether unsatisfactory, as already pointed out by Copeland. Only one species (G. truncata) is common and widely distributed in the lowlands of Malaysia. The others are all mountain plants; they some- times form small thickets, but none are as large or as rampant as G. truncata. Gleichenia hirta (as here interpreted) is the most widely dis- tributed; G. hispida and G. vestita have also wide distributions. A few species appear to occur only in New Guinea, and one only in Sumatra. Two species have been described from immature or stunted plants (G. erecta C. Chr. and Sticherus pinnatus Copel.); these are here united with other species. In this subgenus the characters of the scales are always important for the discrimination of species. Fronds which are not yet fully expanded (immature fronds as distinct from immature plants) are always com- pletely covered with scales and hairs for a time; the extent to which these scales and hairs persist on the various parts of a frond is characteristic of a species (or of varieties within a species), but this kind of character is not always easy to judge from dried specimens. Plants from higher altitudes or more exposed positions also may be more scaly than plants of the same species from lower elevations or more sheltered places. The number of times a primary branch of a frond is forked may also be characteristic of a species, but again this is not easy to judge from dried specimens; more observations on variation of branching habit of plants in one locality is needed for a proper understanding of some species (notably of the varieties of G. hirta). Gleichenia alstonii Holtt., sp. nov. Rhachis primaria c. 4 mm diametiente; rami primarii vulgo bis furcati, rami ultimi solum foliiferi; rami ordinis primi raro ultra furcam primam proliferati; paleae apicum dormitantium 4 mm longae, basi 0.7 mm latae, apice anguste acuminatae, brunneae, margine basin versus pilis multis patentibus brevibus pallidis, apicem versus setis raris obliquis atrobrunneis, ciliatae.; costae infra paleis atrobrunneis 1.5 mm longis, 0.2 mm latis, margine sparse oblique ciliatis praeditae; rami ordinis primi 4—5 cm longi, ordinis secundi 6-—7 cm longi; rami ultimi 30—35 cm longi, 3.5 cm lati, basis 10—15 mm longa interdum stipitiformis; costulae 4 mm inter se distentes, leviter obliquae; segmenta laminae c. 17 mm longa, supra basin 3 mm lata, tenuia, apice fere integra; venae utrinque leviter prominentes; costulae infra paleis pallidis angustissimis sparse instructae, venae pilos stellifomes laxos pallidos ferentes; sori 3 vel 4 sporangiis constituti. 268 R E I N W A R D T I A [VOL. 4 N. SUMATRA. Res. Tapanuli, Lae Pondon, East of Sidikalang, Alston 14981 (BM, type). This has the general aspect of some varieties of G. hirta, but is larger, and differs in its scales and in the entire apices of the segments of the lamina; also in the lamina being confined to the ultimate branches in almost all cases (occasionally a penultimate branch bears a few scattered lamina-segments). There is apparently no near relative of this species in Sumatra. GLEICHENIA ERECTA C. Chr. Gleichenia ereeta C. Christensen, in Brittonia 2: 269. 1937. Sticherus habbemensis Copel. in Philip. J. Sci. 75: 355, pi. 3. 1941. Copeland's species was based on much larger plants than the type of G. erecta, but as regards size every gradation between the two is repre- sented in various collections. The only differences are that the original G. erecta had lamina-segments at a more acute angle to the eosta, had fewer hairs on the veins and more sparsely fringed scales than the type of Sticherus habbemensis. Large plants, which have the main lateral branches of the frond twice forked, have small stipular leaflets at the bases of the primary branches (these leaflets at first protect the resting apex of the main rhachis) ; such leaflets are lacking on small plants which have the primary branches simple or once forked. Some little-branched plants have unusually long lamina-segments, and often these segments curve towards the costa when dried. GLEICHENIA HIRTA Bl. Gleichenia hirta Blume, En. PI. Jav. 250. 1828. The type of this species was collected by Reinwardt on the island of Tidore (Moluccas). Subsequent collections have shown that ferns of similar general character, differing in various details, occur on mountains almost throughout Malaysia (not in Java nor the Lesser Sunda Islands except Ban). Nine species have been based on these collections, but it seems to me preferable to regard all as varieties of G. hirta. Some are quite clearly distinguishable, but others, especially in the Moluccas and New Guinea, are much less easy to characterize, and I am not at all satisfied with the arrangement here proposed. More data about variation in growth habit in a single locality are needed in order that one may know how to interpret dried specimens. 1957] R. E. HOLTTUM: Gleicheniaceae 269 GLEICHENIA HIRTA Bl. var. amoena (v.A.v.R.) Holtt., stat. nov. Gleichenia, amoena van Alderwerelt van Rosenburgh in Bull. Jard. bot. Buit. II, 23: 12. 1916. — G. peninsidaris Copel. in Univ. Cal. Publ. Bot. 17: 387. 1931. This variety occurs in Sumatra, the Malay Peninsula, and the Lingga Archipelago. It is characterized by large ultimate branches and sparse very narrow costal scales. GLEICHENIA HIRTA Bl. var. amboinensis (v.A.v.R.) Holtt., stat. nov. Gleichenia amboinensis van Alderwerelt van Rosenburgh in Bull. Dep. Agr. Ind. Neerl. 18: 3. 1908. This variety agrees with the typical form of the species in scaliness, but has somewhat smaller ultimate branches. Its chief distinctive character is in the invariable proliferation of the primary branches beyond their first forking; but some plants of the typical form of the species (es- pecially in the Philippines) occasionally show this character. GLEICHENIA HIRTA Bl. var. PALEACEA (Bak.) C. Chr. Gleichenia hirta Bl. var. paleacea (Bak.) C. Christensen in Gard. Bull. Str. S. 7: 212. 1934. — G. vestita var. paleacea Baker in J. Bot. 17: 38. 1879. — G. hallieri Chr. in Ann. Jard. bot. Buit. II, 5: 138. 1905. — G. barbula C. Chr. in Dansk bot. Ark. 9, 3: 67, 1937. This variety, collected in various parts of Borneo at moderate ele- vations on the mountains, is more or less intermediate between var. amoena and the typical form of the species. The scaliness is confined to the costae of fully developed fronds, and old fronds are sometimes almost glabrous. The type collection of G. barbula C. Chr. was unusually large and much- branched; the frond also was not quite fully expanded, so that it does not show the final size of the segments of the lamina. GLEICHENIA HIRTA Bl. var. Candida (Rosenst.) Holtt., comb. nov. Gleichenia Candida Rosenstock in Fedde Rep. 5: 33. 1908. — Sticherus hirtus var. candidus Nakai in Bull. nat. Sci. Mus. Tokyo 29: 19. 1950. In its almost glabrous mature fronds and in its usually persistent pale glaucous lower surface, this variety agrees with var. amoena of Western Malaysia, but is never so large as var. amoena and has broader scales. GLEICHENIA HIRTA Bl. var. omamentalis (Rosenst.) Holtt., stat. nov. Gleichenia omamentalis Rosenstock in Nova Guinea 8: 715. 1912. — G. ornamen- talis var. lanuginosa v.A.v.R. in Nova Guinea 14: 23. 1924. — Sticherus lamianus Copel. in Philip. J. Sc. 75: 356, pi. 6. 1941. 270 REINW ARDT I A [VOL. 4 GLEICHENIA HIRTA Bl. var. virescens (Hieron.) Holtt., comb. nov. Gleichenia dolosa var. virescens Hieronymus in Brause Bot. Jahrb. 56: 209. 1920. — Sticherus hirtus var. virescens Nakai in Bull. nat. Sci. Mus. Tokyo 29: 19.1950. In describing this variety, Brause cited two specimens, both from eastern New Guinea; these two specimens are not alike. I propose to rank the specimen first cited as the type (Ledermann 9935, in herb. Berlin). The second specimen (Schlechter 19842) agrees with others which I have included in G. hirta var. Candida. It should be noted that though Hieronymus gave the name virescens, the collector's note states that the under surface of the lamina was almost white (i.e. glaucous) when fresh; such a glaucous surface can be completely destroyed by heat during drying. This variety approaches the typical form of the species in scaliness, but the scales are more sparsely fringed. It is possible that var. virescens should be united with the typical form of the species. GLEICHENIA HISPIDA Mett. ex Kuhn Gleichenia hispida Mettenius ex Kuhn in Verh. K. K. zool.-bot. Ges. Wien 25: 600. 1875. — G. koordersii Chr. in Ann. Jard. bot. Buit. 15: 76, pi. IS /. 1. 1897. — Sticherus caudatus Copel. in Philip. J. Sci. 75: 354, pi. 2. 1941. — S. pinnatus Copel, in Philip. J. Sci. 83: 98, pi. 3. 1954. I have seen specimens from the type colections of all three species cited as synonyms. Sticherus pinnatus was described from plants with simple fronds such as are produced in the early stages of growth by all species in this subgenus. Apart from small size and lack of branching, they only differ from typical plants of G. hispida in having less rigid lamina-segments which dry almost flat, not with edges recurved. In the characteristic scales, with long flexuous hair-tips, the type collection of S. pinnatus agrees exactly with G. hispida. Mr Alston found once-branched fronds of this species bearing sori in Sumatra, so that simple fertile fronds are not a sharply distinguishing feature. GLEICHENIA MILNEI Bak. Gleichenia milnei Baker, Syn. Fil. 449. 1874 — G. warburgii Chr. in Ann. Jard. bot. Buit. 15: 78 1897. (?). — G. kajewskii Copel. in Philip. J. Sci. 60: 102, pi. 6. 1936. Gleichenia, milnei is closely related to G. truncata (see below), differ- ing in its alternate unequal branching (in which it behaves like some varie- ties of Dicranopteris linearis), in its smaller size of ultimate branches, and in having additional stipular leaflets just above each main fork on the 1957] R. E. HOLTTUM: Gleicheniaceae 271 outside. Gleichenia kajewskii Copel. certainly agrees with Baker's type of G. milnei, but I have not seen the type of G. warburgii, and Christ's description of it is not clear. Specimens from Amboina, named G. war- burgii by v.A.v.R. in the Bogor herbarium are certainly like Baker's species, but I have also seen so named by Christ specimens which are one of the much-branched varieties of Dicranopteris linearis, so that it is possible that the original G. warburgii (from Celebes) was a Dicranopteris. Gleichenia pulchra (Copel.) Holtt., comb. nov. Sticherus pulcher Copeland in Philip. J. Sci. 75: 355, pi. If. 1941. GLEICHENIA TRUNCATA (Willd.) Spr. Gleichenia truncata (Willd.) Sprengel, Syst. Veg. ed. 16, 4: 25. 1827. — Mertensia truncata Willd. in Kongl. vet. Ak. Nya Handl. 25: 169, t. 5 f. a. 1804. — M. laevigata Willd., Sp. PI. 5: 75. 1810. — G. laevigata Hook., Sp. Fil. 1: 10. 1844. — S. myriapoda Nakai in Bull. nat. Sci. Mus. Tokyo 29: 12, /. 1 1950. (?). — G. attenuata Wall. (nom. nud., cited by Nakai I.e. 13). Both M. truncata and M. laevigata were described from specimens from Java; I have seen photographs of these specimens, and do not doubt that they represent the same species, which is by far the commonest of this subgenus in Western Malaysia. The original specimen of M. truncata is immature, the segments of the lamina not fully expanded and so appearing truncate. The species has usually been known by the name laevigata, but truncata is earlier. Gleichenia truncata is a polymorphic species, and three varieties are here distinguished. Sticherus myriapoda Nakai was described from a low elevation in the Riau Archipelago; in this region, near sea level, G. truncata is the only known species of this subgenus. Nakai's detailed description and drawing agree with G. truncata except that he describes and figures the main rhachis of the fronds as climbing by twining, an occurrence other- wise unknown in any member of the family. The main rhachis of high- climbing fronds of G. truncata is often rather regularly curved from one node to the next, and I suggest that perhaps Nakai mistook this for a twining of the rhachis, the drawing being made subsequently from memory. GLEICHENIA TRUNCATA (Willd.) Spr. var. bracteata (v.A.v.R.) Holtt., comb. nov. Gleichenia laevigata var. bracteata van Alderwerelt van Rosenburgh, Handb. Mai. Perns Suppl. 85. 1917. — G. bracteata Bl. ex. Hk. & Bak., Syn Fil. 14. 1865 (as synonym, with description). This is a very distinct variety, apparently common at about 1000 m altitude in Java. It has shorter lamina-segments than the typical form of 272 R E I N W A R D T I A [VOL. 4 the species (less than 2 cm long) and additional stipular leaflets about 1 cm below each main fork of the rhachis. GLEICHENIA TRUNCATA (Willd.) Spr. var. plumaeformis (Presl) Holtt., stat. nov. Mertensia plumaeformis Presl in Abh. (K.) bohm. Ges. Wiss. M.-N. Cl. 5: 338, 1848; Epim. Bot. 24, t, 15. 1851. This agrees with typical G. truncata in size, in scales (as regards shape) and in the shape and nature of the lamina-segments; it differs in having each main rhachis-branch only twice forked, with long ultimate branches (25-35 cm long), and in the copious rather large scales of the costae. The original specimen came from Malacca (Mt Ophir, Cuming 377; in some herbaria erroneously labelled Luzon). Similar specimens have been found on other mountains in the Malay Peninsula and in Sumatra; perhaps also in Borneo, but the Bornean specimens do not have such large scales. Other Peninsula specimens are: G. Tahan, Ridley 15999; G. Ter- bakar, Henderson 10981. GLEICHENIA TRUNCATA (Willd.) Spr. var. involuta Holtt., var. nov. Rami ter furcati; rami ultimi interdum quam ceteros multo longi- ores; furcae ultimae angulos paulo infra 90° formantes; costae supra dense paleaceae, paleae persistentes, c. 1 mm longae, pallide ciliatae; costae infra paleis ferrugineis ciliatis 0.5 mm longis vestitae; margines segmentorum laminae in sicco multo revolutae. N. SUMATRA. Batak lands. J. Winkler (Rosenst., Fil. Sumatr. exsicc. no-182) (L, type; BM, P). GLEICHENIA LOHERI Chr. Gleichenia loheri Christ in Bull. Herb. Boiss. II, 6: 1009. 1906. — Sticherus perpaleaceus Copel. in- Philip. J. Sci. 81: 3. 1952. The type collection of S. perpaleaceus differs from other specimens of G. loheri (all from Luzon) in slightly more abundant scaliness, and in having the primary rhachis-branches 3 times forked; the ultimate and penultimate branches together are about equal in length to the ultimate branches on some other specimens. GLEICHENIA LOHERI Chr. var. major Holtt., var. nov. Rami ultimi 20—30 cm longi, 2.5—3.5 cm lati. N. BORNEO. Kinabalu. Topping 1755 (SING, .US). PHILIPPINES. N e g r o s . Canlaon Volcano, Merrill 92U (P, US). SW. CELEBES. G. Bonthain, 2500 m, Biinnemeijer 11965 (BO, K, L, type) ; 5500—7000 ft, E. G. Smith 222U (BM, K) ; 2000 m, Biinnemeijer 12082 (BO). 1957] R. E. HOLTTUM: Gleicheniaceae 273 This variety comes about half way between typical G. loheri and G. brassii C. Chr. from New Guinea. All agree in character of scaliness. DICRANOPTERIS The Malaysian plants of Dicranopteris have usually all been regarded as varieties of a comprehensive species D. linearis (Burm.) Und., and few of these varieties have been clearly distinguished. I have found that two of the most distinct of them have bilateral spores instead of the tetra- hedral spores of typical D. linearis. These are the varieties named var. tenera and var. pubigera long ago by Blume, and I now regard them as separate species. There is also another species, at present only known from Mount Kinabalu in North Borneo, which has bilateral spores. The rest of Malaysian Dicranopteris I prefer still to regard as varieties of D. linearis, and here describe some of them for the first time. Some varieties are very distinct, others much less so; some of the former will probably be regarded as distinct species when they are better known. The varieties which come nearest to the typical form of the species are var. subferruginea (Hieron) Nakai and var. alternans (Mett.) Holtt., the former in New Guinea and the latter in Sumatra and the Malay Peninsula. The important characters distinguishing these and other varieties are the nature of pubescence and the pattern of branching. Both these characters need more study in the field, especially the latter; a herbarium specimen can never show a whole frond of one of these plants. Dicranopteris speeiosa (Presl) Holtt., comb. nov. Hicriopteris speciosa Presl, Epim. bot. 27. 1851. •— Gleichenia opposita v.A.v.R. in Bull. Jard. bot. Buit. II, 11: 13. 1913; Holtt., Rev. PI. Malaya 2: 70, /. lUf. 1955. — G. parallela Ridl. in J. Mai. Br. Roy. As. Soc. 4: 3. 1926. Presl stated that his species had pinnate branching like that of the subgenus Diplopterygium, but that alternate pinnae had buds in their axils and that at the apex there were two equal pinnae with a bud between them. This is exactly the condition of G. opposita v.A.v.R. (see my figure cited above) and of no other member of the family. Presl also stated that the veins of his specimen were twice forked and that the scales were piliform; these two statements are true of Dicranopteris and of no other section of the family. Presl appears to have considered important his observation that the veins anastomose at the margin of the lamina. In fact the raised veins run to join a thickened margin which is not vascular. By courtesy of the Director of the Botanical Institute of the Charles University, Praha, I have been able to examine the type specimen of 274 R E I N W A R D T 1 A [VOL. 4 Presl's species. It agrees exactly with Gleichenia opposita v.A.v.R., of which I have also seen the type. The locality of origin of Presl's specimen is stated to be Pendschab (Punjab), but that must be wrong, as no member of the family grows in that region. Probably the specimen came from Penang, the only locality at present known where the species is at all common; several early collectors visited Penang. DlCRANOPTERIS PUBIGERA (Bl.) Nakai Dicranopteris pubigera (Bl.) Nakai in Bull. nat. Sci. Mus. Tokyo 29: 68. 1950. — Gleichenia hermannii var. pubigera BL, En. PI. Jav. 249. 1828. This species, which occurs abundantly on mountain in Sumatra and Java, and as far eastwards as Flores, has bilateral spores. It differs from D. curranii Copel in its smaller size and in its usually persistent hairiness; from D. clemensiae in its shorter more coriaceous and differently hairy ultimate branches. There appear to be two forms of this species, but I cannot see that a sharp line can be drawn between them. One form has very thick coriaceous fronds (drying very rigid with reflexed margins), the veins when dry distinctly grooved on the lower surface; old fronds of this are often almost glabrous though younger ones are copiously hairy on the costules of the lower surface. The other form has smaller ultimate branches with a thinner lamina which dries nearly flat, the veins not grooved on the lower surface and the costules more persistently hairy. Some specimens however are intermediate between these two extremes. Mettenius (in Ann. Mus. bot Lugd.-Bat. 1: 50. 1863) confused the thick rigid form of this species with G. hermannii var. rigida Bl., which has tetrahedral spores, veins not grooved and a quite glabrous lamina. See below for a further note on this variety. DICRANOPTERIS CURRANII Copel. Dicranopteris curranii Copeland in Philip. J. Sci. 81: 4. 1952. — Gleichenia hermannii var. tenera BL, En. PI. Jav. 249. 1828. — G. hermannii var. venosa BL, I.e. — G. dichotoma var. malayana Chr. in Ann. Jard. bat. Buit. 15: 77. 1898. — G. linearis var. malayana v.A.v.R., Handb. Mai. Ferns 59. 1907; Holtt., Rev. PL Malaya 2: 70. 1955. This very large species, common throughout the lowlands of Western Malaysia and occurring also in the Philippines, has bilateral spores like those of D. pubigera; this fact has not hitherto been reported. Gleichenia weatherbyi Fosb. described from the Caroline Islands (in Am. Fern Journ. 40: 140. 1950), is rather like D. curranii but even larger; it also has bilateral spores. . . 1957] R. E. HoLTTUM: Gleicheniaceae Dicranopteris clemensiae Holtt., sp. nov. 27S Habitu et sporis bilateralibus D. curranii Copel. congruens, differt ramis ultimis angustioribus, costis infra copiose pilis rigidis nitidis atro- rubris vestitis. Rhachis primaria 7 mm vel ultra diametro; rami primarii bis aeque furcati, prope furcas pilis rigidis atrorubris persistentibus vestiti; rami ultimi 50—55 cm longi, 5—6 cm lati, segmenta infima ex- teriora interdum elongata et plus minusve lobata; rami adjuncti furcarum penultimarum 16—18 cm longi et 5 cm lati; foliola stipuliformia furcarum rhacheos primariae 3 cm longa, lobi 3 mm lati; costulae ramorum ulti- morum 5—6 mm inter se distantes; segmenta laminae c. 3.5 mm lata, tenuia, in sicco fragilia; venae vulgo ter furcata (ramus primus basi ipse venae egrediens) utrinque leviter prominentes; costae infra pilis crassis atrorubris nitidis pluricellularibus 1—1.5 mm longis basi ramos breves . ferentibus primo omnino vestitae, denique glabrescentes et basibus pilorum asperulae; costulae infra pilis ferrugineis (non nitidis) leviter crispis et implicatis basi solum pauciramosis instructae; sori sporangiis 10—15 constati; vena unica ramosa saepe soros duos ferens (sori in ramulis extremis sedentes); sporae bilaterales, c. 32 x 16 µ,. N. BORNEO. Mt Kinabalu, Tenompok, alt. 5000 ft (1600 m), Clemens 28745 (BM, type; K, L, US). Only known from one collection. DICRANOPTERIS LINEARIS (Burm.) Und. Dicranopteris linearis (Burm.) Underwood in Bull. Torr. bot. Club. 34: 249. 1907. — Polypodium lineare Burman, Fl. Ind. 235, t. 67 f. 2. 1768. The following are new varieties which I have recognized during my revision of the family for Flora Malesiana, and varieties recognized by former authors for which name-changes are necessary. A special group of such varieties are distinguished by having always accessory branches at the ultimate forkings of the frond. Though these varieties agree in this particular character, they differ so much in other ways that they must be regarded as having arrived at this form of branching on different evolutionary lines. DICRANOPTERIS LINEARIS (Burm.) Und. var. demota Holtt., var. nov. Furcae ramorum frondium alterne inaequales; furcae ultimae ramis adjunctis praeditae; rami adjuncti furcarum infimarum infra furcam egredientes; rami ultimi ad 18 cm longi, 6 cm lati, apicem versus sensim angTastati; costulae inter se 5 mm distantes; lamina tenuis, venae supra distincte, infra leviter prominentes; costae, costulae, laminaque infra glabrae. 276 R E I N W A R D T I A [VOL. 4 1957] R. E. HOLTTUM: Gleieheniaceae 277 N. BORNEO. Mt Kinabalu, 5000 ft, Clemens 29535 (BM, K, type; L, U S ) ; Holttum S.F. 25US8 (K, S I N G ) ; Topping 1821 (US). — N E W GUINEA. D u t c h N o r t h N. G. Bernhard Camp, Idenburg River, 1700 m, Brass 12319 (BM, L ) . P a p u a . Boridi, 6000 ft, Carr U57U (BM, K, L, SING). The accessory branches, especially the lower ones, are attached dis- tinctly below the forks. DICRANOPTERIS LINEARIS (Burm.) Und. var. montana Holtt., var. nov. Rami primarii plurime aeque furcati; rami ultimi 15—25 cm longi, 3.5—6 cm lati, apicem versus sensim angustati; furcae ultimae ramis adjunctis, quam ramos ultimos duplo minoribus, praeditae; costulae 6—7 mm inter se distantes; lamina coriacea, subtus glabra et glauca; venae utrinque prominentes. MALAYA PENINSULA. P e r a k, Maxwell's Hill, 4000 ft, Molesworth Allen 2720 (SING, type). This variety was described in English in my Ferns of Malaya, but the description did not validate publication of the name. The variety is very widely distributed, occurring in Sumatra, Java, Borneo, Ternate, Ceylon and S. India, and in Sikkim. Some specimens from outside Malaya are smaller than the type, with costules only 5 mm apart, and somewhat less coriaceous. DICRANOPTERIS LINEARIS (Burm.) Und. var. altissima Holtt., var. nov. Ehachis primaria interdum alte scandens, ad 10 mm diametiente; rami primarii aeque quater vel quinquies furcati, rami ultimi c. 3 mm supra basin ramis adjunctis praediti; rami ultimi ad 15 cm longi et 3 cm lati; lamina tenuis, subtus glauca et pilos ferrugineos paucos ferens; costulae inter se 3.5—4.5 mm distantes; venae supra prominentes, subtus fere planae. MALAY PENINSULA. J o h o r e , Mawai (Sungei Sedili), Corner S.F. 31W<~ (K, L, SING, type). I refer also to this variety specimens from Palawan, Luzon, New Guinea, Talaud Islands, and Solomon Islands. Among them are specimens distributed by Cuming under his number 270; one of these received the MS name Gleichenia dichotoma var. divaricata, published as a nomen nudum by Moore (Index, p. 377). Cuming also distributed under number 270 specimens of the typical form of D. linearis and apparently one of these was the type of Mertensia pteridifolia Presl (figured in Epim. Bot. t. 14). The name var. altissima, with description in English, was published in my Ferns of Malaya. DICRANOPTERIS LINEARIS (Burm). Und. var. rigida (Bl.) Holtt., comb. nov. Gleichenia hermannii R. Br. var. rigida Blume, En. PI. Jav. 249. 1828. — Mertensia crassifolia Presl in Abh. (K.) bohm. Ges. Wiss. M.-N. Cl. V, 5: 339. 1848. — G. dichotoma var. rigida Mett. in Ann. Mus. bot. Lugd.-Bat. 1: 50. 1863, p.p. The type of Blume's variety was a specimen from Tidore (Reinwardt). In my opinion it does not differ from the specimens of Cuming number 136, from Luzon, one of which was the type of Mertensia crassifolia Pr. Mettenius also ascribed specimens from Java to this variety, but I think he confused it with D. pubigera. DICRANOPTERIS LINEARIS (Burm.) Und. var. latiloba Holtt., var. nov. Rami primarii ad quater furcati; rami ultimi ad 25 cm longi et 9 cm lati, segmenta exteriora laminae interdum elongata et deflexa sed non pinnatifida; costulae 7—8 mm inter se distantes; segmenta laminae 4—5 mm lata, tenuia, subtus glabra, margines in sicco leviter reflexae; venae utrinque paulum prominentes; rami adjuncti ad 20 cm longi et 6—8 cm lati; foliola stipuliformia furcarum infimarum 4 cm longa, late lobata. PHILIPPINES. L u z o n . Benguet Subprov., Merrill 975 (US, type). This appears to be a distinct variety, occurring throughout the Philippines; specimens from N. Celebes (Alston 15797) and from the Caroline Islands (Ledermann 13609a) also appear referable to it. The distinct features are the very wide spacing of the costules, the broad ultimate and accessory branches, and the glabrous rather thin lamina. DICRANOPTERIS LINEARIS (Burm.) Und. var. dichotoma (Thunb.) Holtt., stat. nov. Potypodium dichotomum Thunberg, Fl. Jap. 338, t. 37. 1784. The specimens of D. linearis which I have seen from Japan, and some from China, differ from the typical form of the species (described from Ceylon) in the thinner lamina with the veins distinctly prominent on the upper surface. I can see no other clear difference, and believe that the Japanese fern should rank as a variety of D. linearis. It appears that the fronds die in the winter, whereas in typical D. linearis they persist inde- finitely; the texture of the lamina is perhaps related to this behaviour. DICRANOPTERIS LINEARIS (Burm.) Und. var. subpectinata (Chr.) Holtt., comb. nov. Gleichenia subpectinata Christ in Bot. Tidsskr. 24: 111. 1901. — G. pteridifolia Ridl. in 3. Mai. Br. Roy. As. Soe. 4 : 4. 1926. (non Cesati). — G. linearis var. 278 R E I N W A R D T I A [VOL. 4 alternans sensu Holtt., Rev. Fl. Malaya 2: 70. 1954 (non Mett.). — D. ivarburgii sensu Nakai in Bull. nat. Sci. Mus. Tokyo 29: 70. 1950 {non Chr.). DiCRANOPTERIS LINEARIS (Burm.) U n d . v a r . ferruginea (Bl.) Holtt., comb. Tiov. Gleichenia ferruginea Blume, En. PI. Jav. 249. 1828. — D. ferruginea Copel. in Philip. J. Sci. 75: 349. 1941. — G. linearis var. fermginea Racib., Fl. Buit. Pterid. 13. 1898. DiCRANOPTERIS LINEARIS (Burm.) Und. v a r . subspeciosa Holtt., var. nov. Habitu var. alternante (Mett.) congruens; costulae subtus pilis ten- uissimis pallidis implicatis persistentibus vestitae, pili crassiores nulli; venae supra prominentes, confertae. N. BORNEO. Mt Kinabalu, Kiau, Topping 1516 (SING, US, type). Ths variety agrees exactly in characters of lamina and pubescence with Gleichenia opposita v.A.v.R. (Hicriopteris speciosa Presl), but in branching it agrees with var. alternans (Mett.) Holtt. and var. subpecti- nata (Chr.) Holtt. It has been found in the Malay Peninsula, Sarawak, North Borneo, Luzon, Mindoro and Mindanao, and possibly in New Guinea. DiCRANOPTERIS LINEARIS (Burm.) Und. var. inaequalis (Rosenst.) Holtt., comb. nov. Gleichenia linearis var. inaequalis Rosenstock in Fedde, Rep. 13: 212. 1915. DiCRANOPTERIS LINEARIS (Burm.) Und. v a r . alternans (Mett.) Holtt., comb. nov. Gleichenia dichotoma var. alternans Mettenius in Ann. Mus. bot. Lugd.-Bat. 1: 51. 1863. — G. linearis var. alternans v.A.v.R., Handb. Mai. Ferns Suppl. 84. 1917. INDEX TO GLEICHENIACEAE New published names are in bold face type; synonyms are in italics; pages where descriptions are given, are in bold face type. Acropterygium Nakai 260, 261. Calymella Presl 261, 262; sect. Eucaly- mella Ching 262; sect. Gleicheniastrum (Pr.) Ching- 262. Dicranopteris Bernhardi 258, 259, 260, 261, 266; sub gen. Acropterygium (Diels) Holtt. 261; subgen. Dicrano- pteris 261; clemensiae Holtt. 274, 275; curranii Copel. 274, 275; dichotoma (Thunb.) Bernh. 261; ferruginea Copel. 278; linearis (Burm.) Und. 260, 270, 271, 273, 275, 276, 277, var. alternans (Mett.) Holtt. 273, 278, var. altissima 1957] R. E. HOLTTUM: Gleicheniaceae 279 Holtt. 276, var. demota Holtt. 275, var. dichotoma (Thunb.) Holtt. 277, var. ferruginea (Bl.) Holtt. 278, var. inae- qualis (Rosenst.) Holtt. 278, var. mon- tana Holtt. 276, var. latiloba Holtt. 277, var. pubigera Bl. 273, var. rigida (Bl.) Holtt. 277, var. subferruginea (Hieron.) Nakai 273, var. subpectinata (Chr.) Holtt. 277, 278, var. subspeciosa Holtt. 278, var. tenera Bl. 273; pubi- gera (Bl.) Nakai 274, 277; speciosa (Presl) Holtt. 273; warburgii sensu Nakai (non Christ) 278. Dicranopteris sensu Nakai 261. Dicranopteris sensu Underw. 261. Diplopterygium Nakai 260, 261. Gleiehenella Ching 261. Gleichenia Sm. 258, 259, 260, 261, 262; subgen. Dicranopteris (p.p.) Bower 261; subgen. Diplopterygium (Diels) Holtt. 259, 261, 262; subgen. Eudicranopteris Bower 261; subgen. Eugleichenia Diels 261; subgen. Gleichenia 259, 261, 262, 266; subgen. Mertensia (Willd.) Diels 259, 261, 266, sect. Acropterygium Diels 261, sect. Diplopterygium Diels 261, sect. Heteropterygium. Diels 261, sect. Holopterygium Diels 261; alstonii Holtt. 267; amboinensis v.A.v.R. 269; amoena v.A.v.R. 269; angustiloba Holtt. 263; attenuata Wall, ex Nakai 271; barbula C. Chr. 269; bracteata Bl. ex Hook. & Bak. 271; brassii C. Chr. 273; brevipinnula Holtt. 263, 264; Candida Rosenst. 269; diearpa R. Br. 258; di- chotoma var. alternans Mett. 278, var. divaricata Moore MS. 276, var. ma- lay ana Chr. 274, var. rigida Mett. 277; dolosa var. virescens Hieron. 270; erecta C. Chr. 267, 268; ferruginea Bl. 278; glauca (Thunb.) Hook, 261; hal- Heri Chr. 269; hermannii var. pubi- gera Bl. 274, var. rigida Bl. 274, 277, var. tenera Bl. 274, var. venosa Bl. 274; hirta Bl. 267, 268, var. amboinen- sis (v.A.v.R.) Holtt. 269, var. amoena (v.A.v.R.) Holtt. 269, var. Candida (Rosenst.) Holtt. 269, 270, var. orna- mentalis (Rosenst.) Holtt. 269, var. paleacea (Bak.) C. Chr. 269, var. vi- rescens (Hieron.) Holtt. 270; hispida Mett. ex Kuhn 267, 270; kajewskii Copel. 270, 271; koordersii Chr. 2.70; laevigata (Willd.) Hook. 271, var. bracteata v.A.v.R. 271; linearis var. alternans (Mett.) v.A.v.R. 278, var. ferruginea Racib. 278, var. inaequalis Rosenst. 278, var. malayana v.A.v.R. 274; loheri Chr. 272, 273, var. major Holtt. 272; longissima Bl. 263, 266; matthewii Holtt. 265; microphylla R.Br. 258, 262; milnei Bak. 259, 271, 272; norrisii Mett. 263; novoguineensis Brause 263; opposita v.A.v.R. 273, 274, 278; ornamentalis Rosenst. 269, var. lanuginosa v.A.v.R. 269; paleacea (Copel.) Holtt. 265, 266; papuana Holtt. 266, var. membranacea Holtt. 266, var. papuana 266; parallela Ridl. 273; pel- tophora Copel. var. schizolepis C. Chr. ex Holtt. 262; peninsularis Copel. 269; polypodioides (L.) Sm. 261; pteridifo- lia sensu Ridl. (non Cesati) 277; pul- chra (Copel.) Holtt. 271; sordida Copel. 265; subpectinata Chr. 277; sumatrana Holtt. 264; truncata (Willd.) Spr. 261, 267, 270, 271, 272, var. bracteata (v.A.v.R.) Holtt, 271, var. involuta Holtt. 272, var. plumaeformis (Presl) Holtt. 272; vestita Bl. 267, var. palea- cea Bak. 269; volubilis Jungh. var. peninsularis Holtt. 265; vulcanica Bl. 258; warburgii Chr. 270, 271; weather- byi Fosb. 274. Gleichenastrum Presl 261. Subfam. Gleichenioideae Nakai 261. Hicriopteris (non Presl) Ching, Copel. 259, 261. Hicriopteris Presl 258, 259, 261; astro- tricha Copel. 263; paleacea Copel. 265; speciosa Presl 259, 273, 278. Mertensia Roth 266. Mertensia Willd. 261, 266; crassifolia Presl 277; laevigata Willd. 271; pecti- nata Willd. 261; plumaeformis Presl 280 R E I N W A R D T I A [VOL. 4 272; pteridifolia PresL 276; truneata Willd. 261, 271. Platyzoma R. Br. 257. Polypodium dichotovvum Thunb. 261, 277; glaucum, Thunb. 261; lineare Burm. 261, 275. Subfam. Sticherioideae Nakai 260. Sticherus Presl 258, 259, 260, 261, 266; caudatus Copel. 270; habbemensis Copel. 268; hirtus var. candidus Nakai 269, var. virescens Nakai 270; lami- anus Copel. 269; myriapoda Nakai 271; perpaleaceus Copel. 272; pinnatus Co- pel. 267, 270; jmlcher Copel. 271. Stromatopteris Mett. 257, 260, 261; mo- niliformis Mett. 261. Subfam. Stromatopteroideae Nakai 261. ADDENDUM In Gleichenia subgenus Diplopterygium the following species must be placed: Gleichenia deflexa Holtt., sp. nov. Rami rhacheos 120 cm vel ultra longi, maximi 35 cm lati, rami mi- nores (steriles) interdum 18 cm lati; pinnulae omnes angulo c. 75° deflexae; pinnulae maximae 16—20 cm longae, 3.5—3.8 cm latae, costae 4—5 cm inter se distantes (costae ramorum minorum sterilium 2.5—3.5 cm inter se distantes) ; costulae 4—5 mm inter se distantes, medii rectangulariter patentes, infimae interdum leviter deflexae; lamina tenuis, segmenta c. 20 paria basi leviter constricta (i.e., supra basin ampliata), ala angustissima laminae conjuncta, segrnentum infimum foliolum disjunctum constatum; venae utroque basi prominentes, cetera planae; sori sporangiis 3—5 con- stati; f oliola stipulif ormia.ad 4 cm longa, bipinnatif ida, segmenta 1 mm lata. Paleae apicis rhacheos c. 10 mm longae, haud 1 mm latae, pallidae, margine brunneo setis brevibus obliquis nitidis praeditae; rami rhacheos costaeque utroque primo paleis parvis et pilis stelliformibus coarctis paleis elongatis angustis intermixtis vestiti, demum plus minusve glabri et verruculosi; venae infra pilis stelliformibus pallidis adspersis praeditae. NEW GUINEA. P a p u a . Fergusson Island, 800 m, climbing to 6 m in rather open rain forest in steep ravine, Brass 27171 (L, typs). Normanby Island, Mt Pali- nama, 850 m, scrambling to 7 m in tall mossy forest, Brass 25736 (L). R E I N W A R D T I A Published by Herbarium Bogoriense, Kebun Raya Indonesia Volume 4, Part. 2, pp. 281-310 (1957) THE GENUS FIRMIANA MARSILI (Sterculiaceae) A. J. G. H. KOSTERMANS * SUMMARY 1. Firmiana Marsili and Erythropsis Lindley are congeneric. 2. Firmiana Marsili links Sterculia L. and Scaphium Schott & Endl. 3. Eight species are recognized: Firmiana eolorata (Roxb.) R. Br.; F. diversifolia A. Gray; F. fidgens (Wall, ex King) Corner; F. hainanensis Kosterm.; F. major Hand.-Mazz.; F. papuana Mildbr.; F. philippinensis Kosterm.; F. simplex (L.) W. F. Wight. 4. Seventeen binomials are referred to other genera. While a paper on Firmiana was in the press, I received some addition- al, important information from the Kew Herbarium and from Mr. J. E. Dandy of the British Museum. These "Additional Notes" were placed at. the end of the article. In a subsequent letter that reached me when the article already had been printed and issued**, the problem of Firmiana simplex was solved, thanks to the tenacity of Mr. Dandy. Moreover, F. fulgens came into flower in the Bogor Botanical Garden, which enabled me to add some field notes and photographs. We had in mind to incorporate the paper in "Reinwardtia" in its original form, but due to the additions, mentioned above, it was considered better that it should be rewritten to this new, separate publication. F I R M I A N A * Marsili Firmiana Marsili in Saggi Scientifici e Letterari dell' Accademia Padova 1: 106-116, tab. 1 & 2. 1786; Lamarck-Podret, Encycl. meth. 7: 432. 1806 (as a syn. under Sterculia platanifolia L.f.; excl. syn. Culhamia Forsk.) ; Steudel (Firmiana Medic.), Nomencl. 343. 1821; ed. 2, 1: 642. 1840 (as a syn. cf Sterculia L.) ; DC, Prodr. 1: 481. 1824 (as a syn. of Sterculia L.); Bartling, Ordin. 340. 1831 (as a syn. of Sterculia L.) ; Schott & Endlicher, Melet. bot. 33. 1832; Spach, Hist. Veg. phan. ' D. Sc.j Botanist, Forest Service of Indonesia; cooperator Herbarium Bogoriense. * Published under the title "The Genus Firmiana Marsili" as Communication no. 54̂ of the Forest Research Institute, Indonesia, in December 1956, p. 3-33. *** Named for count K. J. von Firmian, born 1716 at Deutsch-Metz in Tirol, died 1782 Milan, an Austrian statesman who was Governor of Lombardy under Maria Theresia. — 281 — Binder15 Rein.Vol 4,Part 2,pp 119-310_Page_01 Rein.Vol 4,Part 2,pp 119-310_Page_71 Rein.Vol 4,Part 2,pp 119-310_Page_72 Rein.Vol 4,Part 2,pp 119-310_Page_73 Rein.Vol 4,Part 2,pp 119-310_Page_74 Rein.Vol 4,Part 2,pp 119-310_Page_75 Rein.Vol 4,Part 2,pp 119-310_Page_76 Rein.Vol 4,Part 2,pp 119-310_Page_77 Rein.Vol 4,Part 2,pp 119-310_Page_78 Rein.Vol 4,Part 2,pp 119-310_Page_79 Rein.Vol 4,Part 2,pp 119-310_Page_80 Rein.Vol 4,Part 2,pp 119-310_Page_81 Rein.Vol 4,Part 2,pp 119-310_Page_82 Rein.Vol 4,Part 2,pp 119-310_Page_83